Semiology of Temporal Lobe Seizures: Value in Lateralizing the Seizure Focus

Transcription

1 Epilepsia, 39(7): , 1998 Lippincott-Raven Publishers, Philadelphia 0 International League Against Epilepsy Semiology of Temporal Lobe Seizures: Value in Lateralizing the Seizure Focus William J. Marks, Jr. and Kenneth D. Laxer Department of Neurology and Northern California Comprehensive Epilepsy Center, University of California, San Francisco, San Francisco, California, U.S.A. Summary: Purpose: To determine the lateralizing value of the clinical manifestations of seizures in patients with temporal lobe epilepsy (TLE), we made a retrospective videotape analysis of complex partial seizures (CPS) in 55 patients who underwent temporal lobectomy and were seizure-free postoperatively for >2 years. Methods: Blinded to clinical details, we reviewed videotapes from video-eeg telemetry monitoring with attention paid to seizure semiology. Results: Useful lateralizing features included unilateral clonic activity (with the seizure focus contralateral in all patients), unilateral dystonic or tonic posturing (with the seizure focus contralateral in 90 and 86%, respectively), unilateral automatisms (with the seizure focus ipsilateral in 80%), and ictal speech preservation (with the seizure focus contralateral to the language-dominant hemisphere in 80%). Versive head rotation occumng < 10 s before seizures secondarily generalized consistently predicted a contralateral focus. Seizure manifestations less predictive but suggestive of lateralization included ictal speech arrest and postictal speech status, with predictive values of 67%. Seizure manifestations not providing reliable lateralizing information included eye deviation, type of aura, and versive head movements occumng at times other than immediately before seizures secondarily generalized. Conclusions: In TLE, several clinical seizure manifestations are useful in lateralizing the seizure focus, although some provide no reliable information. Therefore, ictal semiology can assist in the evaluation of patients for seizure surgery, providing additional information in the lateralization of TLE. Key Words: Temporal lobe epilepsy-complex partial seizures- Semiology-Lateralization. Seizure semiology has long fascinated neurologists and continues to be the clinical touchstone from which seizure classification derives and from which inferences about localization of the epileptogenic focus begin. Investigators have reported the localizing and lateralizing value of various individual seizure phenomena (1-16). Others have catalogued the clinical semiology of complex partial seizures (CPS) (17-24). In the present study, we analyzed the clinical seizure phenomenology of a large group of patients with temporal lobe epilepsy (TLE) in whom the seizure focus had been determined with a high degree of certainty to determine which features are of lateralizing value. METHODS The study entailed a retrospective videotape analysis of all patients with CPS arising from either temporal lobe Accepted February 26, Address correspondence and reprint requests to Dr. K. D. Laxer at Northern California Comprehensive Epilepsy Center, University of California, San Francisco, Box 0138, San Francisco, CA , U.S.A. who had undergone temporal lobectomy (anterior temporal lobectomy and amygdalohippocampectomy) between 1987 and 1992 at the University of California, San Francisco who had >2-year postoperative seizure-free outcome (Engel class I) and for whom follow-up information was available. Patients meeting these criteria were identified by 1 of the authors (K.D.L.), and videotapes archived from video-eeg telemetry monitoring were reviewed by the other author (W.J.M.), who was blinded to clinical details. On telemetry, these patients typically had undergone rapid taper of antiepileptic drugs (AEDs) to facilitate the recording of seizures. Patient characteristic are shown in Table 1. Although the EEG is encoded on the videotapes, only the video images of patient behaviors were viewed on a monitor that did not display the EEG. Audiovisual analysis of 276 CPS in 55 patients (median of 4 seizures per patient, average of 5 seizures per patient, range of 1-18 seizures per patient, and 44 patients with at least 3 seizures) was performed, with attention paid to the following features: sustained versive head rotation and tilt, deviation of face and eyes, unilat- 72 1

2 722 W. J. MARKS, JR. AND K. D. LAXER TABLE 1. Patient characteristics Demographics (n = 55) Sex Age (yr) Side of temporal lobectomy Postoperative follow-up duration (mo) 33 Male, 22 Female Mean 29.5 (range 10-46) 29 L, 26 R Mean 61.8 (range 32-99) Language dominance n Left hemisphere predominant 45 (82) Right hemisphere predominant 4 (7) Equal or indeterminate Pathology 6(11) Mesial temporal sclerosis 41 (75) Tumor 4 (7) Encephalitis 1(2) Normal 1(2) Others 6 (11) Unknown Telemetry 2 (4) Lateralizing ictal scalp recordings 45 (82) Intracranial recordings required 10 (18) Imaging (MRI) (n = 54) Lateralizing temporal abnormality 35(65) Incidental abnormality 7 (13) Normal 12 (22) MRI, magnetic resonance imaging. era1 automatisms, unilateral tonic and dystonic posturing, unilateral clonic activity, and speech status during the ictal and postictal periods. Observations were scored according to absence, presence, and laterality (when appropriate) of each phenomenon in a standardized manner. A feature was considered present if obvious to the reviewer during at least one of the patient's seizures, but it was not required to be present in every seizure. Versive head movements were considered present only if they appeared involuntary and forced, consisted of at least 45" of rotation or tilt, and persisted for it least 10 s. Tonic posturing was considered present if a limb simply assumed a stiffened appearance, whereas dystonic posturing required maintenance of a sustained, ab- normal, contorted limb posture. The presence of hemiparesis was ascertained from the motor examination performed in the postictal period. Automatisms, posturing, and clonic movements were required to be present for a minimum of 10 s. Ictal vocalizations were characterized as complex verbalization (formed words) or nonmeaningful vocalization, either spontaneous or in response to questioning, during a time in which the patient's awareness appeared altered. Postictal speech was assessed during the first 30 s after apparent cessation of clinical seizure activity and was characterized as intact, dysfunctional (if word-finding difficulties or paraphasic errors occurred), or arrested according to a standardized lapguage assessment protocol administered by the nursing staff. After this blinded analysis, clinical information was obtained from a computerized database and medical records. Information on aura type was obtained from patient description on videotape and/or from medical records. Hemispheric dominance for speech function had been determined by the intracarotid amobarbital (Wada) test and was classified as left- or right-dominant based on language performance during the amobarbital-active phase of the study (25). We performed nonparametric statistical analysis using the Sign test to evaluate the significance of the frequency with which the various motor signs occurred. Ictal and postictal speech findings were evaluated by one-tailed Fisher's exact test; a one-tailed test was chosen because a priori hypotheses concerning speech function had been generated on the basis of previous studies. RESULTS Excluding auras and speech status, patients exhibited an average of 2 signs (range 0-6). Twelve of the patients (22%) displayed none of the 'signs of interest. Table 2 shows the frequency and predictive values of the various motor signs. Versive head rotation was the most,common clinical sign, present in nearly half of the TABLE 2. Motor signs Predictive value Frequency, (lateralization relative Feature n to seizure focus)" Versive head rotation C10 s before generalization At any time during partial seizure 15 (27) 25 (45) 100% contralateral (p < 0.01) 60% contralateral (NS) Unilateral dystomc posturing 10 (18) 90% contralateral (p = 0.05) Unilateral tonic posturing 7 (13) 86% contralateral (NS) Eye deviation 14/51 (27) 57% contralateral (NS) Unilateral automatisms 5 (9) 80% ipsilateral (NS) Unilateral clonic movement 6 (11) 100% contralateral (p = 0.05) Head tilt 5 (9) 60% ipsilateral (NS) Face deviation 1 (2) 100% contralateral (NS) Postictal hemiparesis 1 (2) 100% contralateral (NS) a NS, not significant at the 0.05 significance level by the Sign test. Epilepsia, Vol. 39, No. 7, 1998

3 SEMIOLOGY OF TLE 723 patients. It occurred soon (S10 s) before seizures secondarily generalized in 27% of the patients. Unilateral posturing (dystonic or tonic) occurred in nearly one third of the patients, and deviation of the eyes occurred in about one fourth. The other motor signs of interest occurred less frequently. Of the motor signs, the most reliable predictors of seizure lateralization were unilateral clonic activity and versive head rotation just before generalization, indicating contralateral lateralization in all patients. Although facial deviation and postictal hemiparesis indicated a contralateral seizure focus, these signs were present in only 1 patient each. Unilateral posturing (tonic and dystonic combined) predicted a contralateral focus in 88%; unilateral dystonic posturing was significantly associated with a contralateral seizure focus, and although unilateral tonic posturing was associated with contralateral lateralization, the predictive value did not reach statistical significance in the few patients manifesting this sign. Twelve patients manifested more than one motor sign (Table 3). For all but 1 of these patients, who demonstrated ipsilateral dystonic posturing and contraversive head deviation relative to the seizure focus, the associated signs in each patient suggested the same lateralization. Speech findings are shown in Table 4 for patients in whom evaluation from videotapes was possible. Speech arrest commonly (in 74%) occurred during the ictus and indicated seizure lateralization ipsilateral to the language-dominant hemisphere in two thirds of patients. Ictal speech preservation, though less common (occurring in 19% of patients), indicated a seizure focus contralateral to the language-dominant hemisphere in 80% (p = 0.01, Fisher s exact test). Speech function in the immediate postictal period could be adequately evaluated in 41 of the patients. Two thirds of patients with postictal speech preservation demonstrated a seizure focus contralateral to their language-dominant hemisphere, and two thirds with postical speech impairment (either TABLE 3. Patients with multiple motor signs Signs (lateralization relative to seizure focus) Contralateral tonic posturing, contralateral clonic activity, Contralateral clonic activity, Ipsilateral automatisms, contralateral dystonic posturing Contralateral dystonic posturing, Contralateral tonic posturing, contralateral clonic activity Ipsilateral dystonic posturing, No. of patients TABLE 4. Speech findings Predictive value Frequency, (relative to language- Feature n (%) dominant hemisphere) Ictal speech preservation 10/53 (19) 80% contralateral Ictal speech arrest 39/53 (74) 67% ipsilateral Postictal speech preservation 19/41 (46) 68% contralateral Postictal speech impairment 21/41 (51) 67% ipsilateral a p = 0.01, one-tailed Fisher s exact test. p = 0.03, one-tailed Fisher s exact test. speech arrest or dysfunction) proved to lateralize ipsilatera1 to their language-dominant hemisphere (p = 0.03, Fisher s exact test). Twenty-four patients (44%) reported auras (Table 5). The type of aura did not predict seizure lateralization. DISCUSSION We evaluated the lateralizing value of various clinical features of CPS in 55 patients, all with clear temporal lobe localization. At the time the study was designed and undertaken, no large-scale assessment of the lateralizing value of CPS semiology had been reported. Subsequently, Chee et al. published their observations in a study (23) that analyzed 166 seizures in 38 patients, 27 who had undergone temporal lobe resection and 11 who had undergone extratemporal resection. They established that clinical semiology could be readily ascertained from videotape review, with good interobserver agreement. In the present study, unilateral clonic activity, versive head rotation just before seizures secondarily generalized, and unilateral dystonic posturing were the most reliable predictors of seizure lateralization, predicting a contralateral seizure focus in % of patients. Unilateral posturing or clonic activity occurring in combination with ictal speech preservation was always associated with a contralateral, nonlanguage-dominant focus. The study of Chee et al. (23) demonstrated similar findings, with the predictive value for dystonic posturing 93% and that for ictal spech 83%; unilateral tonic posturing and clonic activity were not assessed. In the present study, unilateral dystonic and tonic posturing had similar predictive values (90 and 86%, respectively); prediction of a TABLE 5. Auras Lateralization Frequency Type na (%) % L % R p-valueb Viscerosensory 11 (20) NS Experiential 7 (13) NS Somatosensory 4 (7%) NS Diffuse warmth 3 (5) NS Cephalic sensation 2 (4) NS Twenty-four patients experienced auras, 3 with features in more than one category. NS, not significant at the 0.05 significance level by the Sign test. Epilepsia, Vol. 39, No. 7, 1998

4 724 W. J. MARKS, JR. AND K. D. LAXER contralateral seizure focus reached statistical significance for dystonic posturing only, however. In a study of temporal lobe seizure features by King and Ajmone- Marsan (19) almost 2 decades ago, unilateral tonic limb posturing was a relatively reliable lateralizing finding, with a contralateral seizure focus in 77% of patients. Kotagal et al. (lo), however, in their report documenting the excellent lateralizing value of unilateral dystonic posturing, commented that tonic posturing... although frequently contralateral, is not a reliable lateralizing sign. We noted that unilateral automatisms overall suggested ipsilateral lateralization in 80%. Unilateral automatisms accompanying dystonic posturing of the opposite extremity (3 patients) correctly lateralized the seizure focus in all (with automatic movements in the ipsilateral limb and dystonic posturing in the contralateral limb). In the 2 patients with isolated unilateral automatisms, however, the sign was unreliable (1 ipsilateral, 1 contralateral). Kotagal et al. (10) reported similar findings. In our study, when patients demonstrated multiple motor signs, the signs typically were in agreement and predicted correct lateralization, except in 1 patient. The most common combinations were clonic activity with head deviation contralateral to the seizure focus and ipsilateral automatisms paired with contralateral dystonic posturing. Many of the motor signs under consideration were observed in only a few of the patients. The low prevalence of motor findings can be explained by the stringent criteria we required for a sign to be considered present. In addition, some of the manifestations may have been present but not readily apparent on the videotape (e.g., posturing occurring under bed covers). Similarly, postictal hemiparesis that was subtle or persisted for only a brief time may not have been detected. The presence of postictal neglect, which could have been misinterpreted as postictal hemiparesis, was not specifically^ assessed. Ictal speech preservation was documented in nearly one fifth of the patients and predicted seizure lateralization contralateral to the language-dominant hemisphere in 80%. Koerner and Laxer (8) and Gabr et al. (9) reported the capacity for ictal speech to be reliable in this regard, with predictive values of 92 and 83%, respectively. Morrell et al. (12) did not find the presence of ictal speech useful, however, with fluent speech as likely to occur in patients with dominant mesial temporal lobe seizure onsets as in those with seizures arising from the nondominant mesial temporal lobe. Two thirds of our 39 patients with ictal speech arrest lateralized to the language-dominant hemisphere. In studies by Gabr et al. (9) and Morrell et al. (12), speech arrest occurred in patients with temporal lobe seizures arising from the language-dominant hemisphere in 1 of 3 and 3 of 5 patients, respectively. Postictal speech status has previously been reported to be a more reliable lateralizing feature (8,9). We noted that postictal speech was more likely to be preserved when the seizures arose from the nonlanguage-dominant hemisphere (68%) and was more likely to be impaired if the focus was in the language-dominant hemisphere (67%). Koerner and Laxer demonstrated similar reliability of these features (8), and Gabr et al. (9) considered postictal speech status to be an even more reliable indicators of lateralization (9). Considerable controversy exists regarding the utility of head movements as reliable lateralizing signs. In their series cataloguing the clinical features of nearly 300 patients with TLE, King and Ajmone-Marsan (19) reported that tonic head deviation provided no consistent lateralizing information. In one of the first reports to address this issue specifically, Robillard et al. (4) studied patients with partial seizures arising from frontal, temporal, or parietooccipital regions in whom head turning was the first obvious clinical manifestation of the seizure (4). Head rotation, whether slight, moderate, or marked, failed to provide reliable lateralizing information. Similarly, Ochs et al. (5) reported that initial head-turning in a siezure had no lateralizing significance. Wyllie et al. (6) distinguished between versive (forced and sustained) and nonversive (mild and unsustained) head turning in their analysis of partial seizures of predominantly frontal and temporal localization and observed that versive, but not nonversive, movement was a uniformly reliable sign lateralizing to the contralateral hemisphere. Using similar criteria in a study of temporal lobe seizures published in the same year, Quesney did not replicate this finding, however (22). Echoing the report of Wyllie et al. (6), however, McLachlan (7) reported that forced head turning, as determined by history, was contralateral to the seizure focus in 92% of patients. In their report on dystonic posturing as a lateralizing sign in seizures of temporal lobe onset, Kotagal et al. (10) noted that version, when it occurred, was always toward the side of dystonic limb posturing and always indicated a contralateral focus (10). Jayakar et al. (14) suggested that head movements provide accurate lateralization only when assessed in the context of the different actions of the sternomastoid and cleidomastoid muscle divisions. Chee et al. (23) reported versive head deviation to be a reliable lateralizing sign, with 94% of patients having a contralateral seizure focus apd nonversive head turning predicting an ipsilateral focus in 80%. Finally, Kernan et al. (16) published their observations on head deviation in secondarily generalized seizures, concluding that head deviation was contralateral to the epileptogenic focus in >90% of seizures if the movement either occurred in the 10 s before generalization or continued as the seizure generalized. The reason for the variability of the reliability of this sign in the various studies may be due to the subjective nature of assessing head movements, consideration of head move- Epilepsia, Vol. 39, No. 7, 1998

5 SEMIOLOGY OF TLE 725 ments at different phases during the seizure, and variability of seizure spread patterns in the study populations. In the present study, we also noted that versive head rotation, when it occurred in the 10 s before generalization or as the seizure generalized, indicated a contralatera1 seizure focus in all patients. Versive head rotation at any time during the partial seizure, however, predicted a contralateral focus in only 60% of patients. We did not specifically analyze whether head turning that occurred very early in a CPS was of lateralizing value. We observed head deviation to be toward the side of limb posturing (either tonic or dystonic) in 5 of 6 patients who exhibited both these signs; therefore, in contrast to the data of Kotagal et al. (lo), in our study these signs were not universally concordant. Versive eye deviation occurred in 27% of the patients in our study and was not predictive of seizure lateralization. In the study of Chee et al. (23), eye deviation was considered only when it accompanied versive head movements. In the present study, aura type did not provide useful lateralizing information, which is in agreement with results of several previous studies (1 1,15), although Gupta et al. suggested that autonomic and psychic auras are more often associated with a right-sided focus (3). Because we wished to evaluate the value of seizure semiology in patients with confident seizure localization, only patients who remained seizure-free for at least 2 years after temporal lobectomy were included in our study. Thus, the patients constitute a selected subset of patients with medically refractory TLE, mainly those whose preoperative evaluation was sufficiently concordant for localization that surgery could be offered and those who also obtained a successful postoperative outcome. Our findings may not be completely generalizable to all patients with medically refractory TLE; therefore, one must use caution in applying these findings to the TLE population in general. Although many of our patients had medial TLE, evidenced by pathological demonstration of mesial temporal sclerosis in 75%, some of them probably had seizures originating from the temporal neocortex. There may be differences in seizure semiology in patients with mesial versus neocortical TLE (24), although whether the lateralizing value of the features considered in the present study differs in the two conditions is not clear. In the evaluation of patients with TLE being considered for seizure surgery, ictal semiology often contributes helpful lateralizing information. Although many of the ictal clinical phenomena occur as the result of seizure spread from the site of onset, a stereotypy in these pathways of spread appears to be common. While seizure semiology is but one element to be considered in determination of the epileptogenic region, its concordance or discordance with EEG and neuroimaging findings con- tributes to the level of confidence with respect to this determination. Particularly when the seizure features we have shown to demonstrate reliable lateralization conflict with other aspects of the evaluation, clinicians may wish to proceed with further investigation to arrive at more confident lateralization. Acknowledgment: We thank Dr. Paul A. Garcia for thoughtful comments REFERENCES 1. Hecaen H, Piercy M. Paroxysmal dysphasia and the problem of cerebral dominace. J Neurol Neurosurg Psychiatry 1956; 19: Serafetinides EA, Falconer MA. Speech disturbances in temporal lobe seizures: a study in 100 epileptic patients submitted to anterior temporal lobectomy. Brain 1963;86: Guptka AK, Jeavons PM, Hughes RC, Covanis A. Aura in temporal lobe epilepsy: clinical and electroencephalographic correlation. J NeuroE Neurosurg Psychiatry 1983;46: Robillard A, Saint-Hilaire JM, Mercier M, Bouvier G. The lateralizing and localizing value of adversion in epileptic seizures. Neudogy 1983;33: Ochs R, Gloor P, Quesney F, Ives J, Olivier A. Does head-turning during a seizure have lateralizing or localizing significance? Neurology 1984;34: Wyllie E, Liiders H, Moms HH, Lesser RP, Dinner DS. The lateralizing significance of versive head and eye movements during epileptic seizures. Neurology 1986;36: McLachlan RS. The significance of head and eye turning in seizures. Neurology 1987;37: Koerner M, Laxer KD. Ictal speech, postictal language dysfunction, and seizure lateralization. Neurology 1988;38: Gabr M, Liiders H, Dinner D, Moms H, Wyllie E. Speech manifestations in lateralization of temporal lobe seizures. Ann Neurol 1989;25:82-7. Kotagal P, Liiders H, Morris HH, et al. Dystonic posturing in complex partial seizures of temporal lobe onset: a new lateralizing sign. Neurology 1989; Janati A, Nowack WJ, Dorsey S, Chesser MZ. Correlative study of interictal electroencephalogram and aura in complex partial seizures. Epifepsia 1990;31:41-6. Morrell MJ, Phillips CA, O Connor MJ, SperIing MR. Speech during partial seizures: intracranial correlates. Epilepsia 1991; 32: Privitera MD, Moms GL, Gilliam F. Postictal language assessment and lateralization of complex partial seizures. Ann Neurol ; Jayakar P, Dyhowny M, Resnick T, Alvarez L. Ictal head deviation: lateralizing significance of the pattern of head movement. Neurology Palmini A, Gloor P. The localizing value of auras in partial seizures: a prospective and retrospective study. Neurology 1992; 42: Keman JC, Devinsky 0, Lucian0 DJ, Vazquez 8, Perrine K. Lateralizing significance of head and eye deviation in secondary generalized toniqclonic seizures. Neurology 1993;43: Bingley T. Mental symptoms in temporal lobe epilepsy and temporal lobe gliomas. Acta Psychiatr Neurol 1958;120(suppl 1): Escueta AV, Kunze U, Waddell G, Boxley J, Nadel A. Lapse of consciousness and automatisms in temporal lobe epilepsy: a videotape analysis. Neurology 1977;27: King DW, Ajmone-Marsan C. Clinical features and ictal patterns in epileptic patients with EEG temporal lobe foci. Ann Neurol 1977;2: Escueta AV, Enrile BF, Treiman DM. Complex partial seizures on closed-circuit television and EEG: a study of 691 attacks in 79 patients. Ann Neurol 1982;11: Theodore WH, Porter RJ, Penry JK. Complex patial seizures: Epikpsia, Vol. 39, No. 7, 1998

6 726 W. J. MARKS, JR. AND K. D. LAXER clinical characteristics and differential diagnosis. Neurology 1983;33: Quesney LF. Clinical and EEG features of complex partial seizures of temporal lobe origin. Epilepsiu 1986;27(suppl 2):S Chee MWL, Kotagal P, Van Ness PC, Gragg L, Murphy D, Luders HO. Lateralizing signs in intractable partial epilepsy: blinded multiple-observer analysis. Neurology 1993;43: Anand I, Kotagal P, Hammel J, Mascha E. Seizure semiology of lateral vs mesial temporal lobe epilepsy using statistical analysis. Neurology 1997;48:A Rausch R, Silfvenius H, Wieser H-G, Dodrill CB, Meador KJ, Jones-Gotman M. Intraarterial amobarbital procedures. In: Engel J Jr, ed. Surgical treatment of the epilepsies, 2nd ed. New York: Raven Press, 1993: Epilepsia. Vol. 39, No. 7, 1998