Clinical Study Stereotact Funct Neurosurg 2014;92:354–359 DOI: 10.1159/000365575
Received: November 13, 2013 Accepted after revision: June 29, 2014 Published online: October 28, 2014
Surgical Treatment of MRI-Negative Temporal Lobe Epilepsy Based on PET: A Retrospective Cohort Study Rui Feng a Jie Hu a, b Li Pan a Jiali Shi d Chun Qiu c Liqin Lang a Xin Gu b Jun Guo b
Department of Neurosurgery, Huashan Hospital, b Department of Neurosurgery, Jing’an Branch of Huashan Hospital and c PET Center of Huashan Hospital, Fudan University, Shanghai, and d Mathematics and Software Science College, Sichuan Normal University, Chengdu, PR China
Key Words Temporal lobe epilepsy · Hippocampal sclerosis · Magnetic resonance imaging-negative epilepsy · Positron emission tomography · Standard anterior temporal lobectomy
Abstract Introduction: Using retrospective and comparative methods, we aim to discuss the surgical treatment of magnetic resonance imaging (MRI)-negative temporal lobe epilepsy (TLE) presented with positive positron emission tomography (PET) results. Methods: From the viewpoint of semiology, demography, surgical treatment and prognosis evaluation, we compared 19 MRI-negative, PET-positive TLE patients to 41 TLE with hippocampal sclerosis patients, and then statistically analyzed the differences between these 2 cohorts. Results: Under intraoperative electrocorticography monitoring, all patients underwent successful standard anterior temporal lobectomy. It appears that there is no significant difference between the surgical outcome of MRI-negative/ PET-positive TLE (Engle class I: 68.4%, Engle class I + II: 84.2%)
© 2014 S. Karger AG, Basel 1011–6125/14/0926–0354$39.50/0 E-Mail [email protected] www.karger.com/sfn
and TLE with hippocampal sclerosis (Engle class I: 68.3%, Engle class I + II: 80.5%). The analysis also shows that to some extent MRI-negative, PET-positive TLE might be distinct from TLE with hippocampal sclerosis as a clinical entity, i.e. the former is not a subtype of the latter. History of febrile convulsion and occurrence of secondary generalized tonic-clonic seizure may possibly differentiate them from each other. Conclusion: Successful resective surgery of MRI-negative TLE based on PET can yield similar favorable results to TLE with hippocampal sclerosis. This study demonstrates that with reasonable presurgical workup, such TLE subtypes can be surgically treated without invasive intracranial electrode implantation. © 2014 S. Karger AG, Basel
Introduction
Temporal lobe epilepsy (TLE), which presents with excellent surgical prognosis, is the most common type of intractable epilepsy. TLE with hippocampal sclerosis (HS + Jie Hu or Li Pan Department of Neurosurgery, Huashan Hospital Shanghai Medical College, Fudan University 12 Wulumuqi Zhong Road, Shanghai 200040 (PR China) E-Mail jiehu68 @ yahoo.com or lipanmr @ sina.com
Downloaded by: Freie Universität Berlin 130.133.8.114 - 5/2/2017 2:52:46 PM
a
Methods This study has been carried out in accordance with the Ethical Principles for Medical Research Involving Human Subjects (Declaration of Helsinki). All clinical data were retrospectively collected from the database of our department, ranging from January 2002 to January 2013. All patients were diagnosed as TLE and underwent surgical treatments in our department. According to MRI features, we divided the patients into 2 cohorts: HS + TLE and MRI– TLE. Patient Inclusion and Exclusion Criteria (1) All patients were treated by multiple antiepilepsy drugs for at least 1 year, which yielded poor control of epilepsy. Patients who were medically treated for less than 1 year before surgery were excluded. (2) Diagnoses of TLE were basically confirmed by EEG and semiology. We excluded patients whose interictal or ictal scalp EEG
MRI-Negative PET-Positive Temporal Lobe Epilepsy
Fig. 1. Thin-sliced MRI showed no meaningful abnormality which
could possibly be epileptogenic focus.
showed epileptiform discharges that were confined contralaterally to localization by MRI or PET studies. (3) In MRI– TLE cases, all PET studies demonstrated striking unilateral temporal hypometabolism. Cases with slight or bilateral hypometabolism in temporal lobes were excluded. Patients who had obvious 18FDG hypometabolism in regions other than the anterior temporal lobe, hippocampus, thalamus or cerebellum were excluded. (4) Cases with multiple cerebral lesions were excluded. (5) Diagnosis of HS + TLE were confirmed by at least 2 radiologists of our hospital. Cases which were contentious in diagnosis were excluded. General Data and Semiology Evaluation We statistically compared general data and semiology of the 2 cohorts: (1) General data included age, seizure duration, history of febrile convulsion, sex ratio and lateralization. (2) Semiology [7, 8] evaluation included type of auras, ictal presentation and postictal patient status. (3) Student’s test was used to compare age and seizure duration of the 2 cohorts. The χ2 test was used to compare differences between history of febrile convulsion, sex ratio, lateralization, aura occurrence rate, ratio of Engel class I and ratio of Engel class I + II. Differences between aura types and seizure types of the 2 cohorts were also compared with the χ2 test method. Presurgical Workup All patients received more than 2 kinds of noninvasive epileptogenic focus localization tools as presurgical workup, including PET, long-term monitoring EEG and 1.5- and 3-tesla MRI (fig. 1).
Stereotact Funct Neurosurg 2014;92:354–359 DOI: 10.1159/000365575
355
Downloaded by: Freie Universität Berlin 130.133.8.114 - 5/2/2017 2:52:46 PM
TLE), which constitutes at least 60–70% of TLE [1, 2], has the best surgical indication. On magnetic resonance imaging (MRI), HS + TLE usually displays temporal lobe atrophy and signal increase on fluid-attenuated inversion recovery (FLAIR) imaging. Positron emission tomography (PET) usually indicates involvement of the temporal lobe with hypometabolism. After successful standard anterior temporal lobectomies (sATLs), most patients can achieve excellent outcome. However, epilepsy surgeons frequently run into TLE without meaningful abnormality on MRI, which is the so-called MRI-negative TLE (MRI– TLE), and this accounts for 20–30% of all TLE [3, 4]. Multiple onecenter retrospective studies reported that the rate of favorable surgical prognosis (Engel grade I + II) of MRI– TLE is approximately 45–90% [2, 5]. Among these MRI– TLE, many displayed unilateral temporal hypometabolism in PET studies (PET-positive/MRI– TLE). According to one meta-analysis involving 46 independent studies, the predictive value of PET to good surgical outcome (Engel I + II) of MRI– TLE is 80% [6]. Although preoperative invasive electroencephalogram (EEG) is still applied by many centers for such types of TLE, more and more epilepsy centers operate on these patients directly without invasive intracranial electrode implantation [2, 4, 5]. This study retrospectively analyzes 60 cases of unilateral TLE which were surgically treated in our department, including 41 cases of HS + TLE and 19 cases of MRI– TLE. PET studies revealed hypometabolism of the unilateral temporal lobe in all cases of MRI– TLE. Retrospectively, we compare MRI– TLE with HS + TLE, which is well known for the most favorable surgical prognosis, summarizing the clinical characteristics, diagnosis, surgical treatment and prognosis of this entity.
PET Studies Altogether, 49 patients underwent interictal FDG-PET studies, among whom there were 30 cases of HS + TLE and 19 cases of MRI– TLE. All studies were carried out in the PET center of our hospital. PET hypometabolism was determined by unilateral decreased 18FDG uptake in the temporal lobe by visual inspection of more than 2 experienced physicians who had no prior knowledge of the clinical data of these patients (fig. 2). Long-Term Monitoring EEG Long-term monitoring scalp EEG was monitored by a 16- or 32-channel EEG system. Off-line EEG analysis was conducted and reported by 3 experienced EEG experts. Ictal descriptions were written by EEG technicians and epileptologists.
Resective Surgeries and Histopathology Studies Preresection and postresection electrocorticography (EcoG) were performed during each surgery. Operation technique: we performed tailored sATL which included unilateral anterior temporal lobe resection and amygdalohippocampectomy. All histopathology studies were performed in the department of neuropathology in our hospital. Follow-up Follow-ups were conducted by telephone or when patients made return visits to our hospital for consultation. Surgical outcomes were evaluated as follows according to the Engel grading system [9]: Engel class I: seizure free or free of disabling seizures; class II, rare seizures per year (less than 3 seizure days); class III, effective (seizure decreased by at least 80%), and class IV, no improvement. Seizures within the first month after operation were excluded. Surgical effectiveness defined by the Engel class of the 2 cohorts was statistically analyzed using the χ2 test method.
Results
General data and semiology of HS + TLE and MRI– TLE are summarized in table 1. We compared statistical differences between the 2 cohorts. General Data There was no significant difference between the 2 cohorts in relation to average age at operation, sex ratio, lateralization, and seizure duration (p > 0.05). However, history of febrile convulsion differed significantly between the 2 cohorts (HS + TLE: 22.0%, MRI– TLE: 10.5%, p < 0.05). 356
Stereotact Funct Neurosurg 2014;92:354–359 DOI: 10.1159/000365575
Fig. 2. Interictal FDG-PET showed striking left temporal hypometabolism, indicating the epileptogenic zone (red and yellow indicate the highest while blue and purple indicate the lowest hypometabolism).
Table 1. General data and semiology of the two cohorts
Cases, n Sex ratio (male:female) Lateralization (left:right) Febrile convulsion, % Age at operation, years Period of disease, months Aura, % Emotional Experiencing Cephalic sensation (dizziness) Visceral-sensory or autonomic Hallucination None Ictal presentation, % Automatisms Language alterations Absence Secondary GTCS Removing clothes Ictal laughter Spitting Unilateral dystonic posturing Head version Eye deviation
HS + TLE
MRI– TLE
41 25:16 20:21 22 22.8 109.9
19 11:8 10:9 10.5 24.4 137.4
12.1 17.1 14.6 31.7 7.3 22.0
10.5 10.5 10.5 31.6 5.3 36.8
36.6 4.9 26.8 41.5 2.4 0 0 24.4 4.9 4.9
47.4 15.8 36.8 73.7 0 0 0 26.3 10.5 0
A patient may complain of more than 1 type of aura or ictal presentation.
Feng/Hu/Pan/Shi/Qiu/Lang/Gu/Guo
Downloaded by: Freie Universität Berlin 130.133.8.114 - 5/2/2017 2:52:46 PM
MRI Studies All patients in the 2 cohorts were studied with hippocampus scanning, which included sagittal T1-weighted image (T1WI) localization (5 mm), axial T2WI scan (7 mm) and tilted coronal T1WI, T2WI and FLAIR scans (3 mm). MRI scanners were either 1.5 or 3.0 T.
Presurgical Workup PET Studies Among all 49 patients who were the subjects of 18FDGPET studies, 29 HS + TLE and 19 MRI– TLE exhibited unilateral temporal lobe hypometabolism (HS + TLE: 96.7%, MRI– TLE: 100%). Long-Term Monitoring EEG Long-term monitoring EEG revealed interictal or ictal epileptic discharges. We excluded cases which displayed no abnormality on EEG or cases in which epileptic discharges were confined to the contralateral side of the operation, i.e. cases with ipsilateral (concordant with PET or MRI) or bilateral discharges were both included. The relationship between EEG patterns and the prognosis of MRI– TLE is summarized in table 2.
Table 2. EEG patterns and prognosis of the MRI– TLE cohort
EEG pattern (interictal and ictal) Ipsilateral Bilateral
Cases
Engel I Engel I + II
14 5
9/14 4/5
12/14 5/5
Resective Surgeries All 60 resective surgeries were performed by 5 experienced neurosurgeons. All of these operations were sATLs under EcoG monitoring. The extent of temporal neocortex resection was 3–4 cm from the temporal pole on the left side and 4–5 cm on the right side. The resection range of the hippocampus was 3 cm from the hippocampal head; 1 patient suffered from postoperative complication of acute epidural hematoma, but survived with no neurological deficit. None of the patients underwent postoperative disabling neurological deficit or surgery-related death. Histopathology For HS + TLE and MRI– TLE cases, histopathology studies can demonstrate hypoxic, ischemic and dysplastic change of neocortical and hippocampal neurons; sometimes laminar disorganization of layering as well as gliocyte proliferation are present. Follow-up These patients were followed up postoperatively from 0.6 to 11.0 years – average 4.0 years (HS + TLE: 4.7 years, MRI– TLE: 4.4 years). To evaluate the surgical prognosis of the 2 cohorts, we statistically compared the Engel class I ratio as well as the Engel class I + II ratio. There seemed to be no significant difference between the 2 cohorts when comparing the Engel class I ratio (p > 0.05; table 3). When comparing the Engel class I + II ratio, we found that there was no significant difference between the 2 cohorts (HS + TLE: 80.5%, MRI– TLE: 84.2%, p > 0.05; table 3).
MRI Studies All 60 patients took a 1.5- or 3.0-tesla MRI examination, to which we added thin-sliced coronal scans for mesial TLE in order to detect hippocampal lesions. We divided our cases into 2 cohorts according to MRI presentation, as follows: (1) HS + TLE (41 cases): MRI showed unilateral atrophy of mesial temporal structures and hyperintensity in FLAIR sequence with/without unilateral enlargement of the temporal horn; we confined this cohort to real HS in order to prevent interference of non-HS cases to surgical prognosis evaluation and (2) MRI– TLE (19 cases): MRI revealed no meaningful structural or signal anomaly in temporal lobes or other cortex.
MRI showed high specificity (83%) and sensitivity (97%) for the diagnosis of HS [10]. However, even if the resolution of MRI keeps elevating, there are still many TLE cases in which no meaningful abnormality can be
MRI-Negative PET-Positive Temporal Lobe Epilepsy
Stereotact Funct Neurosurg 2014;92:354–359 DOI: 10.1159/000365575
Discussion
357
Downloaded by: Freie Universität Berlin 130.133.8.114 - 5/2/2017 2:52:46 PM
Semiology We divided aura types into experiencing auras (déjà vu, forced thoughts, undefinable feeling), visceral-sensory or autonomic auras (abdominal sensation or epigastric rising sensation, vomiting, alterations in cardiac frequency and rhythm, cyanosis, pallor, piloerection), emotional auras (fear and anxiety), hallucination, cephalic sensation (dizziness), and no auras. All the aura types between the 2 cohorts were not significantly different (p > 0.05). We divided ictal presentation types into oral alimentary and manual automatisms, secondary generalized tonic-clonic seizure (GTCS), language alterations, unilateral dystonic posturing, absence, ictal laughter, removing clothes, eye deviation, head version, and ictal spitting. Most of them were not significantly different, except for the occurrence of secondary GTCS which differed between the 2 cohorts (HS + TLE: 41.5%, MRI– TLE: 73.7%, p < 0.05).
Cases Average postoperative period, years Engel grading, % I II I + II III IV
HS + TLE
MRI– TLE
41 4.43
19 4.10
68.3 12.2 80.5 17.1 2.43
68.4 15.8 84.2 10.5 5.3
observed. Such cases are classified as MRI– TLE, accounting for approximately 20–30% of TLE [3, 4, 10, 11]. Recently, many MRI– TLE are surgically treated and the prognoses are satisfactory. Some studies show that surgical specimens (specifically temporal neocortex and hippocampus) from MRI– TLE patients are similar to HS + TLE histopathologically [3]. Therefore, there is an existing viewpoint which considers MRI– TLE as HS without mesial temporal volume atrophy and signal change, which stands for a distinct subtype of HS. But there are disagreements: many neurologists attempted to prove that MRI– TLE was distinct from HS + TLE in aspects of clinical presentation, semiology, EEG features, histopathology, and surgical prognosis. Chassoux et al. [12] and Abou-Hamden et al. [13] found that histopathologically MRI– TLE resulted from microencephalocele or focal cortical dysplasia in the temporal lobe. Carne et al. [2] demonstrated that the extent of hypometabolism of MRI– TLE is broader than HS + TLE, suggesting that they are distinct from each other. Data from our study show that history of febrile convulsion and occurrence of secondary GTCS differ significantly between MRI– TLE and HS + TLE, supporting the point that to some extent they might be distinct from each other. Among many noninvasive presurgical workup tools for MRI– TLE, PET has unique a lateralizing value and is associated with good surgical prognosis [6, 14, 15]. Although 18FDG is limitedly harmful to the human body, we suggest, based on evidence from multiple studies (including our study) [2, 4, 12], that all MRI– TLE patients should undergo PET examination to see whether they are potential surgical candidates. The reason for hypometabolism in the temporal lobe of most TLE patients may not be the HS itself, but rather the long-term epileptic discharges which cause functional changes and disconnections in the neurons. 358
Stereotact Funct Neurosurg 2014;92:354–359 DOI: 10.1159/000365575
It is well known that the range of PET hypometabolism, which stands for the functional deficit zone, tends to be larger than the real epileptogenic focus, thus the lateralization value of PET is more important than the localization value, and resective surgeries should be tailored under EcoG monitoring. With regard to surgical treatment, one consensus is that the surgical prognosis of TLE is intimately associated with the resection extent of the temporal lobe. The study of Ojemann [16] showed that tailored resection of the neocortical temporal lobe under EcoG monitoring can result in excellent surgical prognosis while performing limited amygdalohippocampectomy, while on the contrary, Wyler et al. [17] and Nayel et al. [18] believed that the resective extent of the mesial instead of the neocortical temporal lobe is associated with surgical prognosis. Our study did not focus on such quantitative studies. However, we recommend the application of subdural EcoG monitoring during sATL surgeries. The resection extent of the neocortical temporal lobe is usually 3–4 cm on the left side and 4–5 cm on the right side from the temporal pole. Intraoperative EcoG are deemed as one of multiple indicators for the resection extent. For the hippocampus, the resection extent would be at least 3 cm from the hippocampal head. Many epilepsy centers considered that the surgical outcome of MRI-negative epilepsy is worse than lesional epilepsy [4, 19]. Thus, a huge number of such patients have to undergo intracranial electrode implantation to localize the epileptogenic focus, or even do not have the chance to receive surgery. In our clinical work, for MRInegative/PET-positive TLE, we did not take the invasive electrode implantation as a routine presurgical workup procedure, but its surgical prognosis was as good as HS + TLE, which is considered as one of the best surgical candidates. Moreover, previous literature supported the hypothesis that for such intractable epilepsy all nonconcordant factors should be excluded in order to perform ‘safe’ surgeries, which include the concordant side of EEG abnormal discharges with PET localization [2, 4]. Similarly, in our 2 cohorts, we excluded patients who had scalp EEG findings that were confined contralaterally to lateralization of PET or MRI studies. We support that PET-positive, MRI– TLE can be surgically treated by sATL, regardless of whether EEG findings are confined to the ipsilateral or bilateral sides (table 2). Although the comparison may not be statistically significant because of the low sample size, one can tell that the prognosis of the ‘bilateral’ group does not have to be poorer than the ‘ipsilateral’ group. Furthermore, for the current status of epiFeng/Hu/Pan/Shi/Qiu/Lang/Gu/Guo
Downloaded by: Freie Universität Berlin 130.133.8.114 - 5/2/2017 2:52:46 PM
Table 3. Comparison of Engel grades between the two cohorts
lepsy surgery, if the PET study of MRI– TLE is negative or ambiguous, surgical treatment should not be performed until successful intracranial electrode implantation is performed to localize the epileptogenic focus. Standardized semiology and EEG evaluation should be repeated to make sure such cases are real TLE. There may be different views on whether MRI can reveal minor abnormal structures or signals if its resolution is elevated high enough. What we emphasize is that even
if MRI cannot provide sufficient evidence for lateralization or localization, unilateral temporal hypometabolism in a PET study should be one of the important indicators of sATL surgery in TLE patients.
Disclosure Statement The authors have no conflicts of interest to declare.
References
MRI-Negative PET-Positive Temporal Lobe Epilepsy
7 Pascual MRQ: Temporal lobe epilepsy: clinical semiology and neurophysiological studies. Semin Ultrasound CT MR 2007;28:416–423. 8 Marks WJ, Laxer KD: Semiology of temporal lobe seizures: value in lateralizing the seizure focus. Epilepsia 1998;39:721–726. 9 Wieser HG, Blume WT, Fish G, Goldensohn E, Hufnagel A, King D, Sperling MR, Luders H; Commission on Neurosurgery of the International League against Epilepsy: Proposal for a new classification of outcome with respect to epileptic seizures following epilepsy surgery. Epilepsia 2001;42:282–286. 10 Berkovic SF, McIntosh AM, Kalnins RM, Jackson GD, Fabinyi GCA, Brazenor GA, Bladin PF, Hopper JL: Preoperative MRI predicts outcome of temporal lobectomy – an actuarial analysis. Neurology 1995;45:1358–1363. 11 Wiebe S, Blume WT, Girvin JP, Eliasziw M; Effectiveness and Efficiency of Surgery for Temporal Lobe Epilepsy Study Group: A randomized, controlled trial of surgery for temporal-lobe epilepsy. New Engl J Med 2001; 345:311–318. 12 Chassoux F, Rodrigo S, Semah F, Beuvon F, Landre E, Devaux B, Turak B, Mellerio C, Meder JF, Roux FX, Daumas-Duport C, Merlet P, Dulac O, Chiron C: FDG-PET improves surgical outcome in negative MRI Taylortype focal cortical dysplasias. Neurology 2010; 75:2168–2175.
13 Abou-Hamden A, Lau M, Fabinyi G, Berkovic SF, Jackson GD, Mitchell LA, Kalnins R, Fitt G, Archer JS: Small temporal pole encephaloceles: a treatable cause of ‘lesion negative’ temporal lobe epilepsy. Epilepsia 2010; 51:2199–2202. 14 Radtke RA, Hanson MW, Hoffman JM, Crain BJ, Walczak TS, Lewis DV, Beam C, Coleman RE, Friedman AH: Temporal-lobe hypometabolism on PET – predictor of seizure control after temporal lobectomy. Neurology 1993;43:1088–1092. 15 Dupont S, Semah F, Clemenceau S, Adam C, Baulac M, Samson Y: Accurate prediction of postoperative outcome in mesial temporal lobe epilepsy – a study using positron emission tomography with 18fluorodeoxyglucose. Arch Neurol 2000;57:1331–1336. 16 Ojemann GA: Surgical therapy for medically intractable epilepsy. J Neurosurg 1987; 66: 489–499. 17 Wyler AR, Hermann BP, Somes G: Extent of medial temporal resection on outcome from anterior temporal lobectomy – a randomized prospective study. Neurosurgery 1995; 37: 982–990. 18 Nayel MH, Awad IA, Luders H: Extent of mesiobasal resection determines outcome after temporal lobectomy for intractable complex partial seizures. Neurosurgery 1991;29:55–61. 19 Tellez-Zenteno JF, Ronquillo LH, Moien-Afshari F, Wiebe S: Surgical outcomes in lesional and non-lesional epilepsy: a systematic review and meta-analysis. Epilepsy Res 2010;89: 310–318.
Stereotact Funct Neurosurg 2014;92:354–359 DOI: 10.1159/000365575
359
Downloaded by: Freie Universität Berlin 130.133.8.114 - 5/2/2017 2:52:46 PM
1 Cascino GD, Jack CR, Parisi JE, Sharbrough FW, Hirschorn KA, Meyer FB, Marsh WR, Obrien PC: Magnetic resonance imagingbased volume studies in temporal lobe epilepsy – pathological correlations. Ann Neurol 1991;30:31–36. 2 Carne RP, O’Brien TJ, Kilpatrick CJ, MacGregor LR, Hicks RJ, Murphy MA, Bowden SC, Kaye AH, Cook MJ: MRI-negative PETpositive temporal lobe epilepsy: a distinct surgically remediable syndrome. Brain 2004;127: 2276–2285. 3 Bien CG, Szinay M, Wagner J, Clusmann H, Becker AJ, Urbach H: Characteristics and surgical outcomes of patients with refractory magnetic resonance imaging-negative epilepsies. Arch Neurol 2009;66:1491–1499. 4 LoPinto-Khoury C, Sperling MR, Skidmore C, Nei M, Evans J, Sharan A, Mintzer S: Surgical outcome in PET-positive, MRI-negative patients with temporal lobe epilepsy. Epilepsia 2012;53:342–348. 5 Immonen A, Jutila L, Muraja-Murro A, Mervaala E, Aikia M, Lamusuo S, Kuikka J, Vanninen E, Alafuzoff I, Ikonen A, Vanninen R, Vapalahti M, Kalviainen R: Long-term epilepsy surgery outcomes in patients with MRInegative temporal lobe epilepsy. Epilepsia 2010;51:2260–2269. 6 Willmann O, Wennberg R, May T, Woermann FG, Pohlmann-Eden B: The contribution of 18F-FDG PET in preoperative epilepsy surgery evaluation for patients with temporal lobe epilepsy a meta-analysis. Seizure 2007; 16:509–520.
Received: November 13, 2013 Accepted after revision: June 29, 2014 Published online: October 28, 2014
Surgical Treatment of MRI-Negative Temporal Lobe Epilepsy Based on PET: A Retrospective Cohort Study Rui Feng a Jie Hu a, b Li Pan a Jiali Shi d Chun Qiu c Liqin Lang a Xin Gu b Jun Guo b
Department of Neurosurgery, Huashan Hospital, b Department of Neurosurgery, Jing’an Branch of Huashan Hospital and c PET Center of Huashan Hospital, Fudan University, Shanghai, and d Mathematics and Software Science College, Sichuan Normal University, Chengdu, PR China
Key Words Temporal lobe epilepsy · Hippocampal sclerosis · Magnetic resonance imaging-negative epilepsy · Positron emission tomography · Standard anterior temporal lobectomy
Abstract Introduction: Using retrospective and comparative methods, we aim to discuss the surgical treatment of magnetic resonance imaging (MRI)-negative temporal lobe epilepsy (TLE) presented with positive positron emission tomography (PET) results. Methods: From the viewpoint of semiology, demography, surgical treatment and prognosis evaluation, we compared 19 MRI-negative, PET-positive TLE patients to 41 TLE with hippocampal sclerosis patients, and then statistically analyzed the differences between these 2 cohorts. Results: Under intraoperative electrocorticography monitoring, all patients underwent successful standard anterior temporal lobectomy. It appears that there is no significant difference between the surgical outcome of MRI-negative/ PET-positive TLE (Engle class I: 68.4%, Engle class I + II: 84.2%)
© 2014 S. Karger AG, Basel 1011–6125/14/0926–0354$39.50/0 E-Mail [email protected] www.karger.com/sfn
and TLE with hippocampal sclerosis (Engle class I: 68.3%, Engle class I + II: 80.5%). The analysis also shows that to some extent MRI-negative, PET-positive TLE might be distinct from TLE with hippocampal sclerosis as a clinical entity, i.e. the former is not a subtype of the latter. History of febrile convulsion and occurrence of secondary generalized tonic-clonic seizure may possibly differentiate them from each other. Conclusion: Successful resective surgery of MRI-negative TLE based on PET can yield similar favorable results to TLE with hippocampal sclerosis. This study demonstrates that with reasonable presurgical workup, such TLE subtypes can be surgically treated without invasive intracranial electrode implantation. © 2014 S. Karger AG, Basel
Introduction
Temporal lobe epilepsy (TLE), which presents with excellent surgical prognosis, is the most common type of intractable epilepsy. TLE with hippocampal sclerosis (HS + Jie Hu or Li Pan Department of Neurosurgery, Huashan Hospital Shanghai Medical College, Fudan University 12 Wulumuqi Zhong Road, Shanghai 200040 (PR China) E-Mail jiehu68 @ yahoo.com or lipanmr @ sina.com
Downloaded by: Freie Universität Berlin 130.133.8.114 - 5/2/2017 2:52:46 PM
a
Methods This study has been carried out in accordance with the Ethical Principles for Medical Research Involving Human Subjects (Declaration of Helsinki). All clinical data were retrospectively collected from the database of our department, ranging from January 2002 to January 2013. All patients were diagnosed as TLE and underwent surgical treatments in our department. According to MRI features, we divided the patients into 2 cohorts: HS + TLE and MRI– TLE. Patient Inclusion and Exclusion Criteria (1) All patients were treated by multiple antiepilepsy drugs for at least 1 year, which yielded poor control of epilepsy. Patients who were medically treated for less than 1 year before surgery were excluded. (2) Diagnoses of TLE were basically confirmed by EEG and semiology. We excluded patients whose interictal or ictal scalp EEG
MRI-Negative PET-Positive Temporal Lobe Epilepsy
Fig. 1. Thin-sliced MRI showed no meaningful abnormality which
could possibly be epileptogenic focus.
showed epileptiform discharges that were confined contralaterally to localization by MRI or PET studies. (3) In MRI– TLE cases, all PET studies demonstrated striking unilateral temporal hypometabolism. Cases with slight or bilateral hypometabolism in temporal lobes were excluded. Patients who had obvious 18FDG hypometabolism in regions other than the anterior temporal lobe, hippocampus, thalamus or cerebellum were excluded. (4) Cases with multiple cerebral lesions were excluded. (5) Diagnosis of HS + TLE were confirmed by at least 2 radiologists of our hospital. Cases which were contentious in diagnosis were excluded. General Data and Semiology Evaluation We statistically compared general data and semiology of the 2 cohorts: (1) General data included age, seizure duration, history of febrile convulsion, sex ratio and lateralization. (2) Semiology [7, 8] evaluation included type of auras, ictal presentation and postictal patient status. (3) Student’s test was used to compare age and seizure duration of the 2 cohorts. The χ2 test was used to compare differences between history of febrile convulsion, sex ratio, lateralization, aura occurrence rate, ratio of Engel class I and ratio of Engel class I + II. Differences between aura types and seizure types of the 2 cohorts were also compared with the χ2 test method. Presurgical Workup All patients received more than 2 kinds of noninvasive epileptogenic focus localization tools as presurgical workup, including PET, long-term monitoring EEG and 1.5- and 3-tesla MRI (fig. 1).
Stereotact Funct Neurosurg 2014;92:354–359 DOI: 10.1159/000365575
355
Downloaded by: Freie Universität Berlin 130.133.8.114 - 5/2/2017 2:52:46 PM
TLE), which constitutes at least 60–70% of TLE [1, 2], has the best surgical indication. On magnetic resonance imaging (MRI), HS + TLE usually displays temporal lobe atrophy and signal increase on fluid-attenuated inversion recovery (FLAIR) imaging. Positron emission tomography (PET) usually indicates involvement of the temporal lobe with hypometabolism. After successful standard anterior temporal lobectomies (sATLs), most patients can achieve excellent outcome. However, epilepsy surgeons frequently run into TLE without meaningful abnormality on MRI, which is the so-called MRI-negative TLE (MRI– TLE), and this accounts for 20–30% of all TLE [3, 4]. Multiple onecenter retrospective studies reported that the rate of favorable surgical prognosis (Engel grade I + II) of MRI– TLE is approximately 45–90% [2, 5]. Among these MRI– TLE, many displayed unilateral temporal hypometabolism in PET studies (PET-positive/MRI– TLE). According to one meta-analysis involving 46 independent studies, the predictive value of PET to good surgical outcome (Engel I + II) of MRI– TLE is 80% [6]. Although preoperative invasive electroencephalogram (EEG) is still applied by many centers for such types of TLE, more and more epilepsy centers operate on these patients directly without invasive intracranial electrode implantation [2, 4, 5]. This study retrospectively analyzes 60 cases of unilateral TLE which were surgically treated in our department, including 41 cases of HS + TLE and 19 cases of MRI– TLE. PET studies revealed hypometabolism of the unilateral temporal lobe in all cases of MRI– TLE. Retrospectively, we compare MRI– TLE with HS + TLE, which is well known for the most favorable surgical prognosis, summarizing the clinical characteristics, diagnosis, surgical treatment and prognosis of this entity.
PET Studies Altogether, 49 patients underwent interictal FDG-PET studies, among whom there were 30 cases of HS + TLE and 19 cases of MRI– TLE. All studies were carried out in the PET center of our hospital. PET hypometabolism was determined by unilateral decreased 18FDG uptake in the temporal lobe by visual inspection of more than 2 experienced physicians who had no prior knowledge of the clinical data of these patients (fig. 2). Long-Term Monitoring EEG Long-term monitoring scalp EEG was monitored by a 16- or 32-channel EEG system. Off-line EEG analysis was conducted and reported by 3 experienced EEG experts. Ictal descriptions were written by EEG technicians and epileptologists.
Resective Surgeries and Histopathology Studies Preresection and postresection electrocorticography (EcoG) were performed during each surgery. Operation technique: we performed tailored sATL which included unilateral anterior temporal lobe resection and amygdalohippocampectomy. All histopathology studies were performed in the department of neuropathology in our hospital. Follow-up Follow-ups were conducted by telephone or when patients made return visits to our hospital for consultation. Surgical outcomes were evaluated as follows according to the Engel grading system [9]: Engel class I: seizure free or free of disabling seizures; class II, rare seizures per year (less than 3 seizure days); class III, effective (seizure decreased by at least 80%), and class IV, no improvement. Seizures within the first month after operation were excluded. Surgical effectiveness defined by the Engel class of the 2 cohorts was statistically analyzed using the χ2 test method.
Results
General data and semiology of HS + TLE and MRI– TLE are summarized in table 1. We compared statistical differences between the 2 cohorts. General Data There was no significant difference between the 2 cohorts in relation to average age at operation, sex ratio, lateralization, and seizure duration (p > 0.05). However, history of febrile convulsion differed significantly between the 2 cohorts (HS + TLE: 22.0%, MRI– TLE: 10.5%, p < 0.05). 356
Stereotact Funct Neurosurg 2014;92:354–359 DOI: 10.1159/000365575
Fig. 2. Interictal FDG-PET showed striking left temporal hypometabolism, indicating the epileptogenic zone (red and yellow indicate the highest while blue and purple indicate the lowest hypometabolism).
Table 1. General data and semiology of the two cohorts
Cases, n Sex ratio (male:female) Lateralization (left:right) Febrile convulsion, % Age at operation, years Period of disease, months Aura, % Emotional Experiencing Cephalic sensation (dizziness) Visceral-sensory or autonomic Hallucination None Ictal presentation, % Automatisms Language alterations Absence Secondary GTCS Removing clothes Ictal laughter Spitting Unilateral dystonic posturing Head version Eye deviation
HS + TLE
MRI– TLE
41 25:16 20:21 22 22.8 109.9
19 11:8 10:9 10.5 24.4 137.4
12.1 17.1 14.6 31.7 7.3 22.0
10.5 10.5 10.5 31.6 5.3 36.8
36.6 4.9 26.8 41.5 2.4 0 0 24.4 4.9 4.9
47.4 15.8 36.8 73.7 0 0 0 26.3 10.5 0
A patient may complain of more than 1 type of aura or ictal presentation.
Feng/Hu/Pan/Shi/Qiu/Lang/Gu/Guo
Downloaded by: Freie Universität Berlin 130.133.8.114 - 5/2/2017 2:52:46 PM
MRI Studies All patients in the 2 cohorts were studied with hippocampus scanning, which included sagittal T1-weighted image (T1WI) localization (5 mm), axial T2WI scan (7 mm) and tilted coronal T1WI, T2WI and FLAIR scans (3 mm). MRI scanners were either 1.5 or 3.0 T.
Presurgical Workup PET Studies Among all 49 patients who were the subjects of 18FDGPET studies, 29 HS + TLE and 19 MRI– TLE exhibited unilateral temporal lobe hypometabolism (HS + TLE: 96.7%, MRI– TLE: 100%). Long-Term Monitoring EEG Long-term monitoring EEG revealed interictal or ictal epileptic discharges. We excluded cases which displayed no abnormality on EEG or cases in which epileptic discharges were confined to the contralateral side of the operation, i.e. cases with ipsilateral (concordant with PET or MRI) or bilateral discharges were both included. The relationship between EEG patterns and the prognosis of MRI– TLE is summarized in table 2.
Table 2. EEG patterns and prognosis of the MRI– TLE cohort
EEG pattern (interictal and ictal) Ipsilateral Bilateral
Cases
Engel I Engel I + II
14 5
9/14 4/5
12/14 5/5
Resective Surgeries All 60 resective surgeries were performed by 5 experienced neurosurgeons. All of these operations were sATLs under EcoG monitoring. The extent of temporal neocortex resection was 3–4 cm from the temporal pole on the left side and 4–5 cm on the right side. The resection range of the hippocampus was 3 cm from the hippocampal head; 1 patient suffered from postoperative complication of acute epidural hematoma, but survived with no neurological deficit. None of the patients underwent postoperative disabling neurological deficit or surgery-related death. Histopathology For HS + TLE and MRI– TLE cases, histopathology studies can demonstrate hypoxic, ischemic and dysplastic change of neocortical and hippocampal neurons; sometimes laminar disorganization of layering as well as gliocyte proliferation are present. Follow-up These patients were followed up postoperatively from 0.6 to 11.0 years – average 4.0 years (HS + TLE: 4.7 years, MRI– TLE: 4.4 years). To evaluate the surgical prognosis of the 2 cohorts, we statistically compared the Engel class I ratio as well as the Engel class I + II ratio. There seemed to be no significant difference between the 2 cohorts when comparing the Engel class I ratio (p > 0.05; table 3). When comparing the Engel class I + II ratio, we found that there was no significant difference between the 2 cohorts (HS + TLE: 80.5%, MRI– TLE: 84.2%, p > 0.05; table 3).
MRI Studies All 60 patients took a 1.5- or 3.0-tesla MRI examination, to which we added thin-sliced coronal scans for mesial TLE in order to detect hippocampal lesions. We divided our cases into 2 cohorts according to MRI presentation, as follows: (1) HS + TLE (41 cases): MRI showed unilateral atrophy of mesial temporal structures and hyperintensity in FLAIR sequence with/without unilateral enlargement of the temporal horn; we confined this cohort to real HS in order to prevent interference of non-HS cases to surgical prognosis evaluation and (2) MRI– TLE (19 cases): MRI revealed no meaningful structural or signal anomaly in temporal lobes or other cortex.
MRI showed high specificity (83%) and sensitivity (97%) for the diagnosis of HS [10]. However, even if the resolution of MRI keeps elevating, there are still many TLE cases in which no meaningful abnormality can be
MRI-Negative PET-Positive Temporal Lobe Epilepsy
Stereotact Funct Neurosurg 2014;92:354–359 DOI: 10.1159/000365575
Discussion
357
Downloaded by: Freie Universität Berlin 130.133.8.114 - 5/2/2017 2:52:46 PM
Semiology We divided aura types into experiencing auras (déjà vu, forced thoughts, undefinable feeling), visceral-sensory or autonomic auras (abdominal sensation or epigastric rising sensation, vomiting, alterations in cardiac frequency and rhythm, cyanosis, pallor, piloerection), emotional auras (fear and anxiety), hallucination, cephalic sensation (dizziness), and no auras. All the aura types between the 2 cohorts were not significantly different (p > 0.05). We divided ictal presentation types into oral alimentary and manual automatisms, secondary generalized tonic-clonic seizure (GTCS), language alterations, unilateral dystonic posturing, absence, ictal laughter, removing clothes, eye deviation, head version, and ictal spitting. Most of them were not significantly different, except for the occurrence of secondary GTCS which differed between the 2 cohorts (HS + TLE: 41.5%, MRI– TLE: 73.7%, p < 0.05).
Cases Average postoperative period, years Engel grading, % I II I + II III IV
HS + TLE
MRI– TLE
41 4.43
19 4.10
68.3 12.2 80.5 17.1 2.43
68.4 15.8 84.2 10.5 5.3
observed. Such cases are classified as MRI– TLE, accounting for approximately 20–30% of TLE [3, 4, 10, 11]. Recently, many MRI– TLE are surgically treated and the prognoses are satisfactory. Some studies show that surgical specimens (specifically temporal neocortex and hippocampus) from MRI– TLE patients are similar to HS + TLE histopathologically [3]. Therefore, there is an existing viewpoint which considers MRI– TLE as HS without mesial temporal volume atrophy and signal change, which stands for a distinct subtype of HS. But there are disagreements: many neurologists attempted to prove that MRI– TLE was distinct from HS + TLE in aspects of clinical presentation, semiology, EEG features, histopathology, and surgical prognosis. Chassoux et al. [12] and Abou-Hamden et al. [13] found that histopathologically MRI– TLE resulted from microencephalocele or focal cortical dysplasia in the temporal lobe. Carne et al. [2] demonstrated that the extent of hypometabolism of MRI– TLE is broader than HS + TLE, suggesting that they are distinct from each other. Data from our study show that history of febrile convulsion and occurrence of secondary GTCS differ significantly between MRI– TLE and HS + TLE, supporting the point that to some extent they might be distinct from each other. Among many noninvasive presurgical workup tools for MRI– TLE, PET has unique a lateralizing value and is associated with good surgical prognosis [6, 14, 15]. Although 18FDG is limitedly harmful to the human body, we suggest, based on evidence from multiple studies (including our study) [2, 4, 12], that all MRI– TLE patients should undergo PET examination to see whether they are potential surgical candidates. The reason for hypometabolism in the temporal lobe of most TLE patients may not be the HS itself, but rather the long-term epileptic discharges which cause functional changes and disconnections in the neurons. 358
Stereotact Funct Neurosurg 2014;92:354–359 DOI: 10.1159/000365575
It is well known that the range of PET hypometabolism, which stands for the functional deficit zone, tends to be larger than the real epileptogenic focus, thus the lateralization value of PET is more important than the localization value, and resective surgeries should be tailored under EcoG monitoring. With regard to surgical treatment, one consensus is that the surgical prognosis of TLE is intimately associated with the resection extent of the temporal lobe. The study of Ojemann [16] showed that tailored resection of the neocortical temporal lobe under EcoG monitoring can result in excellent surgical prognosis while performing limited amygdalohippocampectomy, while on the contrary, Wyler et al. [17] and Nayel et al. [18] believed that the resective extent of the mesial instead of the neocortical temporal lobe is associated with surgical prognosis. Our study did not focus on such quantitative studies. However, we recommend the application of subdural EcoG monitoring during sATL surgeries. The resection extent of the neocortical temporal lobe is usually 3–4 cm on the left side and 4–5 cm on the right side from the temporal pole. Intraoperative EcoG are deemed as one of multiple indicators for the resection extent. For the hippocampus, the resection extent would be at least 3 cm from the hippocampal head. Many epilepsy centers considered that the surgical outcome of MRI-negative epilepsy is worse than lesional epilepsy [4, 19]. Thus, a huge number of such patients have to undergo intracranial electrode implantation to localize the epileptogenic focus, or even do not have the chance to receive surgery. In our clinical work, for MRInegative/PET-positive TLE, we did not take the invasive electrode implantation as a routine presurgical workup procedure, but its surgical prognosis was as good as HS + TLE, which is considered as one of the best surgical candidates. Moreover, previous literature supported the hypothesis that for such intractable epilepsy all nonconcordant factors should be excluded in order to perform ‘safe’ surgeries, which include the concordant side of EEG abnormal discharges with PET localization [2, 4]. Similarly, in our 2 cohorts, we excluded patients who had scalp EEG findings that were confined contralaterally to lateralization of PET or MRI studies. We support that PET-positive, MRI– TLE can be surgically treated by sATL, regardless of whether EEG findings are confined to the ipsilateral or bilateral sides (table 2). Although the comparison may not be statistically significant because of the low sample size, one can tell that the prognosis of the ‘bilateral’ group does not have to be poorer than the ‘ipsilateral’ group. Furthermore, for the current status of epiFeng/Hu/Pan/Shi/Qiu/Lang/Gu/Guo
Downloaded by: Freie Universität Berlin 130.133.8.114 - 5/2/2017 2:52:46 PM
Table 3. Comparison of Engel grades between the two cohorts
lepsy surgery, if the PET study of MRI– TLE is negative or ambiguous, surgical treatment should not be performed until successful intracranial electrode implantation is performed to localize the epileptogenic focus. Standardized semiology and EEG evaluation should be repeated to make sure such cases are real TLE. There may be different views on whether MRI can reveal minor abnormal structures or signals if its resolution is elevated high enough. What we emphasize is that even
if MRI cannot provide sufficient evidence for lateralization or localization, unilateral temporal hypometabolism in a PET study should be one of the important indicators of sATL surgery in TLE patients.
Disclosure Statement The authors have no conflicts of interest to declare.
References
MRI-Negative PET-Positive Temporal Lobe Epilepsy
7 Pascual MRQ: Temporal lobe epilepsy: clinical semiology and neurophysiological studies. Semin Ultrasound CT MR 2007;28:416–423. 8 Marks WJ, Laxer KD: Semiology of temporal lobe seizures: value in lateralizing the seizure focus. Epilepsia 1998;39:721–726. 9 Wieser HG, Blume WT, Fish G, Goldensohn E, Hufnagel A, King D, Sperling MR, Luders H; Commission on Neurosurgery of the International League against Epilepsy: Proposal for a new classification of outcome with respect to epileptic seizures following epilepsy surgery. Epilepsia 2001;42:282–286. 10 Berkovic SF, McIntosh AM, Kalnins RM, Jackson GD, Fabinyi GCA, Brazenor GA, Bladin PF, Hopper JL: Preoperative MRI predicts outcome of temporal lobectomy – an actuarial analysis. Neurology 1995;45:1358–1363. 11 Wiebe S, Blume WT, Girvin JP, Eliasziw M; Effectiveness and Efficiency of Surgery for Temporal Lobe Epilepsy Study Group: A randomized, controlled trial of surgery for temporal-lobe epilepsy. New Engl J Med 2001; 345:311–318. 12 Chassoux F, Rodrigo S, Semah F, Beuvon F, Landre E, Devaux B, Turak B, Mellerio C, Meder JF, Roux FX, Daumas-Duport C, Merlet P, Dulac O, Chiron C: FDG-PET improves surgical outcome in negative MRI Taylortype focal cortical dysplasias. Neurology 2010; 75:2168–2175.
13 Abou-Hamden A, Lau M, Fabinyi G, Berkovic SF, Jackson GD, Mitchell LA, Kalnins R, Fitt G, Archer JS: Small temporal pole encephaloceles: a treatable cause of ‘lesion negative’ temporal lobe epilepsy. Epilepsia 2010; 51:2199–2202. 14 Radtke RA, Hanson MW, Hoffman JM, Crain BJ, Walczak TS, Lewis DV, Beam C, Coleman RE, Friedman AH: Temporal-lobe hypometabolism on PET – predictor of seizure control after temporal lobectomy. Neurology 1993;43:1088–1092. 15 Dupont S, Semah F, Clemenceau S, Adam C, Baulac M, Samson Y: Accurate prediction of postoperative outcome in mesial temporal lobe epilepsy – a study using positron emission tomography with 18fluorodeoxyglucose. Arch Neurol 2000;57:1331–1336. 16 Ojemann GA: Surgical therapy for medically intractable epilepsy. J Neurosurg 1987; 66: 489–499. 17 Wyler AR, Hermann BP, Somes G: Extent of medial temporal resection on outcome from anterior temporal lobectomy – a randomized prospective study. Neurosurgery 1995; 37: 982–990. 18 Nayel MH, Awad IA, Luders H: Extent of mesiobasal resection determines outcome after temporal lobectomy for intractable complex partial seizures. Neurosurgery 1991;29:55–61. 19 Tellez-Zenteno JF, Ronquillo LH, Moien-Afshari F, Wiebe S: Surgical outcomes in lesional and non-lesional epilepsy: a systematic review and meta-analysis. Epilepsy Res 2010;89: 310–318.
Stereotact Funct Neurosurg 2014;92:354–359 DOI: 10.1159/000365575
359
Downloaded by: Freie Universität Berlin 130.133.8.114 - 5/2/2017 2:52:46 PM
1 Cascino GD, Jack CR, Parisi JE, Sharbrough FW, Hirschorn KA, Meyer FB, Marsh WR, Obrien PC: Magnetic resonance imagingbased volume studies in temporal lobe epilepsy – pathological correlations. Ann Neurol 1991;30:31–36. 2 Carne RP, O’Brien TJ, Kilpatrick CJ, MacGregor LR, Hicks RJ, Murphy MA, Bowden SC, Kaye AH, Cook MJ: MRI-negative PETpositive temporal lobe epilepsy: a distinct surgically remediable syndrome. Brain 2004;127: 2276–2285. 3 Bien CG, Szinay M, Wagner J, Clusmann H, Becker AJ, Urbach H: Characteristics and surgical outcomes of patients with refractory magnetic resonance imaging-negative epilepsies. Arch Neurol 2009;66:1491–1499. 4 LoPinto-Khoury C, Sperling MR, Skidmore C, Nei M, Evans J, Sharan A, Mintzer S: Surgical outcome in PET-positive, MRI-negative patients with temporal lobe epilepsy. Epilepsia 2012;53:342–348. 5 Immonen A, Jutila L, Muraja-Murro A, Mervaala E, Aikia M, Lamusuo S, Kuikka J, Vanninen E, Alafuzoff I, Ikonen A, Vanninen R, Vapalahti M, Kalviainen R: Long-term epilepsy surgery outcomes in patients with MRInegative temporal lobe epilepsy. Epilepsia 2010;51:2260–2269. 6 Willmann O, Wennberg R, May T, Woermann FG, Pohlmann-Eden B: The contribution of 18F-FDG PET in preoperative epilepsy surgery evaluation for patients with temporal lobe epilepsy a meta-analysis. Seizure 2007; 16:509–520.
