The relevance of somatosensory auras in refractory temporal lobe epilepsies

Transcription

1 BRIEF COMMUNICATION The relevance of somatosensory auras in refractory temporal lobe epilepsies Ghazala Perven, Ruta Yardi, Juan Bulacio, Imad Najm, William Bingaman, Jorge Gonzalez-Martinez, and Lara Jehi SUMMARY Ghazala Perven recently completed her fellowship at the Cleveland Clinic and will be joining the University of Texas Southwestern as an assistant professor. The purpose of this study is to look at the prevalence, characteristics, and prognostic value of somatosensory auras (SSAs) in patients who have undergone temporal lobe epilepsy (TLE) surgery to treat drug-resistant focal epilepsy. We retrospectively reviewed all patients with drug-resistant epilepsy who underwent TLE surgery at Cleveland Clinic between 2005 and 2010 (n = 333) to study the prevalence, characteristics, and prognostic implications of SSA in the context of TLE surgery. Analyses were performed using two seizure outcome definitions: complete seizure freedom and Engel classification. Of the 333 patients, 26 (7.8%) had SSA. Almost half (12 patients) had unilateral sensory symptoms, whereas the rest had bilateral symptoms. Tingling and numbness were the most frequently reported sensations. Compared to their non-ssa counterparts, patients with SSA had the same clinical and imaging characteristics, but had a higher rate of breakthrough seizures (p = 0.03), although most (54%) were still able to achieve Engel class of I (p = 0.02). Based on our results we would encourage detailed presurgical testing, which may include an invasive evaluation to analyze the extent of the epileptogenic zone in patients with SSA and suspected TLE. KEY WORDS: Somatosensory aura, Refractory temporal lobe epilepsy, Surgical outcome. Somatosensory auras (SSAs) are reported infrequently in temporal lobe epilepsy (TLE). Their presence in presumed TLE raises the question of potential extratemporal seizure onset, a possibility that may hypothetically compromise the chances of seizure freedom following resective TLE surgery. The postoperative seizure outcome literature available on SSA in the context of TLE surgery is in fact conflicting: Some studies have reported a favorable outcome, and hypothesize that the aura is due to rapid ictal spread to extratemporal sites, 1,2 whereas other studies found a higher risk for surgical failure and suggest that the SSA reflect an extratemporal primary Accepted July 8, 2015; Early View publication August 7, Cleveland Clinic Epilepsy Center, Neurological Institute, Cleveland Clinic, Cleveland, Ohio, U.S.A. Address correspondence to Lara Jehi, S51 Cleveland Clinic Epilepsy Center, Cleveland Clinic, Cleveland, OH 44195, U.S.A. jehil@ ccf.org Wiley Periodicals, Inc International League Against Epilepsy sensory cortex or supplementary somatosensory cortexepileptogenicity. 3 These conflicting published assessments of the prognostic significance of SSA in TLE may be due to heterogeneous designs, as few studies analyzed SSA among other auras in TLE, all had a small sample size, and all except one 1 were retrospective. In this article, we will examine a large, well-studied cohort of patients who have undergone TLE surgery to treat drug-resistant epilepsy. We aim to determine the prevalence, characteristics, and prognostic value of SSAs in this context. Methods Patient selection We retrospectively reviewed all patients who had TLE surgery at Cleveland Clinic between 2005 and 2010 to treat drug-resistant focal epilepsy. We excluded patients with <6 months of follow-up. Data collected included clinical e143

2 e144 G. Perven et al. characteristics, radiologic and pathologic findings, and specific somatosensory aura information. All patients underwent scalp video electroencephalography (EEG) monitoring and high-resolution magnetic resonance imaging (MRI). After case discussions in a multidisciplinary patient management conference, a decision was made about further surgical management. Outcome information was obtained from follow-up clinic visits or direct patient calls. Patients were classified as either seizure-free or as exhibiting seizure recurrence. Two levels of the seizure-free definitions were recorded for each patient: In the first definition, a patient was considered seizure-free only if he/she did not have any seizures after surgery. In the second definition, seizure freedom was the equivalent of an Engel class I at last follow-up (allowing for auras, breakthrough seizures with missed antiepileptic drugs, and some initial postoperative seizures granted eventual seizure remission occurred). Statistical method Prior to modeling, the data were summarized with descriptive statistics for each variable including means, medians, and standard deviations for continuous variables and frequencies for categorical variables. For exploratory purposes, an initial analysis of the data was first performed using chi-square and Fisher s exact tests to compare: 1 Patients with SSA to those without SSAs. This comparison allowed us to study the characteristics of the SSAs in detail. 2 Seizure-free patients to those with seizure recurrence, regardless of follow-up time. This allowed identification of potential seizure outcome prognostic indicators, including the presence of SSAs. Variables with a significance level of 10% on initial univariate analysis were then tested in a multivariate Cox proportional hazards regression model, and were subsequently considered statistically significant at the 5% level. This method allowed for the testing of the correlation of specific variables with outcome while taking into account any interactions and associations among those variables, and their variation with time. Results Overall patient characteristics Three hundred thirty-three patients who had undergone temporal lobectomy for drug- resistant TLE fulfilled study criteria and were analyzed. These included 264 patients (78.2%) who proceeded directly to a resective surgery after the multidisciplinary management conference, whereas the remainder required an invasive evaluation (subdural electrodes [SDEs], or SDEs + depths, or stereo-eeg). Table 1 summarizes the patients baseline clinical, imaging, pathologic characteristics, and seizure outcome. Table 1. Clinical characteristics of the overall cohort Overall group Characteristics (n = 333) Female (%) 171 (51) Side of temporal lobectomy Left (%) 188 (56) Right (%) 145 (44) MRI Normal (%) 52 (15) Abnormal (%) 280 (84) PET Localized (%) 284 (85) Nonlocalized (%) 5 (1) Not done (%) 44 (14) Preoperative GTCS present (%) 259 (77) Median preoperative seizure frequency/month (range) 6 ( ) Median age at seizure onset, years (range) 12 (0.3 65) Median age at surgery, years (range) 34 (0.9 74) Younger than 17 years old (%) 64 (19) 17 years or older (%) 269 (81) Median epilepsy duration, years (range) 14 (0.3 64) Auras present (%) Somatosensory aura 26 (8) Psychic aura 80 (24) Abdominal aura 93 (28) Invasive evaluation performed (%) 94 (28) Pathologic subgroups (%) MTS 118 (35) MCD 56 (16) Vascular 35 (10) Gliosis 70 (21) Tumor 41 (12) Other 15 (6) None 5 (1) Mean follow-up duration, years (range) 3.9 ( ) Seizures recurred (%) 179 (54) Causes of failure (%) AED reduction 64 (19) No reason 113 (34) Others 2 (0.6) No recurrence 154 (46) Engel class (%) I 245 (73) II 44 (14) III 31 (9) IV 13 (4) PET, positron emission tomography; GTCS, generalized tonic clonic seizures; MTS, mesial temporal sclerosis; MCD, malformation of cortical development; AED, antiepileptic drug. Somatosensory aura: prevalence and patient characteristics Of the 333 patients, 26 (7.8%) had SSA. Of these 26 patients, 15 were male, 20 had a prior history of generalized tonic clonic seizures (GTCS), 20 had an abnormal MRI, and 23 had an abnormal positron emission tomography (PET). Ten patients with SSAs underwent invasive evaluation prior to proceeding to resection, with the majority (9 of 10) undergoing subdural grid with some depth electrodes. Only one patient underwent stereo-eeg. Seventeen patients had left temporal lobectomy, whereas nine had right temporal resection. Almost half (12 patients) had unilateral

3 e145 Somatosensory Aura in TLE sensory symptoms, with three reporting symptoms ipsilateral to the epileptogenic zone, whereas nine patients reported contralateral symptoms. Tingling and numbness were the most frequently reported sensations. Other reported sensations were warmth, cold, and sizzling. Twenty patients reported other types of auras as well. Three patients had associated psychic aura and three patients had associated abdominal aura, whereas one patient had both types of auras. Other types of auras reported were gustatory, olfactory, auditory, and unclassified. Table 2 summarizes the clinical, imaging, and pathologic characteristics and outcomes of the SSA cohort. There were no differences in these clinical, imaging, and pathologic characteristics between the group of patients with SSA and those without SSA in this cohort. Seizure outcome analysis On the initial univariate screening, preoperative history of GTCS, invasive evaluation, and presence of SSA predicted an unfavorable outcome using both seizure outcome measures (with respective p-values of 0.02, <0.01, and 0.03 for the complete seizure freedom analysis; and respective p- values of 0.04, 0.02, and 0.02 for the Engel class I analysis). On univariate analysis, PET abnormalities, invasive evaluation, pathology, and duration of epilepsy did not show any significant impact on the prognosis. Multivariate analysis Upon applying proportional hazard modeling, the history of previous GTCS retained its significance as an independent predictor of unfavorable outcome for both seizure recurrence and Engel class I (p = 0.04 and p = 0.04, respectively). Invasive evaluation retained significance only insofar as predicting any seizure recurrence (p =<0.01), whereas the presence of SSA remained significant only in its prediction of the Engel class I (p = 0.02). The laterality of the SSA (unilateral vs. bilateral), its association with other aura types (psychic vs. abdominal vs. other aura types), and its nature (tingling vs. other sensations) did not affect its relation with outcomes. Patients with SSA and an abnormal MRI did worse than patients without SSA (p = 0.02). Discussion Somatosensory auras The reported prevalence of SSA in TLE varies between 1.7% and 11%. 1,4 6 The significance of SSA in TLE is largely unknown. In this study, we are reporting a prevalence of 7.8% of SSA in patients who underwent temporal lobectomy to treat presumed drug-resistant TLE. This prevalence is concordant with the prior range of % 5 reported previously, specifically in cohorts with drug-resistant TLE. The lack of any clinical, imaging, or pathologic differences between the patient group with SSA as compared to that without SSA suggests that this semiologic seizure manifestation is purely an electroclinical manifestation of the epilepsy, rather than a feature that should be used to infer conclusions on etiology or pathologic substrate. The presence of SSA correlated with lower chances of complete seizure freedom and eventual remission. This finding held true even in patients with an abnormal brain MRI, possibly suggesting a false sense of security about the extent of the epileptogenic zone in the presence of a lesion. This would hypothetically lead to avoidance of a more extensive evaluation to analyze the extent of the epileptogenic zone by means of intracranial recordings. The observation of poorer seizure outcomes in patients with SSA compared to those without SSA, despite similar clinical characteristics otherwise between the two groups, supports that SSAs corroborate a poorer prognosis regardless of the remaining clinical picture and should trigger a more careful assessment of the localization hypothesis. Prognostic value of other variables In this study we again confirm the presence of GTCS and a history of invasive evaluation as negative markers of prognosis. Invasive EEG When compared to patients who proceeded directly to resective surgery, those with invasive evaluation either subdural grids or stereo-eeg were more likely to have ANY postoperative seizure recurrence on multivariate analysis, although they achieved comparable Engel class I by last follow-up. These results may be attributed to more complicated epilepsies with discordant preoperative diagnostic data in patients who eventually require invasive EEG. 7 Generalized tonic clonic seizures Almost eighty percent of the patients with preoperative GTCS had seizure recurrence, compared to 22% in those who never had GTCS preoperatively (p = 0.04). Previous studies have found similar results In our group, these patients were able to achieve excellent seizure outcome eventually based on Engel score. This observation has been attributed to wider epileptic networks or a lower overall seizure threshold in patients who have GTCS in the context of presumed TLE. Study limitation This is a retrospective study based on chart review so there are various limitations and chances of bias. Documentation of auras depends on the history-taking skills, detailed attention, and documenting skills of the examiner. It is also dependent on the recall of the patient. Another limitation is that this study included only patients who had undergone TLE surgery. Patients who had TLE but refused surgery or who were deemed nonsurgical candidates (for various other reasons) were not included.

4 e146 G. Perven et al. Table 2. Characteristics of patients with somatosensory auras Patient (no; gender) GTCS Type of SSA Unilateral or bilateral Location Associated auras Pathology Radiology Invasive Seizure recurrence SSA recurred Engel outcome Followup period (years) 1; F Y Tingling U/L Lips and left arm and leg 2: M Y Tingling U/L R hand first 2 fingers 3: M N Warm/tingling B/L Warmth in face and head and tingling in fingers Olfactory MTS MRI- ABL Unclassified Gliosis MRI- ABL PETnonlocalizable Unclassified Gliosis MRI- ABL 4; M Y Tingling U/L Right hand None NONSPECIFIC MRI- ABL 5; M N Numbness U/L R hand fingers Gustatory GLIOSIS MRI- ABL PETnonlocalizable 6; M Y Sizzling U/L Left hemibody Psychic TUM MRI- ABL 7; F Y Numbness B/L Whole body Psychic/ abdominal 8; M Y Tingling U/L Left face, arm Auditory/ olfactory 9; M Y Tingling U/L Left arm Gustatory/ olfactory 10; F N Cold/tingling B/L Cold on right body and tingling on left side of head 11; M Y Tingling B/L Whole Body/right leg MTS MRI- NL Infarction MRI- ABL Gliosis MRI- ABL None Gliosis MRI- ABL Auditory MTS MRI- ABL 12; M N Numbness B/L Whole body numb Abdominal MTS MRI- ABL 13; M Y Tingling U/L Right hand Gustatory MTS MRI- ABL 14; F Y Tingling B/L Feet?body None INFARCTION MRI- ABL 15; F Y Numbness B/L Around mouth/ hands Gustatory/ olfactory GLIOSIS MRI- NL PETnonlocalizable 16; F Y Tingling B/L Hands, lips None GLIOSIS MRI- NL 17; M N Tingling B/L Occiput Gustatory NONE MRI- ABL 18; F Y Tingling B/L Whole body Psychic GLIOSIS MRI- ABL N Y Y II 8 Y Y Y IV 5 Y Y Info unavailable I 5 Y Y N I 8 Y Y N I 8 N Y Y III 1 Y N N I 5 Y Y Y II 8 N Y N III 7 N N N I 4 N Y N III 5 N N N I 1 N N N I 0.5 Y Y N I 4 N Y Y IV 5 Y N N I 4 N Y Y II 5 N Y Y I 4 Continued

5 e147 Somatosensory Aura in TLE Table 2. Continued. Patient (no; gender) GTCS Type of SSA Unilateral or bilateral Location Associated auras Pathology Radiology Invasive Seizure recurrence SSA recurred Engel outcome Followup period (years) 19; F Y Tingling/chilly U/L Neck?left side of body 20; M Y Numbness U/L Left head?left arm leg 21; F Y Numbness U/L Right hand (index and middle finger) Unclassified MTS MRI- ABL Abdominal TUM MRI- ABL None TUM CT- ABL 22; M Y Tingling/warm B/L Whole body None MCD MRI- NL 23; F N Tingling B/L Arms Unclassified/ gustatory 24; M Y Tingling B/L Lower back and legs GLIOSIS MRI- NL Psychic INFARCTION MRI- ABL 25; F Y Numbness B/L Legs Abdominal MTS MRI- ABL 26; M Y Tingling U/L Right 5th digit Gustatory TUM MRI- ABL N Y N III 3 N Y Y III 4 N N N I 5 Y Y N I 2 Y N N I 4 N Y N II 5 N Y N I 8 N Y N III 3 SSA, somatosensory aura; ABL, abnormal; N, normal; U/L, unilateral; B/L, bilateral; TUM, tumor; MTS, mesial temporal sclerosis; MCD, malformation of cortical development; MRI, magnetic resonance imaging.

6 e148 G. Perven et al. The presence of SSA in temporal lobe seizures is considered by some as an indicator of neocortical or extratemporal epilepsy. 3,4 In our study we included only the patients who had undergone temporal lobectomy. Therefore, the suspicion of the seizure originating from temporal lobe was very high based on noninvasive and invasive data. Our data suggest, however, that the presence of SSA could highlight a more extensive epileptogenic zone. Only 10 of the 26 patients underwent invasive evaluation. We analyzed our invasive data for the 10 patients with SSA. Nine of the 10 patients were evaluated with grids in combination with depth electrodes covering mainly the mesial temporal structures. Only one patient underwent stereo-eeg with electrodes in insula but not in the parietal operculum and primary sensory cortex. This patient did not have his typical SSA during the recording. Therefore, this study was insufficient to explore the intracranial spread patterns in this population. Conclusion Herein we report a large cohort of patients with TLE who underwent temporal lobectomy to treat their medically intractable epilepsy. We found that the presence of SSA was associated with a higher rate of recurrence, but many patients were still able to achieve excellent seizure control after the initial breakthrough seizure. Based on our results we would encourage detailed presurgical testing in patients with SSA and suspected temporal lobe epilepsy, with particular attention to entertaining alternative localization hypotheses that extend beyond the temporal lobe to include the parietal operculum, insula, and primary somatosensory cortex. Disclosure None of the authors has any conflict of interest to disclose. We confirm that we have read the Journal s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines. References 1. Erickson JC, Clapp LE, Ford G, et al. Somatosensory auras in refractory temporal lobe epilepsy. Epilepsia 2006;47: Tuxhorn IE. Somatosensory auras in focal epilepsy: a clinical, video EEG and MRI study. Seizure 2005;14: Aghakhani Y, Rosati A, Dubeau F, et al. Patients with temporoparietal ictal symptoms and inferomesial EEG do not benefit from anterior temporal resection. Epilepsia 2004;45: Palmini A, Gloor P. The localizing value of auras in partial seizures: a prospective and retrospective study. Neurology 1992;42: Weil AG, Surbeck W, Rahme R, et al. Somatosensory and pharyngolaryngeal auras in temporal lobe epilepsy surgeries. ISRN Neurol 2013;2013: Kotagal P, Luders HO, Williams G, et al. Psychomotor seizures of temporal lobe onset: analysis of symptom clusters and sequences. Epilepsy Res 1995;20: Isnard J, Guenot M, Ostrowsky K, et al. The role of the insular cortex in temporal lobe epilepsy. Ann Neurol 2000;48: Jehi LE, Silveira DC, Bingaman W, et al. Temporal lobe epilepsy surgery failures: predictors of seizure recurrence, yield of reevaluation, and outcome following reoperation. J Neurosurg 2010;113: Jeha LE, Najm IM, Bingaman WE, et al. Predictors of outcome after temporal lobectomy for the treatment of intractable epilepsy. Neurology 2006;66: Hennessy MJ, Elwes RD, Rabe-Hesketh S, et al. Prognostic factors in the surgical treatment of medically intractable epilepsy associated with mesial temporal sclerosis. Acta Neurol Scand 2001;103: McIntosh AM, Kalnins RM, Mitchell LA, et al. Temporal lobectomy: long-term seizure outcome, late recurrence and risks for seizure recurrence. Brain 2004;127: