Predictors of surgical outcome in medically-resistant temporal lobe epilepsy with bilateral features on pre-operative evaluation

Accepted Manuscript Title: Predictors of surgical outcome in medically-resistant temporal lobe epilepsy with bilateral features on pre-operative evaluation Author: Hena Waseem M.D. Katie E. Osborn M.A. Mike R. Schoenberg Ph.D. Valerie Kelley R.N. Ali M. Bozorg M.D. Selim R. Benbadis M.D. Fernando L. Vale M.D. PII: DOI: Reference:

S0303-8467(15)30047-0 http://dx.doi.org/doi:10.1016/j.clineuro.2015.10.016 CLINEU 4211

To appear in:

Clinical Neurology and Neurosurgery

Received date: Revised date: Accepted date:

19-1-2015 9-8-2015 11-10-2015

Please cite this article as: Waseem H, Osborn KE, Schoenberg MR, Kelley V, Bozorg AM, Benbadis SR, Vale FL, Predictors of surgical outcome in medically-resistant temporal lobe epilepsy with bilateral features on pre-operative evaluation, Clinical Neurology and Neurosurgery (2015), http://dx.doi.org/10.1016/j.clineuro.2015.10.016 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Bilateral Features Highlights

•

Ac ce pt e

d

M

an

us

cr

• •

We evaluate prognostic factors for favorable temporal resection outcomes in bilateral cases Lateralized neuropsychological and Wada test were significantly associated with outcomes Seizure reduction and freedom is possible in patients with bilateral features Temporal resection should be considered in patients with lateralized preoperative exams

ip t

•

Page 1 of 27

Title: Predictors of surgical outcome in medically-resistant temporal lobe epilepsy with bilateral features on pre-operative evaluation

cr

ip t

Authors’ names: Hena Waseem M.D.1 Katie E. Osborn M.A. 2 Mike R. Schoenberg Ph.D. 1-3 Valerie Kelley R.N. 3 Ali M. Bozorg M.D. 3 Selim R. Benbadis M.D.3 Fernando L. Vale M.D.1

an

us

Affiliations: 1Department of Neurosurgery and Brain Repair Morsani College of Medicine, University of South Florida 2 Tampa General Circle, USF Health, 7th Floor, Tampa, Florida, 33606, U.S.A. [email protected] [email protected] 2Department of Psychiatry and Behavioral Neurosciences

Ac ce pt e

3Department of Neurology

d

M

Morsani College of Medicine, University of South Florida 3515 East Fletcher Ave., Tampa, Florida, 33613 [email protected] [email protected]

Morsani College of Medicine, University of South Florida 2 Tampa General Circle, USF Health, 6th Floor, Tampa, Florida, 33606, U.S.A. [email protected] [email protected] [email protected] Corresponding Author: Fernando L. Vale, M.D. Professor and Vice Chair Residency Program Director Department of Neurosurgery and Brain Repair Morsani College of Medicine, University of South Florida 2 Tampa General Circle, USF Health, 7th Floor, Tampa, Florida, 33606, U.S.A. Tel: +1-813-259-0605; Fax: +1-813-259-0944 E-mail address: [email protected]

Page 2 of 27

Ac ce pt e

d

M

an

us

cr

ip t

Keywords: Intractable epilepsy seizure disorder epilepsy surgery temporal lobe epilepsy bitemporal Neuropsychology

Page 3 of 27

Abstract

Ac ce pt e

d

M

an

us

cr

ip t

Objective: This study identifies potential prognostic factors for favorable anterior mesial temporal lobe (AMTL) resection outcomes in patients with medically refractory temporal lobe epilepsy (TLE) with bilateral features on pre-operative examination. Methods: Thirty-one patients demonstrated bilateral features defined as: bilateral independent temporal or bitemporal ictal onsets on surface or intracranial EEG, or bitemporal interictal epileptiform abnormalities on surface EEG with bilateral radiographic mesial temporal sclerosis. Surgical outcomes were classified according to reduction in seizure frequency: I (100% reduction), II (>= 75% reduction), III (50-74% reduction), IV (<50% reduction). Results: Of 31 patients, 14 (45%) improved to class I and 9 (29%) had a class II outcome at an average of 4 years after surgery. Eight (26%) patients did not exhibit good surgical outcome (class III, class IV). We found that neuropsychological and Wada memory scores were significantly correlated (p<0.05) with surgical outcome, and logistic regression found neuropsychological evaluation significantly predicted better surgical outcome (p<0.05). Conclusions: When bilateral features are present on pre-operative evaluation, neuropsychological and Wada test results can provide unique data to better identify those patients more likely to achieve substantial seizure reduction.

Page 4 of 27

Introduction Surgical resection in patients with medically intractable temporal lobe epilepsy (TLE) can achieve seizure freedom in cases that permit localization of a

ip t

discrete epileptogenic site[1-3]. The value of pre-operative evaluations in the identification of the site of seizure origin is well demonstrated[1]. Various measures

cr

of lateralization have been correlated with better surgical outcomes, with the

us

clearest prognostic factor being the identification of a lateralized side of seizure origin on EEG[4].

an

Patients presenting with bilateral features on pre-operative testing are not ideal candidates for temporal lobe resection, and are presumed to have worse

M

surgical outcomes due to the difficulty of identifying a unilateral seizure origin, or due to the presence of a genuinely bitemporal seizure origin [5, 6]. A diagnosis of

d

bitemporal lobe epilepsy is ideally made using intracranial EEG monitoring [7].

Ac ce pt e

However, this procedure is not without risks, and even in those patients who receive intracranial monitoring, determination of lateralization may not be absolute [8].

A number of investigations have demonstrated that seizure reduction and

indeed seizure freedom is possible in patients with indeterminate or persistent bilateral features on EEG monitoring [5, 6, 9-13]. However, outcomes are less uniform than in unilateral TLE [5, 6]. The purpose of this study is to identify potential secondary prognostic factors encountered during pre-operative examination that may characterize those cases most likely to benefit from surgical

Page 5 of 27

intervention. A better definition of the effectiveness of surgical intervention in these patients may allow a more inclusive selection of surgical candidates to be identified.

Materials and Methods

ip t

Patient selection

cr

A retrospective review of 501 patients who underwent temporal resection

for the treatment of epilepsy from 1998 to 2013 was performed. Patient data were

us

collected in a prospective epilepsy surgery database established in 1998 after

an

approval from the Institutional Review Board. Informed consent was obtained from all patients. Patients were selected if they demonstrated mesial temporal clinical

M

semiology [14-16] with: (1) bilateral independent ictal onset of temporal origin on surface or intracranial EEG, (2) bitemporal synchronous (non-lateralized) ictal

d

onset on surface or intracranial EEG, or (3) bitemporal interictal epileptiform

Ac ce pt e

abnormalities on surface EEG with bilateral radiographic mesial temporal sclerosis (MTS) (if ictal onset was non-localized). Patients were excluded if they did not undergo temporal resection, or if data indicated exclusively extra-temporal seizure activity, or if seizures were a component of a widespread, congenital neurological disease. Thirty-one patients met the inclusion criteria. All pre-operative findings were discussed at a multi-disciplinary epilepsy

conference. Pre-operative evaluation included a history and physical examination, surface video-EEG long-term monitoring (LTM), magnetic resonance imaging (MRI), Wada testing, positron emission tomography (PET) scan, and neuropsychological examination. Intracranial EEG recording was considered if any of the following was present in the setting of typical temporal semiology: (1) poorly localized surface

Page 6 of 27

EEG ictal onset, (2) inadequate surface EEG ictal recordings due to extensive muscle artifacts, or (3) multiple discordant or inconclusive results among other preoperative tests. Patient pre-operative evaluation protocols have been previously

ip t

described [17]. All patients in this study were pharmaco-resistant, had debilitating seizures with bilateral EEG features, and limited treatment options. These patients

cr

were non-traditional candidates for epilepsy surgery. In these situations, evaluation

us

is usually focused on palliative outcomes.

Selection criteria for surgical resection involved focusing on the predominant

an

area of electrographic seizure activity that correlated with the more clinically significant seizure semiology allowing identification of the most clinically significant

M

lobe. All patients underwent MRI study and most completed a FDG PET study to evaluate for structural and/or metabolic abnormalities, respectively.

d

Neuropsychological evaluation and Wada study were requested for all patients, to

Ac ce pt e

evaluate for functional deficits and assist in predicting functional outcomes. However, some pre-operative tests were not performed because of patient financial constraints, refusal, or non-compliance due to a young age or cognitive deficits. Though invasive monitoring is the gold standard for bitemporal epilepsy, it is not without risks. Due to these concerns some patients decided to undergo surgical resection and forego an invasive evaluation with the understanding of the potential palliative benefit and risks of the surgery.

Surgery and outcome classification

Page 7 of 27

Details of the surgical technique have been previously described [18]. All patients underwent a selective anterior mesial temporal lobe (AMTL) resection (20 right, 11 left) through an inferior temporal gyrus approach. Sixteen patients

ip t

underwent invasive recordings prior to surgery and 15 patients elected to move forward with surgical resection and forego invasive evaluation. Eight of the 31

cr

patients had a history of vagus nerve stimulator (VNS) placement prior to

us

surgery. All patients with prior VNS implantation waited at least one year before a resection was performed. All patients had surgical resection performed by a

an

single neurosurgeon (F.L.V.).

Surgical outcomes were classified according to reduction in seizure

M

frequency: I (seizure freedom, 100% reduction), II (rare, almost seizure free, >= 75% reduction), III (worthwhile improvement, 50-74% reduction), IV (no

d

worthwhile improvement, <50% reduction) [5, 11]. This classification has been

Ac ce pt e

previously used in bitemporal patient populations [5, 11]. For some analyses, outcome was dichotomized as “favorable” (class I or II) or “poor” (class III or IV), which reflects common clinical assessment in epilepsy surgery of a successful or failed surgical resection [9, 10].

Surface and intracranial EEG monitoring Surface video-EEG LTM (International 10-20 system) was conducted using

XLTEK (Oakville, Ontario, Canada). Bilateral basilar temporal placements such as T1, T2 electrodes were routinely used. Sphenoidal electrodes were not used. Intracranial EEG recording was obtained using subdural contact electrode placement along the basal and lateral neocortex.

Page 8 of 27

Seizure totals included both non-localized seizures, where a lobe was not well identified, as well as non-lateralized seizures, including bilateral ictal onsets based on long-term video surface EEG. Two distinct variables derived from EEG

ip t

monitoring were evaluated for prognostic potential: (1) percent ictal lateralization, defined as: (greater number of seizures recorded in a single temporal lobe/total

cr

number of seizures) x 100 = % of lateralization[6, 13, 16, 19]. (2) Ictal lateralization

us

score (ILAT score), defined as:

| %ipsilateral seizure - %contralateral seizure | - (%non-localized + %non-

an

lateralized), where ipsilateral is defined as the side of resection, and contralateral the side opposite resection. ILAT score attempts to better account for non-specific

M

seizures compared with percent ictal lateralization, with a higher ILAT score representing more strongly lateralized findings (highest theoretical ILAT score=1.0

d

would reflect 100% seizures that were ipsilateral with no contralateral or non-

Ac ce pt e

localized/lateralized seizures. The lowest theoretical score would be -1.0 with no ipsilateral seizures and 100 percent of nonlocalizable/nonlateralizable seizures). MRI scan

Patients received imaging with high-resolution MR using a 1.5 or 3.0-T

magnet with special focus on the temporal lobes. Coronal and axial T2-weighted and FLAIR images were obtained and reviewed by a board-certified neuro-radiologist. Abnormal signal on FLAIR and T2-weighted images, decreased volume, and loss of anatomical configuration of the hippocampal formation are considered the hallmarks of radiographically-confirmed MTS [20]. MRI results were classified as ipsilateral MTS, contralateral MTS, bilateral MTS, or no evidence of MTS.

Page 9 of 27

FDG-PET Scan A fluorodeoxyglucose (FDG) PET brain scan was performed to evaluate for a focal abnormality in regional cerebral metabolic rate (rCMR) in the mesial temporal

ip t

lobe. An FDG–PET scan was considered positive when it showed an area of hypometabolism in the ipsilateral hemisphere. PET results were classified as

cr

unilateral dysfunction, bilateral dysfunction, or no evidence of dysfunction.

us

Wada study

Wada testing to evaluate short-term memory asymmetry and language

an

dominance was performed, and a recognition memory asymmetry of > ⅜ was considered lateralized [21]. The procedure was performed according to the protocol

M

of Loring et al., with a methohexital adaptation[22]. Memory score is the total correct recognition score of materials presented minus 0.5 points for any false

d

positive errors (highest raw recognition recall memory = 8). There are 8 foils, and

Ac ce pt e

the greatest negative score is -4 (total correct=0-4 = -4). Two distinct variables derived from Wada study results were evaluated for prognostic potential: (1) Wada memory asymmetry score, defined as the absolute value of the difference in recognition memory scores obtained after right and left injections, (2) Wada dysfunction concordance, classified as ipsilateral dysfunction (concordant with side of resection), contralateral dysfunction (not concordant with side of resection), or no evidence of asymmetry. Neuropsychology study Neuropsychological evaluation included assessment of task engagement using performance validity tests, intellect, attention/executive functioning,

Page 10 of 27

language, learning and memory, visuospatial/visuoconstruction, processing speed, and mood/anxiety symptoms. Tests were administered and scored based on respective manuals and standardized scores were employed in interpretation and

ip t

statistical analysis. These tests were evaluated by a board-certified neuropsychologist blinded to procedures for participant allotment of this study.

cr

Interpretation for abnormalities and potential evidence of focal temporal

us

dysfunction were based on clustered deficits in language and material-specific memory domains [23]. Neuropsychological study results were determined prior to

an

interdisciplinary consultation during epilepsy case conferences.

Neuropsychological findings were classified as one of the following: unilateral

Statistical Analysis

d

or inconclusive.

M

(dominant or non-dominant) dysfunction, diffuse dysfunction, within normal limits,

Ac ce pt e

Differences in demographic and clinical characteristics were completed using

x2 or ANOVA as appropriate. For these analyses, outcome was dichotomized as “favorable” (class I or II) or “poor” (class III or IV). Spearman’s rank-order correlation was used to explore the relationship between outcome class (I, II, III, or IV) and the potential prognostic factors, including disease duration, side of resection, ictal and interictal EEG percent lateralization, ILAT score, MTS with side of resection, PET results, neuropsychological results, Wada memory asymmetry score, and Wada dysfunction concordance. Finally, logistic regression predicted surgical outcome from previously identified potential prognostic factors [5, 21, 24, 25]. The dependent variable was outcome class, dichotomized as “favorable” (class I

Page 11 of 27

or II) or “poor” (class III or IV). Predictors included ILAT score, Wada memory asymmetry score, and neuropsychological exam. Neuropsychological exam results were coded for statistical analyses as: unilateral dysfunction (code=1), diffuse

ip t

dysfunction (code=2), within normal limits (code=3), and inconclusive (code=4). Alpha level was set at p<0.05 for all comparisons. All statistical analyses were

cr

performed with Statistical Package for Social Science (SPSS) version 20 (SPSS Inc.,

us

Chicago, IL, U.S.A.). Results

an

Patient Demographics

Patient age ranged from 19 to 63 years with a mean age of 43 years. Of 31

M

patients, 16 were male and 15 were female. Notably, 29 (93.5%) were right handed. Outcome did not significantly vary by gender [p = 0.47], nor by age at surgical

d

resection [p = 0.35]. The mean post-operative time to follow-up was four years.

Ac ce pt e

Post-operative time to follow-up did not differ significantly between those with “favorable” outcome (class I or II) and “poor” outcome (class III or IV) [p = 0.62]. Wada study found the 29 right handed subjects were also left hemisphere dominant for language. Complete patient demographics can be seen in Table 1.

Page 12 of 27

Class I, II 23

All 8

31

5 3

16 15

41 (11) years 43 years 19 - 59 years

47 (9) years 43 years 38 - 63 years

cr

11 12

ip t

p = 0.47

15 8

5 3

20 11

d

20 (14) years 20 years 2 - 46 years

p = 0.35

43 (11) years 43 years 19 - 63 years

us

an 39 (11) years 34 years 27 - 56 years

M

34 (11) years 34 years 15 - 51 years

Ac ce pt e

Number of patients Gender Male Female Age Mean (SD) Median Range Side of resection Right Left Age at surgical resection Mean (SD) Median Range Epilepsy duration Mean (SD) Median Range Age of diagnosis Mean (SD) Median Range Follow up time Mean (SD) Median Range Handedness Right Left Language hemisphere Left Right Bilateral

Class III, IV

Differences by outcome group (I, II vs. III, IV)

p=0.35 35 (11) years 34 years 15 - 56 years p=0.91

21 (14) years 19 years 7 - 44 years

20 (14) years 20 years 2 - 46 years p=0.44

14 (13) years 11 years 0.5 - 42 years

18 (9) years 15 years 9 - 36 years

15 (12) years 13 years 0.5 - 42 years p = 0.62

4 (3) years 3 years 1 - 11 years

4 (3) years 4 years 1 - 10 years

4 (3) years 3 years 1 - 11 years

21 2

8 0

29 2

p=.31

21 0 2

8 0 0

29 0 2

p=.37

Page 13 of 27

Table 1: Patient demographics Note: Class I = Seizure free, Class II = > 75% reduction in seizures, Class III = 50-74% reduction in seizures, Class IV < 50% reduction in seizures; SD=Standard Deviation.

ip t

Surgical outcome

Favorable surgical outcome (class I or II) was achieved in 23 of 31 patients

cr

(74%), with a total of 14 patients (45%) achieving a seizure-free outcome (class I).

us

No surgical complications were identified. Figure 1 summarizes all 31 surgical outcomes. At follow-up, no patient in the sample exhibited neurological deficit and

an

there was no onset of amnesia, aphasia, ataxia, or apraxia. There was no new onset

Ac ce pt e

d

M

hemiparesis.

Figure 1: Surgical outcomes

Logistic regression accounted for significant variance in seizure free surgical outcome [x2(3) = 9.79, p = 0.02], with an overall correct classification rate of 85.2 (see Table 2). The standard error (SE) values were less than 2.0, indicating no

Page 14 of 27

multicolinearity in dependent variables. When applied to the sample, the logistic regression achieved a positive predictive value of 0.86 and a negative predictive value of 0.80 for an overall diagnostic odds ratio of 25.3. The model correctly

ip t

predicted favorable seizure free outcome (Class I or II) for 95.0% of subjects, but only 57.1% of subjects with poor outcome (Class III or IV). There was a prevalence

cr

rate of 74.1% for favorable outcome. Table 2 reveals that individuals with poor

us

outcome exhibit non-lateralizing neuropsychological studies, but have Wada

asymmetries in memory recognition. There was a trend for these individuals to have

B

Neuropsychological exam EEG ILAT score Wada memory asymmetry Constant -2 Log likelihood = 20.42 Model x2 = 10.48, p = .02 Nagelkerke R2 = .47 PPV = 86.4% NPV= 80.0%

1.83 -2.42 .25 -6.31

Standard Error

Ac ce pt e

d

Variable

M

an

more strongly lateralizing ictal EEG findings (p=0.09).

.91 1.43 .26 2.91

Wald

Significance

Exp(B)

4.02 2.88 .92 4.71

.05 .09 .34 .03

6.20 0.09 1.29 ---

Diagnostic Odds Ratio = 25.33 Overall Classification Rate = 85.2%

Table 2: Logistic regression analysis for predictors of surgical outcome (n=27). 1

PPV = Positive predictive value for favorable (Class I or II) surgical outcome NPV = Negative predictive value for favorable surgical outcome. NOTE: Dependent variable is seizure freedom coded so that 0 – was poor outcome and 1 – was good outcome. Four participants did not have neuropsychological data and were excluded from analyses.

2

Page 15 of 27

Ictal Lateralization Percent ictal lateralization was not found to be significantly correlated with

ip t

outcome class [p = 0.57], however a correlation trend approaching significance was noted between ILAT score and surgical outcome class [Spearman’s r(30)= 0.32, p =

cr

0.08] (data not shown). As a predictor in a multinomial logistic regression, the ILAT score demonstrated a nonsignificant (p=0.09) inverse relationship with outcome,

us

such that with each unit increase in ILAT score, individuals were more likely to have

an

poor outcome. While initially unexpected, the ILAT score can be negative when there are a high number of nonlocalizable seizures. Indeeed, six participants in this

M

study had a negative ILAT score and all of these patients had a favorable outcome (four with Class 1 outcome and two with Class II outcome). A case example may be

d

illustrative. One subject with class II outcome had an ILAT score of -.34 resulting

Ac ce pt e

from having 33% ipsilateral seizures, 17% contralateral seizures and 50% nonlocalizable seizures. Percent ictal lateralization on surface and intracranial EEG is summarized in Figure 2.

Page 16 of 27

M

an

us

cr

ip t

Percent Ictal Lateralization

Ac ce pt e

d

Figure 2: Ictal lateralization by outcome class NOTE: Graphic representation by outcome class (I, II, III, and IV) for the percent of ictal lateralization from video-EEG. The percent of lateralized EEG abnormalities recorded from surface and invasive EEG are similar for participants with Class IV outcome, but there is greater lateralization observed with invasive EEG for subjects with Class I, II, and III outcomes.

MRI and PET

MRI evidence suggested 8 cases of unilateral MTS of which 3 were

contralateral to side of resection, 12 cases of bilateral MTS, 9 cases with no evidence of MTS, and 2 cases with inconclusive findings. MRI findings were not significantly correlated with outcome class [p = 0.66]. PET scan indicated 13 cases of unilateral dysfunction of which 1 was contralateral to side of resection, 7 cases of bilateral dysfunction, and 2 cases with no evidence of dysfunction. PET results were not significantly correlated with outcome class [p = 0.50].

Page 17 of 27

Wada asymmetry and concordance Wada memory asymmetry score studies indicated 16 cases of unilateral mesial temporal dysfunction of which 1 was contralateral to side of resection, 13

ip t

cases of no asymmetry, and 2 cases of inconclusive findings. A correlation trend approaching significance was observed between outcome class and Wada memory

cr

asymmetry score [Spearman’s r(30) = 0.32, p = 0.08]. Wada dysfunction

us

concordance was significantly correlated with outcome class, with ipsilateral

dysfunction being positively associated with a better outcome [Spearman’s r(30) =

an

0.38, p = 0.04]. While not a significant independent predictor [p = 0.25], Wada memory asymmetry score exhibited iterative improvement in diagnostic accuracy of

M

the logistic regression improving correct classification of being seizure free by 1.36

Neuropsychology evaluation

d

times when asymmetric to ipsilateral side of surgery.

Ac ce pt e

Neuropsychological studies indicated 9 cases of unilateral dysfunction of

which 2 were contralateral to side of resection, 16 cases of bilateral/diffuse dysfunction, 1 case within normal limits, and 1 case of inconclusive findings. Neuropsychological study results suggestive of unilateral dysfunction were significantly correlated with a better outcome versus non-lateralized findings [Spearman’s r(26) = 0.46, p = 0.02]. Neuropsychological study results also significantly predicted outcome [p = 0.04] in the regression model, with unilateral dysfunction being predictive of a “favorable” outcome (class I or II) relative to cases with non-lateralizing findings [Exp(B) = 5.18] (Table 2). Pre-operative evaluation results are summarized in Table 3.

Page 18 of 27

Unilateral

Bilateral

Negative

Inconclusive

12 (39%)

9 (29%)

2 (6%)

13 (59%)

7 (32%)

2 (9%)

0 (0%)

16 (52%)

13 (42%)

0 (0%)

2 (6%)

9 (33%)

16 (59%)

1 (4%)

1 (4%)

us

Table 3: Pre-operative evaluation results

ip t

8 (26%)

cr

MTS1 n=31 PET2 n=22 Wada3 n=31 NP4 n=27

1

2 patients had non-relevant radiographic abnormalities 9 patients did not receive PET scan 3 Bilateral refers to “no asymmetry” 4 In 4 patients, NP study results were available only in summary form

Discussion

M

an

2

d

This study adds to the surgical outcome data for a unique patient cohort

Ac ce pt e

considered for epilepsy surgery who are diagnosed with pharmaco-resistant epilepsy with bilateral localization. First, to our knowledge this is one of the larger studies evaluating variables predictive of favorable surgical outcome for patients with pharmaco-resistant epilepsy with bilateral localization. Second, these data support previous work suggesting that patients with bilateral temporal lobe epilepsy can achieve benefit from surgical resection without disabling neurological deficits after surgery [9, 21, 24-26]. Third, these data extend the literature in documenting the difficulty to predict positive outcome (Engel class I or II) for individuals with bilateral temporal lobe seizures[6, 9-12, 25, 27, 28]. Finally, this study contributes to the conflicting literature documenting that in specific cases Wada memory asymmetry and lateralized dysfunction on neuropsychological exam

Page 19 of 27

can contribute unique data to predict likelihood of a favorable surgical intervention for patients with bilateral temporal lobe epilepsy disease [5, 6, 9-13, 25]. Surgical outcome-based studies of individuals with bilateral temporal lobe

ip t

epilepsy are difficult, due to the relatively low incidence of TLE cases exhibiting bilateral features, and because these patients are often considered to have poorer

cr

prognosis and are less frequently offered surgery as a treatment [5, 6]. After

us

applying this study’s inclusion/exclusion criteria to the literature, such that studies with unclear lateralization recordings were excluded [5, 13], there were 103

an

nonduplicated patients. Of the 103 patients, 71 (69%) were reported to have at least a class II outcome, including 51 (50%) who became seizure-free[6, 9-12, 25, 27, 28].

M

This study compares favorably, with 75 percent of patients having at least a class II outcome at an average of 4 years after temporal lobe resection that uses an inferior

d

temporal gyrus approach. These data support prior research that the presence of

Ac ce pt e

contralateral ictal EEG findings alone do not seem to be a major contraindication to a seizure-free outcome for carefully selected patients. The potential value of seizure reduction, or seizure freedom, for patients with bilateral temporal lobe epilepsy is particularly important given the higher morbidity and mortality observed among individuals with refractory epilepsy that is decreased with seizure control [29]. Further, seizure freedom is associated with improved quality of life[30, 31] and even palliative outcomes can improve subjective health status and quality of life [1, 2, 32, 33]. Neuropsychological evaluation, often used to assess functional outcomes from temporal lobe epilepsy surgery, has shown mixed value in predicting seizure-

Page 20 of 27

free outcome from surgery[5, 24, 26, 34, 35]. While not consistently observed [34], neuropsychological data can provide unique variance in predicting seizure freedom in selected patient groups in which ictal or interictal EEG abnormalities are

ip t

incongruent and/or there are no structural abnormalities identified on MRI [5, 24, 26, 35, 36]. The present study contributes to data suggesting the presence of

us

patients with TLE having bilateral disease [5, 24, 26, 35] .

cr

lateralizing neuropsychological results are predictors of favorable outcome among

As with neuropsychological evaluation, Wada testing is primarily interpreted

an

to assess the potential for functional memory deficits following surgical resection to determine the functional adequacy of the ipsilateral temporal lobe, often in

M

conjunction with neuropsychological study findings [22, 37]. The tendency of the Wada memory test scores to also predict seizure reduction in patients with

d

unilateral TLE has been described, although this remains controversial [37]. Sirven

Ac ce pt e

et al. determined that lateralized Wada test results were associated with a seizurefree outcome (class I) in patients with bilateral TLE [25]. Other studies of unilateral MTS and ipsilateral EEG have failed to demonstrate a relationship between Wada results and outcome after other non-invasive prognostic tests are employed [38]. These data found Wada memory asymmetry score using an absolute value difference to correlate significantly with surgical outcome, and show a trend to improve prediction of seizure free outcome in TLE patients with bilateral disease. The prognostic value of percent ictal lateralization has been discussed in the past studies of surgical treatment for bilateral TLE [5, 6, 9-13, 25, 27, 28]. Previous works have defined percent ictal lateralization similar to the definition used in this

Page 21 of 27

study[5, 10-12, 25]. Intracranial EEG derived percent ictal lateralization in this study ranged from 50% to 100% (see Figure 2). This compares similarly to percent ictal lateralization obtained using intracranial EEG recordings of other studies that

ip t

have ranged from 42% to 95% [6, 9-12, 25, 27, 28]. In many of these studies, patients were then carefully selected for surgical candidacy, excluding patients with

cr

non-localizing seizures, inadequate Wada results, less than 50% lateralization,

us

extra-temporal lesions on imaging, or lack of concordant pre-operative tests. These features were encountered in this study, but unique to this study, did not preclude

an

patients from surgical candidacy. Thus, these data are more generalizable to the complex patients that may be encountered at Level 4 Epilepsy Centers and extend

seizure free outcomes.

M

the literature regarding the use of common presurgical diagnostic testing to predict

d

There are several limitations that impact generalizability and clinical

Ac ce pt e

application of this study. First, these data were obtained from a retrospective chart review of a carefully selected population of epilepsy patients seeking surgical treatment in the southeastern United States. As a result, the disease parameters for pharmaco-resistant temporal lobe epilepsy with bilateral disease may not generalize to other clinics/regions. Another limitation is sample size that restricted the power of statistical test modeling. However, we do meet minimum N requirements as recommended by Agresti [39] for logistic regression and employed more conservative nonparametric Spearman rank correlation and Mann-Whitney U tests. Further, the population of interest for this study reflects a low incidence of pharmacoresistant epilepsy with bilateral localization that present for surgical

Page 22 of 27

treatment. The lack of significant prediction for the EEG data was unexpected. The ILAT score (a derived metric intended as a quantitative measurement of lateralizing findings on EEG that considers the proportion of nonlocalizable seizures) did not

ip t

provide significant variance to predict seizure free outcome. However, the ILAT algorithm is confounded by these patients, since subjects with more than 50% of

cr

nonlocalizable seizures will have a negative ILAT score, even if all other seizures

us

were ipsilateral lateralized seizures. A further confound of the ILAT score is that side of seizure surgery was determined, at least in part, by the presence of

an

lateralized EEG findings. Thus, extent of EEG lateralization is confounded with side of surgery resection, and seizure freedom, which may or may not be an index of

M

epilepsy disease. EEG lateralization as a predictor of seizure freedom after 4 years is particularly challenging in this sample, given the difficulty in predicting seizure

d

recurrence after temporal lobectomy in patients with unilateral temporal lobe

Ac ce pt e

epilepsy[40-42]. The study outcome measures were also restricted to seizure freedom and neurological status, such that post-operative neuropsychological deficits and patient self-reported quality of life was not assessed. Finally, not all patients had intracranial recordings, which adds uncertainty to seizure localization. It is recognized that invasive intracranial EEG is the gold standard for seizure localization in patients with bilateral abnormalities; however, this study likely reflects the clinical realities that some patients may refuse invasive electrode placement and/or otherwise not be possible. Despite these limitations, our findings mirror findings of other research for this unique patient population[6, 9-12, 25, 27, 28]. Replication of this study with a larger sample that includes post-operative

Page 23 of 27

neuropsychological and psychiatric function is needed. Research is also needed to identify the potential modifiable factors underlying seizure recurrence after

ip t

successful seizure surgery.

Conclusions

cr

Surgical resection can be considered in carefully selected patients presenting

us

with bilateral temporal lobe epilepsy. In patients with bilateral EEG findings, unilateral dysfunction on neuropsychological examination, and dysfunction

an

concordant with side of resection on Wada testing were associated with better outcome. Selective unilateral temporal lobe resections may be considered for

Ac ce pt e

d

contraindications for surgery.

M

individuals with bilateral temporal lobe epilepsy provided there are no other

Conflict of interest: none

Acknowledgements: none

References 1. Wiebe S, Blume WT, Girvin JP, Eliasziw M, Effectiveness, Efficiency of Surgery for Temporal Lobe Epilepsy Study G. A randomized, controlled trial of surgery for temporal-lobe epilepsy. The New England journal of medicine 2001;345:311-318.

Page 24 of 27

Ac ce pt e

d

M

an

us

cr

ip t

2. Spencer SS, Berg AT, Vickrey BG, Sperling MR, Bazil CW, Shinnar S, Langfitt JT, Walczak TS, Pacia SV, Ebrahimi N, Frobish D, Multicenter Study of Epilepsy S. Initial outcomes in the Multicenter Study of Epilepsy Surgery. Neurology 2003;61:1680-1685. 3. Wieser HG, Ortega M, Friedman A, Yonekawa Y. Long-term seizure outcomes following amygdalohippocampectomy. Journal of neurosurgery 2003;98:751-763. 4. Engel J, Jr., Rausch R, Lieb JP, Kuhl DE, Crandall PH. Correlation of criteria used for localizing epileptic foci in patients considered for surgical therapy of epilepsy. Annals of neurology 1981;9:215-224. 5. Holmes MD, Miles AN, Dodrill CB, Ojemann GA, Wilensky AJ. Identifying potential surgical candidates in patients with evidence of bitemporal epilepsy. Epilepsia 2003;44:1075-1079. 6. Hufnagel A, Elger CE, Pels H, Zentner J, Wolf HK, Schramm J, Wiestler OD. Prognostic significance of ictal and interictal epileptiform activity in temporal lobe epilepsy. Epilepsia 1994;35:1146-1153. 7. Spencer SS. Depth electrodes. Epilepsy research. Supplement 1992;5:135145. 8. Vale FL, Pollock G, Dionisio J, Benbadis SR, Tatum WO. Outcome and complications of chronically implanted subdural electrodes for the treatment of medically resistant epilepsy. Clinical neurology and neurosurgery 2013;115:985990. 9. So N, Olivier A, Andermann F, Gloor P, Quesney LF. Results of surgical treatment in patients with bitemporal epileptiform abnormalities. Annals of neurology 1989;25:432-439. 10. Boling W, Aghakhani Y, Andermann F, Sziklas V, Olivier A. Surgical treatment of independent bitemporal lobe epilepsy defined by invasive recordings. Journal of neurology, neurosurgery, and psychiatry 2009;80:533-538. 11. Hirsch LJ, Spencer SS, Spencer DD, Williamson PD, Mattson RH. Temporal lobectomy in patients with bitemporal epilepsy defined by depth electroencephalography. Annals of neurology 1991;30:347-356. 12. Holmes MD, Dodrill CB, Ojemann GA, Wilensky AJ, Ojemann LM. Outcome following surgery in patients with bitemporal interictal epileptiform patterns. Neurology 1997;48:1037-1040. 13. Olivier A, Gloor P, Ives J. [Research and surgical treatment of bitemporal epilepsy]. L'union medicale du Canada 1980;109:247-250. 14. Gil-Nagel A, Risinger MW. Ictal semiology in hippocampal versus extrahippocampal temporal lobe epilepsy. Brain : a journal of neurology 1997;120 ( Pt 1):183-192. 15. Luders H, Acharya J, Baumgartner C, Benbadis S, Bleasel A, Burgess R, Dinner DS, Ebner A, Foldvary N, Geller E, Hamer H, Holthausen H, Kotagal P, Morris H, Meencke HJ, Noachtar S, Rosenow F, Sakamoto A, Steinhoff BJ, Tuxhorn I, Wyllie E. Semiological seizure classification. Epilepsia 1998;39:1006-1013. 16. Maillard L, Vignal JP, Gavaret M, Guye M, Biraben A, McGonigal A, Chauvel P, Bartolomei F. Semiologic and electrophysiologic correlations in temporal lobe seizure subtypes. Epilepsia 2004;45:1590-1599.

Page 25 of 27

Ac ce pt e

d

M

an

us

cr

ip t

17. Vale FL, Effio E, Arredondo N, Bozorg A, Wong K, Martinez C, Downes K, Tatum WO, Benbadis SR. Efficacy of temporal lobe surgery for epilepsy in patients with negative MRI for mesial temporal lobe sclerosis. Journal of clinical neuroscience : official journal of the Neurosurgical Society of Australasia 2012;19:101-106. 18. Duckworth EA, Vale FL. Trephine epilepsy surgery: the inferior temporal gyrus approach. Neurosurgery 2008;63:ONS156-160; discussion ONS160-151. 19. Hirsch LJ, Spencer SS, Williamson PD, Spencer DD, Mattson RH. Comparison of bitemporal and unitemporal epilepsy defined by depth electroencephalography. Annals of neurology 1991;30:340-346. 20. Sanchez-Alvarez JC, Serrano-Castro P, Canadillas-Hidalgo F. [Refractory epilepsy in adults]. Revista de neurologia 2002;35:931-953. 21. Sabsevitz DS, Swanson SJ, Morris GL, Mueller WM, Seidenberg M. Memory outcome after left anterior temporal lobectomy in patients with expected and reversed Wada memory asymmetry scores. Epilepsia 2001;42:1408-1415. 22. Loring DW, Lee GP, Meador KJ, Flanigin HF, Smith JR, Figueroa RE, Martin RC. The intracarotid amobarbital procedure as a predictor of memory failure following unilateral temporal lobectomy. Neurology 1990;40:605-610. 23. American Academy of Clinical N. American Academy of Clinical Neuropsychology (AACN) practice guidelines for neuropsychological assessment and consultation. The Clinical neuropsychologist 2007;21:209-231. 24. Sawrie SM, Martin RC, Gilliam FG, Roth DL, Faught E, Kuzniecky R. Contribution of neuropsychological data to the prediction of temporal lobe epilepsy surgery outcome. Epilepsia 1998;39:319-325. 25. Sirven JI, Malamut BL, Liporace JD, O'Connor MJ, Sperling MR. Outcome after temporal lobectomy in bilateral temporal lobe epilepsy. Annals of neurology 1997;42:873-878. 26. Seidenberg M, Hermann B, Wyler AR, Davies K, Dohan FC, Jr., Leveroni C. Neuropsychological outcome following anterior temporal lobectomy in patients with and without the syndrome of mesial temporal lobe epilepsy. Neuropsychology 1998;12:303-316. 27. Cukiert A, Cukiert CM, Argentoni M, Baise-Zung C, Forster CR, Mello VA, Burattini JA, Mariani PP. Outcome after cortico-amygdalo-hippocampectomy in patients with severe bilateral mesial temporal sclerosis submitted to invasive recording. Seizure : the journal of the British Epilepsy Association 2009;18:515-518. 28. Cukiert A, Sousa A, Machado E, Buratini JA, Forster C, Argentoni M. Results of surgery in patients with bilateral independent temporal lobe spiking (BITLS) with normal MRI or bilateral mesial temporal sclerosis (MTS) investigated with bilateral subdural grids. Arquivos de neuro-psiquiatria 2000;58:1009-1013. 29. Cockerell OC, Johnson AL, Sander JW, Hart YM, Goodridge DM, Shorvon SD. Mortality from epilepsy: results from a prospective population-based study. Lancet 1994;344:918-921. 30. Gilliam F. Optimizing health outcomes in active epilepsy. Neurology 2002;58:S9-20.

Page 26 of 27

Ac ce pt e

d

M

an

us

cr

ip t

31. Hermann BP, Seidenberg M, Bell B, Woodard A, Rutecki P, Sheth R. Comorbid psychiatric symptoms in temporal lobe epilepsy: association with chronicity of epilepsy and impact on quality of life. Epilepsy & behavior : E&B 2000;1:184-190. 32. Birbeck GL, Hays RD, Cui X, Vickrey BG. Seizure reduction and quality of life improvements in people with epilepsy. Epilepsia 2002;43:535-538. 33. Gilliam F. The impact of epilepsy on subjective health status. Current neurology and neuroscience reports 2003;3:357-362. 34. Bell ML, Rao S, So EL, Trenerry M, Kazemi N, Stead SM, Cascino G, Marsh R, Meyer FB, Watson RE, Giannini C, Worrell GA. Epilepsy surgery outcomes in temporal lobe epilepsy with a normal MRI. Epilepsia 2009;50:2053-2060. 35. Potter JL, Schefft BK, Beebe DW, Howe SR, Yeh HS, Privitera MD. Presurgical neuropsychological testing predicts cognitive and seizure outcomes after anterior temporal lobectomy. Epilepsy & behavior : E&B 2009;16:246-253. 36. Helmstaedter C, Kockelmann E. Cognitive outcomes in patients with chronic temporal lobe epilepsy. Epilepsia 2006;47 Suppl 2:96-98. 37. Rathore C, Kesavadas C, Sarma SP, Radhakrishnan K. Usefulness of Wada test in predicting seizure outcome following anterior temporal lobectomy. Epilepsy research 2013;107:279-285. 38. Kirsch HE, Walker JA, Winstanley FS, Hendrickson R, Wong ST, Barbaro NM, Laxer KD, Garcia PA. Limitations of Wada memory asymmetry as a predictor of outcomes after temporal lobectomy. Neurology 2005;65:676-680. 39. Agresti A. An introduction to categorical data analysis: NJ: Wiley 2007:138 pp. 40. McIntosh AM, Kalnins RM, Mitchell LA, Fabinyi GC, Briellmann RS, Berkovic SF. Temporal lobectomy: long-term seizure outcome, late recurrence and risks for seizure recurrence. Brain : a journal of neurology 2004;127:2018-2030. 41. Bragin A, Wilson CL, Engel J, Jr. Chronic epileptogenesis requires development of a network of pathologically interconnected neuron clusters: a hypothesis. Epilepsia 2000;41 Suppl 6:S144-152. 42. Jiruska P, de Curtis M, Jefferys JG, Schevon CA, Schiff SJ, Schindler K. Synchronization and desynchronization in epilepsy: controversies and hypotheses. The Journal of physiology 2013;591:787-797.

Page 27 of 27