Value of non-invasive testing when there are independent bitemporal seizures in the scalp EEG

Epilepsy Research 68 (2006) 115–122

Value of non-invasive testing when there are independent bitemporal seizures in the scalp EEG Sigmund Jenssen a,∗ , Joyce Liporace b , Maromi Nei b , Michael J. O’Connor c , Michael R. Sperling b a

Department of Neurology, Drexel Medical College, Hahnemann University Hospital, Mail Stop 423, Broad and Vine Streets, Philadelphia, PA 19129, United States b Department of Neurology, Thomas Jefferson University, Philadelphia, PA, United States c Department of Neurosurgery, Thomas Jefferson University, Philadelphia, PA, United States Received 7 March 2005; received in revised form 15 September 2005; accepted 19 October 2005 Available online 28 December 2005

Abstract We investigated the value of non-invasive data for predicting the outcome of intracranial EEG and anterior temporal lobectomy (ATL) (follow-up > 1 year) in patients who have bitemporal independent seizures in the scalp EEG. No previous report has dealt with this patient group. Independent variables were duration of epilepsy, febrile seizures, interictal and ictal scalp EEG, ictal behavior, MRI, [18F]-fluorodeoxyglucose–PET (PET) and Wada test and dependent variables were surgical outcome (seizure free or not) and localized on intracranial EEG (finding all symptomatic seizures from one temporal lobe). Non-parametric statistics were used. Of 24 patients, 20 patients had IEEG, of which 12 were localized and 8 were not. Sixteen patients had ATL and, of these, 13 (81%) became seizure free and the remaining three improved. Lateralized findings on MRI and PET, a history of febrile convulsions and shorter duration of epilepsy were all associated with a focal onset on intracranial EEG, while there was a non-significant trend with ictal behavior. The non-invasive data did not predict surgical outcome. We conclude that some of these patients can do well with surgery. In most cases, intracranial EEG is necessary for localization of seizure focus, but if PET and MRI show focal abnormalities and there is a history of febrile convulsions no further evaluation could be needed. These findings need confirmation. © 2005 Elsevier B.V. All rights reserved. Keywords: Epilepsy; EEG; Temporal lobe; Bilateral; Surgical treatment

1. Introduction ∗

Corresponding author. Tel.: +1 215 7627037; fax: +1 215 7628613. E-mail address: [email protected] (S. Jenssen). 0920-1211/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.eplepsyres.2005.10.007

Temporal lobe epilepsy is a common type of intractable localization related epilepsy that is amendable to surgery (Engel et al., 1993). Surgery abolishes

116

S. Jenssen et al. / Epilepsy Research 68 (2006) 115–122

seizures in most patients, who usually have mesial temporal sclerosis as their pathological substrates (Bengzon et al., 1968). Video–EEG is routinely performed as part of the presurgical evaluation process to record seizures. The ictal scalp EEG usually has lateralized features, with seizures emanating from one temporal lobe (Risinger et al., 1989). On occasion, however, seizures recorded in the scalp EEG arise independently from both temporal lobes. Independent temporal lobe seizures in the scalp EEG pose a conundrum. Do they reflect independent bitemporal epileptogenicity and thereby serve as a warning that surgery is unwise, or can a unilateral anterior temporal lobectomy (ATL) be performed none-the-less, with a reasonable chance of success? The literature to date contains little data to answer this question. This paper reviews our experience in the surgical evaluation and treatment of patients with independent bitemporal seizures recorded in the scalp EEG. We assessed which clinical features correlated with finding unilateral temporal seizure onset in chronic intracranial EEG recordings and predicted surgical outcome. The goal of this project was to establish guidelines for the surgical treatment of patients with independent bitemporal seizures in the scalp EEG.

2. Methods 2.1. Subjects A retrospective analysis was performed. Patients were drawn from a clinical database of individuals treated surgically for refractory epilepsy. They were eligible if they had independent bitemporal seizures on scalp EEG, if the seizures were electrographically and clinically compatible with a temporal lobe onset and if they had undergone intracranial EEG recordings or ATL with a post-surgical follow-up of more than 1 year. The history and examinations were registered for each patient. 2.2. Independent variables 2.2.1. Video and scalp/sphenoidal EEG Scalp/sphenoidal EEG was performed using the international 10–20 electrode placement system and sphenoidal electrodes, with 21-channel playback sys-

tem. Methods of video and EEG recording have been previously described (Sirven et al., 1997). Intercital spikes and sharp waves were defined as temporal when maximal in the sphenoidal, frontotemporal (F7, F8) or midtemporal (T3, T4) contacts. Interictal spikes were considered unilateral if more than 90% appeared from one temporal lobe and otherwise as bilateral and according to appearance during sleep only or also during wakefulness. We defined independent bitemporal seizures as seizures in the scalp/sphenoidal EEG arising independently from right and left temporal lobes on EEG. The clinical behavior during seizures had to be compatible with a temporal lobe origin (Escueta et al., 1982). Seizures were classified as simple partial seizures (SPS), complex partial seizures (CPS) and secondarily generalized tonic–clonic seizures (SGTCS) based on the behavioral testing and analysis of EEG. Ictal behavior was examined in terms of type of aura, motor manifestations, presence or absence of speech or word comprehension during the initial part of the seizure and tendency for secondary generalization. 2.3. MRI, PET, Wada test Protocols for obtaining and analyzing brain MRI and [18F]-fluorodesoxyglucose–PET (PET) have been described previously (Sperling et al., 1994; Manno et al., 1994). All MRI scans were performed using a 1.5 T magnet. They were analyzed visually for findings indicating mesial temporal sclerosis (hippocampal atrophy on T1 sequences and increased signal of the hippocampus on T2 and FLAIR) in which case they were deemed positive (Sirven et al., 1997). Previous reports have indicated that bilateral temporal lobe volume decreased on MRI does not affect surgical outcome when compared to bilateral normal temporal volumes (Jack et al., 1995); only a lateralizing criteria was therefore used. PET scans were analyzed visually for unilateral or asymmetric hypometabolism (Manno et al., 1994). Bilateral symmetric changes on MRI and PET were classified as non-lateralized or negative. The Wada test protocol has been described previously (Sperling et al., 1994). Results were defined as lateralized if a difference of 22% or more in recognition memory score was found between the sides (≥2 of 9 items). As a measure of total recognition memory, the total score for both sides was calculated.

S. Jenssen et al. / Epilepsy Research 68 (2006) 115–122

2.4. Dependent variables 2.4.1. Surgical outcome The above non-invasive tests were compared to the outcome of the ATL and the results of the intracranial EEG recordings. Seizure outcome was categorized in four classes as previously reported (Engel et al., 1993), but for purposes of data analysis seizure outcome was seizure free (with or without auras) or not seizure free. 2.4.2. Intracranial EEG Only symptomatic seizures in the intracranial EEG were counted. The methods of intracranial EEG have been previously described (Sirven et al., 1997). All patients had medial temporal depth electrodes, lateral temporal subdural electrodes and variable extratemporal coverage. We defined the seizures as localized if all seizures emanated from one temporal lobe and non-localized if any seizures originated from the contralateral temporal lobe, from both temporal lobes simultaneously, from an extra-temporal focus or if the site of origin could not be determined. 2.4.3. Statistical methods The non-invasive test results were correlated with surgical outcome and with intracranial IEEG results. To test our exploratory hypothesis, a set of univariate analysis were done given our small sample size. We used the χ2 -test for categorical variables and the t-test for continuous variables. We calculated the positive predictive value (PPV) and the likelihood ratio (LR) using a confidence interval (CI) of 95% (SSPS, Chicago, IL).

117

in any of the patients. Individual patient data is in Table 1. 3.2. Independent variables 3.2.1. Video and scalp EEG Scalp EEG showed interictal spikes in 23 patients (96%), in 16 during wakefulness and sleep and in 7 during sleep only. Bilateral interictal spikes were recorded in 18 patients. Five patients had unilateral interictal spikes. Independent bilateral CPS or SGTCS were seen in 19 patients (79%) and five patients (21%) had unilateral CPS and SGTCS and contralateral SPS. Video analysis of the scalp EEG recordings showed that 8 patients had different behavior during seizures from the two different temporal lobes while 16 patients did not. None of the patients had symptoms or signs of extra-temporal seizure onset. 3.2.2. MRI, PET, Wada test All 24 patients had MRI, of which 12 (50%) showed a lateralized temporal lobe abnormality. In addition to right hippocampal atrophy, one patient had a rightsided parietal lesion (Patient 15, Tables 1 and 3). Nine patients had PET, of which five showed lateralized temporal lobe hypometabolism. Twenty-one patients had Wada test with bilateral injections and all but one showed left hemisphere dominance for language. Thirteen of the Wada tests (62%) showed a lateralized memory deficit while the rest did not. Mean total score on the Wada test was 10.3 (range 2.0–16.0), out of possible maximum of 18. 3.3. Dependent variables

3. Results 3.1. Demographic data A total of 24 patients (8 males, 16 females) with independent bitemporal seizures in the scalp EEG had undergone either ATL or intracranial EEG investigation. Median age at surgery was 33 year (range 22–46 years). Mean duration of epilepsy was 22 years (range 8–38 years). Five patients (21%) had a history of febrile convulsions. Mean full scale IQ was 93 (range 66–121). The neurological exam did not reveal focal findings

3.3.1. Intracranial EEG Twenty of the 24 patients (83%) had intracranial EEG while the remaining 4 did not. Twelve of the intracranial EEG investigations (60%) showed a localized temporal lobe onset and eight (40%) had non-localized ictal onset, and so the likelihood ratio (LR) for having a localized intracranial EEG was 1.5. 3.3.2. Surgical outcome Of the 12 localized patients, 1 refused surgery while the remaining 11 underwent ATL. One patient had

S. Jenssen et al. / Epilepsy Research 68 (2006) 115–122

118 Table 1 Demographic data, history and exam Patient

Age (years)

Sex

Handedness

Full scale IQ

Epilepsy onset (years)

Epilepsy duration (years)

Febrile convulsions

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

33 29 22 22 32 24 41 30 42 47 25 32 50 31 29 30 42 46 24 36 34 27 23 33

M M F F M F F F F F M F F F M F M F F M F F F F

R R R R R R R R R R R R R R R R A R R R R R A

96 121 66 115 95 88 77 89 89 116 82 87 84 94 88 84 85 94 82 121 95 75 110

9 16 3 9 1 11 12 36 20 14 17 12 38 8 23 5 1 5 0.5 22 21 3 12 5

20 13 18 20 22 13 13 15 22 33 8 20 12 23 19 28 38 22 37 16 15 33 38 28

N N N N Y N Y N N Y N N N Y N Y N N N N N N N N

M: male; F: female; R: right; A: ambidextrous; Y: yes; N: no.

in addition to ATL, parietal lobe resection (Patient 15, Table 1). One patient had additional left temporal cortical resection 1 year after the first ATL due to seizure recurrence (Patient 12, Table 1). Of the eight patients with non-localized intracranial EEG, seven were rejected for surgery while one had ATL, because of unilateral hippocampal atrophy on MRI (Patient 17, Table 1). Four patients had ATL without prior intracranial EEG. See Table 2 for details. Of the total of 16 patients who had anterior temporal lobectomy, the surgical outcome was as follows: 13 (81.2%) became seizure free, 1 (6.3%) had rare postoperative seizures (Class 2) and 2 (12.5%) had more than 80% improvement (Class 3). Mean post-surgical follow-up was 7.0 years (range 1.4–15 years). Pathological diagnosis of excised temporal lobe was mesial temporal sclerosis in 10 of the cases (63%), while 4 (25%) had gliosis and 2 (12%) were reported as normal. The additional ipsilateral parietal lobe tissue resected in one of the patients (Patient 15, Table 1) showed changes compatible with contusion.

3.4. Statistical analysis 3.4.1. Correlation with intracranial EEG We correlated the data from the epilepsy history, scalp video–EEG recordings, MRI, PET and Wada test with the results from intracranial EEG. MRI, PET and history of febrile convulsions were associated with having focal seizure onset on intracranial EEG. In brief, febrile convulsions and lateralized abnormalities on MRI and PET were associated with focal ictal onset on intracranial EEG; see Table 3. All five patients with a history of febrile seizures had localized intracranial EEG. MRI always correlated with the intracranial EEG lateralization when this was localized, although a lateralized MRI alone did not predict a localized intracranial EEG. A shorter seizure history was associated with having a localized intracranial EEG (p < 0.05). Age group with later negative findings was not significantly different in the localized group as compared to the nonlocalized.

Table 2 Test results and surgical outcome Patient IIS asleep

IIS awake

Scalp EEG Right seizures

Left seizures

Single ictal behavior

MRI PET WADA Total Wada score

+ + + + + + + − + + − + + + − − + + − −

R R L – – L – – – – R L L L R – R R – – – – – –

Intracranial ATL Outcome Pathology Follow-up EEG Class (years)

SPS CPS/GTCS SPS CPS/GTCS R B/L

R B/L

R B/L L B/L B/L B/L B/L B/L B/L B/L B/L B/L B/L B/L L B/L B/L B/L R B/L B/L

R L

1

B/L R B/L B/L B/L B/L B/L

B/L B/L

1

1 1

3 1

L 5 R 2 B/L

5 5 4 2 1 1 1 2 5 1 2 1 3 1 1 2 1 2 1 5 1 3 2

2 1 4 1

1

3 1

2

3 10 4 1 2 1 1 6 6 3 1 2 5 4 1 3 2 1 1 2

L L L R – –

L

–

R R – R L L –

11.0 8.0 4.5 14.0 12.5 16.0 7.5

R – – – L – – R – L

7.0 10.5 12.0 15.0 9.5 10.5 10.5 8.5 9.0 12.0

– R

15.5 10.0

R R

10.5 2.0

L L L R R R R L L L R L – – – – – – – –

R R L R L L L R R R R La L L Rb

1 2 1 1 1 1 1 1 1 1 1 1 3 1 3

MTS NL MTS MTS GLI NL GLI MTS MTS MTS MTS MTS GLI MTS GLI

2.2 6.6 2.8 8.4 4.3 12.7 3.8 4.7 6.8 9.7 1.4 15.4 11 10.8 4.9

R

1

MTS

6.2

S. Jenssen et al. / Epilepsy Research 68 (2006) 115–122

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

IIS: interictal spikes; R: right; L: left; B/L: bilateral; SPS: simple partial seizure; CPS: complex partial seizure; GTCS: secondarily generalized tonic–clonic seizure; single ictal behavior: see text; MTS: mesial temporal sclerosis; GLI: gliosis; NL: normal. a Patient 12 had additional left cortical resection 1 year later. b Patient 15 had additional simultaneous right parietal resection.

119

120

S. Jenssen et al. / Epilepsy Research 68 (2006) 115–122

Table 3 Non-invasive predictors of localized onset on intracranial EEG Clinical features

Positive predictive value (%)

Sensitivity (%)

Specificity (%)

Likelihood ratio

MRI of PET positive Febrile seizures and MRI positive

88 (72–100) 91 (78–100)

64 (41–86) 83 (67–100)

86 (70–100) 88 (73–100)

4.45 6.67

Fifty percent of the non-localized group had different ictal behavior from one seizure to the other while only 13% of the localized group; this difference was not significant (p < 0.20), but the sample was small. There was no association between the result on intracranial IEEG and the distribution of temporal spikes or different seizure types on scalp EEG, nor with lateralized Wada test result or the total memory score on the Wada test.

ficient statistical power to detect surgical prognostic factors. The selection of good candidates had already taken place. Secondly, finding an association between febrile convulsions, MRI, PET and having a localized ictal onset in the intracranial EEG makes it probable that they are also independent predictors of good surgical outcome.

3.4.2. Correlation with surgical outcome We then correlated epilepsy history, scalp video– EEG, MRI, PET and Wada test with the surgical outcome and found no association between the noninvasive data and seizure free outcome. In all cases when MRI, PET or Wada test were lateralized, they were abnormal on the side that was resected.

This is the second report that deals exclusively with independent bitemporal seizures in scalp–sphenoidal EEG. A previous study of 42 patients found that surgical outcome correlated with lateralized MRI, preponderance of lateralized interictal spikes and lateralized memory deficit. The authors found poor outcome when none of these non-invasive features were concordant with side of ATL, independent of results from intracranial EEG (Holmes et al., 2003). Other studies of so-called bitemporal epilepsy have dealt primarily with bilateral interictal abnormalities in scalp EEG or independent bitemporal seizures recorded with intracranial EEG. These previous studies support the impression that surgery is feasible and that non-invasive test results affect the likelihood of success. A favorable surgical outcome was found in patients with bitemporal interictal spikes in the scalp EEG more often when there were lateralized MRI or PET abnormalities (So et al., 1989; Benbadis et al., 1995; Holmes et al., 1997). A good surgical outcome for some patients with independent temporal lobe seizures during invasive monitoring was also found (Cendes et al., 1996; Sirven et al., 1997), generally related to having lateralized MRI or Wada test results. These results confirm a high specificity for MRI in temporal lobe epilepsy, whether the interictal or ictal EEG is unilateral or bilateral. Some of these studies of patients with bitemporal interictal abnormalities on scalp EEG or bilateral ictal onsets on invasive EEG used visual inspection as the method to analyze the MRI with an emphasis on asymmetry (Hirsch et al., 1991;

4. Discussion 4.1. Main findings We found that patients with independent bitemporal seizures in the scalp–sphenoidal EEG are good candidates for anterior temporal lobectomy. Of the 24 patients, most were offered surgery and all who had resection improved, with most patients becoming seizure free. The presence of unilateral temporal lobe structural or functional deficit, febrile convulsions and shorter seizure history correlated with localized unilateral temporal seizure onset on intracranial EEG. A single stereotyped ictal behavior was more frequently found in patients with localized intracranial EEG, but this association was not statistically significant. Although the analysis shows no association between the non-invasive test results and the seizure outcome when surgery was offered, an association is likely to exist for two reasons. First, since only 16 patients received surgical therapy and 12 became seizure free, lack of specificity is not surprising. There was insuf-

4.2. Comparison to previous reports

S. Jenssen et al. / Epilepsy Research 68 (2006) 115–122

Sperling et al., 1992; Sirven et al., 1997) while another used temporal lobe volume measurements (Cendes et al., 1996). These sensitivity and specificity of these two methods of MRI analysis for hippocampal atrophy have not been compared in this patient group. Febrile convulsions correlated with having a localized intracranial EEG in these patients with independent bitemporal seizures in the scalp EEG. It has previously been reported that in patients with bilateral temporal interictal spikes in scalp EEG, a history of febrile seizures was associated with unilateral ictal onset in intracranial EEG (Hirsch et al., 1991; So et al., 1989). Also an association between febrile convulsions and favorable surgical outcome has been noted (AbouKhalil et al., 1993). The reason here is probably that a history of febrile seizures is associated with increased likelihood of unilateral mesial temporal sclerosis and temporal lobe epilepsy. We have previously found the Wada test to correlate with surgical outcome in patients with independent bitemporal seizures in intracranial EEG (Sirven et al., 1997) but in this study the Wada test did not correlate. No correlation was found with surgical outcome in a similar group of patients (Holmes et al., 2003). This could reflect type II error and lack of power. The finding of a longer duration of epilepsy in the patients with non-localized intracranial EEG also finds some support in previous reports. One study related a poorer outcome with longer duration of epilepsy independent of age at surgery (Sirven et al., 2000) and there are some reports of increased hippocampal atrophy with temporal lobe epilepsy of longer duration (Theodore et al., 1991). The present sample is too small to define a cut-off duration beyond which a poor outcome is expected. 4.3. How is it that these patients become seizure free? There are several explanations why patients who have independent bitemporal seizures on scalp EEG can become seizure free after unilateral ATL. In some cases, seizures really originate from one mesial temporal lobe and have variable propagation to the ipsilateral and contralateral neocortex. Hence, the scalp EEG is misleading. This was shown in many of our patients. Another reason is sampling error. Too few seizures might be recorded in the intracranial EEG as compared

121

to the scalp EEG; but this was not the case in our study since a similar number of seizures were recorded in both modalities. Another mechanism may be that while seizures originate independently in each temporal lobe one focus may be dependent on the other so that the contralateral seizure activity abates when the primary epileptogenic focus is excised. This could be analogous to Goddard’s kindling model, in which the contralateral seizure focus disappears after the primary focus has been resected (Cibula et al., 1997). Temporal lobe epilepsy is frequently a bilateral process evidenced by frequent finding of bilateral mesial temporal sclerosis in autopsy specimen (Margerison and Corsellis, 1966). Indirect evidence of one temporal lobe’s influence on the contralateral side in humans is found in reports of normalization of mesial temporal chemical spectra with MR spectroscopy after successful contralateral temporal lobe resection (Hugg et al., 1996). 4.4. Limitations This study is retrospective, has a limited number of cases, and therefore the conclusions require confirmation. Patient selection for investigation and surgery may also have been based on other less tangible features not recorded in medical records and results might be different in other hands. The relatively small patient number makes type II errors more likely, so it is possible that some preoperative tests may correlate with outcome in operated patients. Also, there is a danger of circular reasoning in assessing test results, since some of the decision making was based on the results of the non-invasive tests. Another area of possible error lies in the usual requirement of having localized intracranial EEG recording before offered surgery. It is possible that some of the patients denied surgery because of bitemporal seizures in the intracranial EEG would have benefited had surgery been offered. None-the-less, the report confirms that patients with independent bitemporal seizures in the scalp EEG can be good surgical candidates and that the non-invasive tests are predictive of outcome.

References Abou-Khalil, B., Andermann, E., Andermann, F., Olivier, A., Quesney, L.F., 1993. Temporal lobe epilepsy after prolonged febrile

122

S. Jenssen et al. / Epilepsy Research 68 (2006) 115–122

convulsions: excellent outcome after surgical treatment. Epilepsia 34 (5), 878–883. Benbadis, S.R., So, N.K., Antar, M.A., Barnett, G.H., Morris, H.H., 1995. The value of PET scan (and MRI and Wada test) in patients with bitemporal epileptiform abnormalities. Arch. Neurol. 52 (11), 1062–1068. Bengzon, A.R., Rasmussen, T.B., Gloor, P., Dussault, J., Stephens, M., 1968. Prognostic factors in the surgical treatment of temporal lobe epileptics. Neurology 18, 717–731. Cendes, F., Dubeau, F., Andermann, F., et al., 1996. Significance of mesial temporal atrophy in relation to intracranial ictal and interictal stereo EEG abnormalities. Brain 119, 1317–1326. Cibula, Jean E., Gilmore, Robin L., 1997. Secondary epileptogenesis in humans. J. Clin. Neurophysiol. 14 (2), 111–127. Engel Jr., J., Van Ness, P.C., Rasmussen, T.B., 1993. Outcome with respect to epileptic seizures. In: Engel Jr., J. (Ed.), Surgical Treatment of the Epilepsies, second ed. New Yoi Raven Press, pp. 609–621. Escueta, A.V., Bacsal, F.E., Treiman, D.M., 1982. Complex partial seizures on closed-circuit television and EEG: a study of 691 attacks in 79 patients. Ann. Neurol. 11, 292–300. Hirsch, L., Spencer, S., Spencer, D., Williamson, P., Mattson, R., 1991. Temporal lobectomy in patients with bitemporal epilepsy defined by depth electroencephalography. Ann. Neurol. 30, 347–356. Holmes, M.D., Dodrill, C.B., Ojemann, G.A., Wilensky, A.J., Ojemann, L.M., 1997. Outcome following surgery in patients with bitemporal interictal epileptiform patterns. Neurology 48 (4), 1037–1040. Holmes, M.D., Miles, A.N., Dodrill, C.B., Ojemann, G.A., Wilensky, A.J., 2003 Aug. Identifying potential surgical candidates in patients with evidence of bitemporal epilepsy. Epilepsia 44 (8), 1075–1079. Hugg, J.W., Kuzniecky, R.I., Gilliam, F.G., Morawetz, R.B., Fraught, R.E., Hetherington, H.P., 1996. Normalization of contralateral

metabolic function following temporal lobectomy demonstrated by 1H magnetic resonance spectroscopic imaging. Ann. Neurol. 40 (2), 236–239. Jack Jr., C.R., Trenerrv, M.R., Cascino, G.D., Sharbrough, F.W., So, E.L., O’Brien, P.C., 1995 Jul. Bilaterally symmetric hippocampi and surgical outcome. Neurology 45 (7), 1353–13358. Manno, E.M., Sperling, M.R., Ding, X., Jaggi, J., Alavi, A., O’Connor, M.J., Reivich, M., 1994. Predictors of outcome after anterior temporal lobectomy: positron emission tomography. Neurology 44 (12), 2331–2336. Margerison, J.H., Corsellis, J.A.N., 1966. Epilepsy in the temporal lobes. Brain 89, 499–530. Risinger, M.W., Engel Jr., J., Van Ness, P.C., Henry, T.R., Crandall, P.H., 1989. Ictal localization of temporal lobe seizures with scalp/sphenoidal recordings. Neurology 39 (10), 1288– 1293. Sirven, J.I., Malamut, B.L., Liporace, J.D., O’Connor, M.J., Sperling, M.R., 1997. Outcome after temporal lobectomy in bilateral temporal lobe epilepsy. Ann. Neurol. 42 (6), 873–878. Sirven, J.I., Malamut, B.L., O’Connor, M.J., Sperling, M.R., 2000. Temporal lobectomy outcome in older versus younger adults. Neurology 54 (11), 2166–2170. So, N., Olivier, A., Andermann, F., Gloor, P., Quesney, L.F., 1989. Results of surgical treatment in patients with bitemporal epileptiform abnormalities. Ann. Neurol. 25, 432–439. Sperling, M.R., O’Connor, M.J., Saykin, A.J., et al., 1992. A noninvasive protocol for anterior temporal lobectomy. Neurology 42, 416–422. Sperling, M.R., Saykin, A.J., Glosser, G., Moran, M., French, J.A., O’Connor, M.J., 1994. Predictors of outcome after anterior temporal lobectomy: the intracarotid amobarbital test. Neurology 44, 2325–2330. Theodore, W.H., Bhatia, S., Hatta, J., et al., 1991. Hippocampal atrophy, epilepsy duration, and febrile seizures in patients with partial seizures. Neurology 52 (1), 132–136.