Focal Abnormalities in Idiopathic Generalized Epilepsy: a Critical Review of the Literature *†Udaya Seneviratne, *Mark Cook, and *Wendyl D’Souza

CRITICAL REVIEW AND INVITED COMMENTARY

Focal abnormalities in idiopathic generalized epilepsy: A critical review of the literature *†Udaya Seneviratne, *Mark Cook, and *Wendyl D’Souza

Epilepsia, 55(8):1157–1169, 2014 doi: 10.1111/epi.12688

SUMMARY Conventionally, epilepsy is dichotomized into distinct “focal” and “generalized” cate- gories. However, many studies have reported so-called focal features among patients with idiopathic generalized epilepsy (IGE) in the domains of semiology, electroencephalography, neuropsychology, neuropathology, and neuroimaging. We sought to review such features and clinical implications. A Web of Science database search was conducted to identify relevant publications. Our search yielded 145 papers describing focal features involving different domains in IGE, with 117 papers analyzed after excluding abstracts and case reports. Focal semiologic features are commonly seen in IGE. There are conﬂicting data from studies in the domains of electroencephalography, neuroimaging, and neuropathology. Studies on neuropsychology are suggestive of Dr. Udaya frontal lobe functional deﬁcits in juvenile myoclonic epilepsy. Most advanced neuroimaging studies demonstrate the involvement of both the thalamus and the cortex dur- Seneviratne is a ing generalized spike-wave discharges (GSWDs). A few electroencephalographic and Consultant neuroimaging studies indicate that the cortex precedes the thalamus at the onset of Neurologist and GSWD. Focal features may contribute to misdiagnosis of IGE as focal epilepsy. How- Epileptologist at ever there are methodologic limitations in the studies that affect the results. Monash Medical KEY WORDS: Electroencephalographic, Epilepsy semiology, Generalized seizures, Centre and St. Partial seizures, Neuroimaging, Neuropsychology. Vincent’s Hospital, Melbourne,Australia.

Idiopathic generalized epilepsy (IGE; genetic general- tion and spread based on epileptic networks. This change is ized epilepsy) constitutes a rubric of several electroclinical reflected in a recent publication by the International Lea- syndromes based on specific clinical features and electro- gue Against Epilepsy (ILAE) on revised terminology of encephalography (EEG) abnormalities.1,2 The traditional seizures and epilepsies.2 This paper conceptualizes focal approach of classifying epilepsies into “focal” and “gener- seizures as those arising from networks confined to a single alized” implies there can be no “focal” features in “gener- cerebral hemisphere, whereas generalized seizures “origi- alized” epilepsies. However, there has been a gradual nate at some point within, and rapidly engage bilaterally paradigm shift in the conceptualization of seizure genera- distributed networks” including cortical and subcortical structures, leaving open the possibility of a focal origin Accepted May 13, 2014; Early View publication June 17, 2014. within the network. *Department of Medicine, St. Vincent’s Hospital, University of Mel- Amid the ongoing debate and discussion about the patho- bourne, Melbourne, Victoria, Australia; and †Department of Neuroscience, genesis and classification of IGE, a body of literature high- Monash Medical Centre, Melbourne, Victoria, Australia Address correspondence to Udaya Seneviratne, Department of Neuro- lighting various focal features of IGE has gradually science, St. Vincent’s Hospital, PO Box 2900, Fitzroy, Melbourne, Vic. accumulated over the years. In this paper, we review the lit- 3065, Australia. E-mails: [email protected]; wusenevi@ erature on focal characteristics of IGE in the domains of optusnet.com.au semiology, EEG, neuropsychology, neuropathology, and Wiley Periodicals, Inc. neuroimaging. © 2014 International League Against Epilepsy

1157 1158 U. Seneviratne et al.

Methods questioning. It is possible that auras in IGE are under- reported, as patients may not volunteer this information We searched the Web of Science database using search unless questioned in a systematic manner. terms “primary generaliz(s)ed epilepsy,”“idiopathic gener- It is unclear what auras represent in IGE, particularly in aliz(s)ed epilepsy,”“absence epilepsy,”“juvenile myoclonic the absence of correlating video-EEG data. Auras may not epilepsy,”“absence seizure,”“myoclonic seizure,” and “gen- necessarily be due to focal seizure activity. It is possible that eraliz(s)ed tonic–clonic seizure” in combination with “focal,” preictal prodromal symptoms may have been reported as “aura,”“semiology,”“EEG,”“neuropsychology,”“magnetic ictal “auras” by patients in these studies. resonance imaging,”“volumetry,”“voxel-based morphometry,”“functional magnetic resonance imaging,”“magnetic Generalized tonic–clonic seizures resonance spectroscopy,”“pathology,” and “positron emis- In classic generalized tonic–clonic seizures (GTCS) of sion tomography.” Publications in all years were taken into IGE, the tonic–clonic activity is expected to be symmetrical consideration. We also searched the reference lists of identi- and generalized from the onset. Numerous focal features fied publications for more papers relevant to the topic. Only have been described in secondarily GTCS (focal seizures original papers published in English were included in this evolving to bilateral convulsive seizures according to the 2 review. Single case reports and abstracts from conferences revised terminology ) of focal epilepsies, helping lateraliza- were excluded. tion and localization of the seizure focus. Several authors have described similar focal features in GTCS of IGE as Results and Discussion well. These include forced head version, eye version, “figure-4-sign” (contralateral elbow extension and ipsilateral Our literature search yielded a total of 145 relevant publi- flexion), focal clonic activity, asymmetry in tonic and clonic cations describing focal features of IGE in different phases, hemiconvulsion, fencing posture, unilateral tonic/ domains. Having excluded 16 abstracts and 12 case reports, dystonic posturing, postictal nose wiping, postictal hemipa- 117 publications were analyzed. resis, and asymmetric seizure termination.8–16 The frequency of such focal features ranges from 35% to 46% among three studies involving 20, 55, and 26 patients based Semiology on video-EEG monitoring (VEM).12,14,15 However, it Aura should be noted that these studies were reported from ter- An aura is a subjective symptom representing a seizure or tiary referral centers, which were likely to receive more part of a seizure that can be recalled by the individual. The atypical and complex cases of IGE. Therefore, the possibil- 1981 ILAE seizure classification describes the aura as a ity of overestimating the frequency due to selection bias simple partial seizure or the signal symptom of a complex should be taken into consideration. partial seizure.3 In the revised terminology, aura is described as a focal seizure “without impairment of con- Myoclonic seizures sciousness or awareness involving subjective sensory or Myoclonic jerks of JME are characterized by bilateral, psychic phenomena only.”2 This implies auras are experi- arrhythmic, irregular jerks of proximal and distal muscles enced only in focal epilepsies. predominantly in the arms.1 However, focal (unilateral) and Our search yielded four papers on auras experienced by asymmetric jerks in JME have been reported. In a detailed patients with IGE. In a prospective study involving 19 video-polygraphic analysis of myoclonic seizures in JME, patients with IGE from an epilepsy monitoring unit (EMU), four of five patients had asymmetric myoclonic jerks.17 70% reported auras.4 This relatively high frequency could Three other studies have reported asymmetric myoclonus in be an overestimate due to selection bias, as IGE patients 61%,14 16.8%,18 and 14%19 of JME patients. Focal myo- with auras are more likely to be referred to EMU with a pro- clonic seizures were detected in 25% of patients with JME visional diagnosis of focal epilepsy. Another study found who underwent VEM.15 Study heterogeneity and sampling special sensory, psychic, and autonomic auras among 54% bias would account for this wide range. of 37 patients diagnosed with juvenile myoclonic epilepsy (JME).5 A questionnaire was used to obtain information on Absence seizures auras. However, auras were reported during a structured Automatisms are generally considered to be an integral interview by only 24% (n = 379) with “generalized epi- component of complex partial seizures (focal seizures with lepsy” from three population-based twin registries.6 Visual impairment of consciousness or awareness2). However, oral auras were experienced by 4 (10%) of 40 patients diagnosed and manual automatisms have been well recognized in asso- with JME.7 Phenomenologically, auras of temporal lobe ciation with typical absence seizures. In a cohort of 70 epilepsy and IGE appeared to be different, except for rising untreated children with absence seizures, automatisms were epigastric sensation, which was reported by both groups.4 detected in 40% of seizures and 76% of patients.20 Automa- In the above studies, auras were elicited by systematic tisms were more likely to occur during hyperventilation and

Epilepsia, 55(8):1157–1169, 2014 doi: 10.1111/epi.12688 1159 Focal Features in IGE in longer absence seizures.20 The origin and underpinning function in JME with FLE, temporal lobe epilepsy (TLE), mechanisms of automatisms are complex, and their pres- and healthy controls. The JME group had significant impair- ence does not necessarily indicate a focal basis in IGE. Yet, ment in executive functions compared to TLE and healthy it is important to be aware of this phenomenon to avoid mis- controls. However, there was no difference between JME diagnosis of absences as complex partial seizures. and FLE.43 Diagnostic criteria of IGE is a confounding factor, as Prospective memory is closely linked to executive func- inadvertently mixed focal epilepsy patients will affect the tions, and one study reported significant deficits in prospec- results on semiology. The 1989 ILAE criteria1 were used in tive memory in JME patients and their unaffected siblings most studies on semiology, whereas criteria were not speci- compared to healthy controls.44 fied in six studies.4,9,12,13,15,19 Verbal and nonverbal episodic memory reflecting functions of the temporal lobe did not appear to be impaired in Electroencephalography JME in comparison to healthy controls.41,44 The existence of focal EEG features in IGE has been More recently, researchers used Iowa Gambling Task known for many decades.21 In a large cohort of IGE (IGT) to evaluate the decision making ability of patients patients, EEG focal abnormalities were found in 56% and with JME. Worse performance was recorded from patients localized to temporal regions in most.22 In a series of IGE in comparison to controls, and further analysis revealed a patients studied with video-EEG, focal interictal epilepti- trend toward poor performance with poor seizure control.45 form discharges (IEDs) and semiologic features of focal sei- Another study found similar results in JME patients with zure onset were observed in 35%. However, no seizures ongoing seizures.46 These results are suggestive of deficits with focal EEG onset were seen.12 Another study found involving cortico-subcortical prefrontal networks in JME.45 focal interictal EEG abnormalities in 2 of 26 patients with Targeted neuropsychological studies in other IGE syn- JME.15 Other studies have reported EEG focalities in dromes are scarce. One study compared intellectual capac- 30–55% of JME patients.23–25 The reported frequency of ity, verbal and nonverbal memory, language, and executive focal or lateralized EEG abnormalities in patients with functions between benign childhood epilepsy with centro- absence seizures ranges from 16% to 37%.21,26–30 Temporal temporal spikes (BECTS), CAE, and healthy controls. Both intermittent rhythmic delta activity was reported among BECTS and CAE patients were in remission off antiepilep- 13% of patients with juvenile absence epilepsy (JAE) com- tic drugs. Executive functions were evaluated with trail- pared to none in JME from a single study based on video- marking, as well as Stroop and Wisconsin card sorting tests. EEG.31 Focal EEG abnormalities appear to be state depen- No significant differences were found between the three dent, and Koutroumanidis et al.32 found focal discharges in groups in the domains of memory, language, and executive children with childhood absence epilepsy (CAE) mainly functions.47 during non–rapid eye movement sleep. Based on these stud- Another study compared neuropsychological functions ies, it appears that focal interictal EEG abnormalities are between children with absence epilepsy and healthy con- found among one third of patients with IGE. trols. Patients who had experienced seizures other than However, it should be noted that patient populations of absences were excluded. Hence the patient group was likely these studies were heterogeneous and the influence of to consist of a mix of CAE and JAE. Compared to the con- potential confounding factors such as age, arousal state, trols, the patient group had significantly lower general cog- length of recording, and AED therapy were not specified. nition, visuospatial skills, nonverbal memory, and delayed Most studies employed diagnostic criteria consistent with recall.48 1989 ILAE classification, although in the minority it was Comparisons are difficult due to methodologic variations, not specified.12,15,19,26,27 Another potential pitfall is label- particularly with respect to cohort differences in age of ing bifrontal fragmented spike-wave discharges as focal onset, method of ascertainment (incident vs. prevalent, IEDs. It should be noted that among the studies reviewed community vs. hospital, prospective vs. retrospective), only one provides a strict definition of focal IEDs.15 duration of disease, treatment status, seizure control, lack of population control, and testing paradigms. As in other Neuropsychology domains, diagnostic criteria employed have an influence on Working memory, prospective memory, and executive the results. Among the 15 studies discussed, only 9 have function are considered to be frontal lobe functions.33 Sev- defined diagnostic features consistent with 1989 ILAE crite- eral neuropsychological studies have revealed functional ria.36–40,43,45–47 In general, there appears to be a tendency deficits involving frontal lobes in JME. On neuropsycholog- for frontal lobe functional deficits in JME on neuropsycho- ical testing, deficits in visual working memory were found logical testing. in both JME and frontal lobe epilepsy (FLE) compared to healthy controls.34 Eight studies have reported deficits in Magnetic resonance imaging executive functions in JME in comparison to healthy By convention, magnetic resonance imaging (MRI) is not controls.35–42 Another study compared frontal cognitive expected to reveal abnormalities in IGE. More convincing

Epilepsia, 55(8):1157–1169, 2014 doi: 10.1111/epi.12688 1160 U. Seneviratne et al. evidence on focal features in IGE has emerged from studies age, gender) and the size of control group.64 Artifacts can using advanced neuroimaging techniques. affect comparisons, and gray matter volumes may vary across different scanners.65 Furthermore, reporting bias in Magnetic resonance imaging volumetry the literature due to selective analysis, and reporting and Studies on volumetric analysis in IGE have yielded publication of positive results should also be taken into con- mixed results. The most commonly reported abnormalities sideration when interpreting the studies on volumetry. This are in the thalamus. Two studies found reduced thalamic bias has been well demonstrated in the studies on brain vol- volumes,40,49 whereas no changes were detected in three ume abnormalities in mental health conditions.66 With the studies.50–52 Both increased and decreased thalamic vol- exception of three studies,41,58,63 1989 ILAE diagnostic cri- umes were found in the cohort of another study.53 The same teria were used. group reported increased volume in a larger cohort.54 Both increased55 and reduced56 lobar volumes have also been Magnetic resonance spectroscopy reported. Volumetry technique is operator dependent, which Magnetic resonance spectroscopy (MRS) is a noninva- can lead to variable results in different studies. With the sive technique of in vivo quantification of brain metabolites. exception of one study,52 all have used 1989 ILAE criteria. Several studies have been carried out in IGE looking at glutamate plus glutamine (GLX), N-acetyl aspartate (NAA), Voxel-based morphometry creatine phosphocreatine (Cr), Choline (Cho), and c-amin- Voxel-based morphometry (VBM) is an automated obutyric acid. Regional changes of these compounds reflect technique of quantitative MRI analysis. It is widely used in differences in local metabolism, cellular integrity, neuronal epilepsy, and our search yielded 13 studies in IGE function, and excitability.67 (Table 1).38,41,49,50,53,55,57–63 Both regional cortical gray From 13 published studies on MRS in IGE, the most con- matter and thalamic volume changes have been reported in sistent findings are reduced NAA in the thalamus and the those studies. The most consistent finding is reduction is frontal lobe. There are conflicting results on GLX in the thalamic volumes seen across the majority of studies. Both frontal lobe, with studies showing both increased and increased and decreased frontal cortical gray matter reduced concentrations (Table 2).59,68–79 volumes have been found (Table 1), whereas one study MRS results are influenced by several variables such as did not find any difference between JME patients and differences in data acquisition, post- processing techniques, controls.41 and age and duration of epilepsy. The diagnostic criteria of These discrepancies could be due to several factors IGE are not defined in five studies,70–72,78,79 but other stud- including heterogeneity of study populations and methodol- ies have used 1989 ILAE criteria. These confounders should ogy. Voxel-based morphometry findings in particular are be taken into consideration in the interpretation and compar- affected by covariates (such as total intracranial volume, ison of results.

Table 1. Voxel-based morphometry studies in IGE References Age range (mean) year Cohort Key ﬁndings in patients 38 (33.61) JME 28, Co 55 Reduced GM in supplementary motor area and posterior cingulate cortex 41 15–45 (24.2) JME 19, Co 20 No difference in GM volume 49 (33.6 Æ 8.6) GTCSO 19, Co 52 Reduced GM in frontal, parietal, temporal cortex, thalami and cerebellum 50 20–50 (30) JME 21, Co 20 Reduced volume anterior thalamus 53 9–60 (30) JME 7, AE 11, GTCSO 4 Focal GM abnormalities (frontal, parietal, temporal, occipital) in 77% 55 9–60 (30 Æ 11) JME 19, Co 19 Reduced GM in prefrontal lobe 57 JME (32 Æ 9) JME 44, AE 24, Increased GM in frontobasal (JME), superior mesiofrontal (AE), AE (27 Æ 12) GTCSO 15, Co 47 no difference (GTCSO) GTCSO (12 Æ 11) 58 (17 Æ 8) CAE 13, Co 109 Reduced GM in thalami, subcallosal gyri of frontal lobe Reduced WM in basal frontal lobe, anterior and posterior cingulate 59 (32) IGE 43, Co 38 Reduced volume in thalami 60 16–35 (22.7 Æ 5.1) JME 25, Co 44 Increased GM mesiofrontal lobe, reduced GM thalami 61 14–55 (26.6) JME 60, Co 30 Reduced GM thalami, insula, cerebellum Increased GM superior frontal, orbital frontal, medial frontal gyri 62 (16) CAE 44, Co 257 Reduced GM in thalami 63 15–37 (25) JME 20, Co 30 Increased cortical GM in mesial frontal lobes 41 15–45 (24.2) JME 19, Co 20 No difference in GM volume

AE, absence epilepsy; CAE, childhood absence epilepsy; Co, controls; GM, gray matter; GTCSO, generalized tonic–clonic seizures only; IGE, idiopathic generalized epilepsy; JME, juvenile myoclonic epilepsy; WM, white matter.

Epilepsia, 55(8):1157–1169, 2014 doi: 10.1111/epi.12688 1161 Focal Features in IGE

Table 2. Studies on magnetic resonance spectroscopy in IGE Age range References (mean) years Cohort On AED Key ﬁndings in patients 59 (32) JME 23, GTCSO 20, Co 38 Yes Increased GLX and reduced NAA in thalamus 68 20–52 (37) JME 12, GTCSO 8, Co 11 NS Reduced NAA/Cr in thalamus 69 (27.2) JME 57, Co 30 Yes Reduced NAA/Cr in frontal lobe; increased GLX/Cr in insula and striatum 70 18–39 (28) IGE 18, Co 25 Yes Reduced NAA, GLX, CHO and MI in frontal lobe, reduced NAA in thalamus 71 25–43 (32.3) AE 9, Co 9 Yes Reduced NAA/Cr in thalamus 72 16–30 (20.3) JME 15, Co 16 Yes Reduced NAA/Cr in thalamus 73 11–19 (14.9) JAE 14, Co 10 Yes Reduced NAA/Cr in thalamus 74 (26.6) JME 60, Co 30 Yes Reduced NAA/Cr in thalamus, motor cortex and medial prefrontal cortex; reduced GLX/Cr in motor cortex, medial prefrontal cortex and posterior cingulate gyrus 75 17–44 (31.6) JME 10, Co 10 NS Reduced NAA/Cr in thalamus 76 26–42 JME 15, Co 10 Yes Reduced NAA in prefrontal cortex 77 18–51 JME 25, GTCSO 20, Co 10 Yes (43), no (2) Reduced NAA in frontal lobe (JME); reduced NAA in thalamus (GTCSO); reduced CHO and MI (JME and GTCSO) 78 17–55 (27) IGE 15, OLE 15, Co 15 Yes Increased GLX and GABA in occipital lobe (both IGE and OLE) 79 NS IGE 21, Co 17 Yes Increased GLX and reduced NAA in frontal lobe

AE, absence epilepsy; AEDs, antiepileptic drugs; CHO, choline; Co, controls; Cr, creatine; GABA, c-aminobutyric acid; GLX, glutamate plus glutamine; GTCSO, generalized tonic–clonic seizures only; IGE, idiopathic generalized epilepsy; JAE, juvenile absence epilepsy; JME, juvenile myoclonic epilepsy; MI, myo-inositol; NAA, N-acetyl aspartate; NS, not speciﬁed; OLE, occipital lobe epilepsy.

Functional magnetic resonance imaging First, most studies have used GSWD for simultaneous Functional magnetic resonance imaging (fMRI) is a non- EEG-fMRI. Although GSWD is the electrographic hall- invasive technique measuring cerebral activity linked to mark, there are other specific EEG features in IGE. Only a cerebral blood flow based on oxygenated and deoxygenated minority of research studies address other electrographic hemoglobin ratio (blood oxygenation level–dependent markers such as polyspike-wave discharges and photopar- contrast, BOLD).67 Increased synaptic activity associated oxysmal response (Table 3). Hence, the inferences drawn with epileptiform discharges is expected to increase the from studies involving GSWDs alone may not provide a BOLD signal in the corresponding region (“activation”). complete functional picture. This principle is used in simultaneous recording of EEG and Second, the findings are affected by various methodolog- fMRI to compare BOLD signal changes with ictal and inte- ic differences such as fMRI paradigms, image processing, rictal epileptiform discharges. The reasons for decreased data acquisition (continuous vs. spike triggered EEG-fMRI, BOLD signal (“deactivation”) in EEG-fMRI are less amount and length of GSWD acquired) and statistical analy- clear. Researchers have postulated it to be a reflection of sis. interruptions to the “default state of the brain” by epileptic Finally, as in all other domains, diagnostic criteria will discharges.80 affect the sample and results. Among the studies discussed, Activation of various brain regions during cognitive tasks the majority have defined diagnostic features consistent has also been tested with fMRI in IGE patients.41 Our litera- with 1989 ILAE criteria.80–85,87,90–93 ture search yielded 16 fMRI studies in IGE.41,80–94 Despite study heterogeneity, common trends emerge from EEG- Positron emission tomography fMRI results. First, almost all studies demonstrate thalamic Positron emission tomography (PET) scans are com- activation with generalized spike-wave discharges (GSWDs). monly used in the presurgical evaluation of focal epilepsies. Second, GSWDs are also associated with activation of Only a few studies have employed this technique in IGE. certain cortical areas, mainly frontal, parietal, temporal, Two early studies demonstrated increased cerebral meta- and insular. Two studies found that cortical activation bolic rate for glucose during GSWD without any focal preceded thalamic changes.83,92 Third, there is deactivation changes.95,96 A later study evaluated fluorodeoxyglucose of certain regions recognized as “default mode areas” (FDG) uptake with visual working memory paradigms in (Table 3). nine JME patients in comparison to 14 healthy controls. The activation of certain cortical regions in EEG-fMRI Only the controls demonstrated increased uptake in frontal can be interpreted as evidence supporting the theory of focal regions during the task. The absence of uptake in JME was basis in IGE. However, these studies should be interpreted interpreted as “cortical disorganization” in the region.97 A and compared with caution due to some limitations. more recent study found increased FDG uptake in the

Epilepsia, 55(8):1157–1169, 2014 doi: 10.1111/epi.12688 1162 U. Seneviratne et al.

Table 3. Studies on functional magnetic resonance imaging in IGE Age range References (mean) year Cohort On AED Key ﬁndings in patients 41 15–45 (24.2) JME 19, Co 20 Yes 16, no 3 fMRI done with cognitive tests; no difference between JME and controls in fMRI activation with working memory paradigms 80 18–66 (35) IGE 15 Yes 14, no 1 Activation in thalamus, mesial frontal, insula and cerebellum; deactivation in anterior frontal, parietal, posterior cingulate gyri 81 18–66 (35) IGE 15 Yes 14, no 1 Multiregional activation and deactivation in 14, thalamus activation in 8, deactivation in 2, both in 2 82 16–34 CAE 4, IGEU 1 Yes 4, no 1 Deactivation in posterior cingulate, no activation in thalamus 83 6–15 (11.9) CAE 42 NS Activation in orbital/medial frontal and medial/lateral parietal cortices >5s before seizure onset; thalamus activation follows 84 6–15 CAE 9 Yes 5, no 4 Activation in thalamus, frontal, primary visual, auditory, somatosensory and motor cortices; deactivation in parietal cortex, cingulate gyrus, and basal ganglia 85 18–74 IGE 32, SGE 14 Yes 43, no 3 Activation in thalamus; deactivation in frontal, parietal, and temporal cortices 86 11–14 (10.25) EMA 4 Yes Activation in thalamus, cerebellum, mesial frontal, middle parietal, temporal and insula r cortices; deactivation in anterior frontal cortex, posterior parietal cortex, and posterior cingulate gyrus 87 NS CAE 8, JAE 1 Yes 3, no 6 Activation in thalamus (94%), frontal and parietal cortex (59%); (17 AS studied) deactivation in caudate (59%), default mode areas (88%) 88 19–58 (35.2) IGE 27, Co 30 Yes 25, no 2 Activation in thalamus, mesial frontal cortex, cerebellum; deactivation in default mode areas 89 8–22 (14) Generalized PPR 6 Yes 2, no 4 Activation in visual, parietal and premotor cortices (with PPR) 90 4–7 (9) CAE 3, MAE 1, Yes 5, no 1 Activation in thalamus; deactivation in frontal, parietal regions, JME 1, IGEU 1 and precuneus (with generalized polyspike and wave discharges) 91 4–12 (8) CAE 10 No Activation in thalamus; deactivation in parietal lobe, precuneus, caudate 92 15–55 JME 5, JAE 1, IGEU 3 Yes Activation in prefrontal and dorsolateral cortex followed by thalamus 93 (31.8) JME 12, IGE Yes Valproate-responsive and valproate-resistant groups were compared. (non-JME) 13 Activation in thalamus, frontal region (all); activation in mesial frontal cortex, paracingulate gyrus, and anterior insula (valproate-resistant patients) 94 (32.8) JME 30, Co 26 Yes 29, no 1 fMRI was done with spatial working memory paradigm. Activation in primary motor cortex and supplementary motor area; impaired deactivation in default mode network

AED, antiepileptic drugs; AS, absence seizures; CAE, childhood absence epilepsy; Co, controls; EMA, eyelid myoclonia with absences; fMRI, functional magnetic resonance imaging; IGE, idiopathic generalized epilepsy; IGEU, idiopathic generalized epilepsy unspeciﬁed; JAE, juvenile absence epilepsy; JME, juvenile myoclonic epilepsy; MAE, myoclonic astatic epilepsy; NS, not speciﬁed; PPR, photoparoxysmal response; SGE, secondary generalized epilepsy (please note only two fMRI studies, Refs 41 and 94, were done with cognitive tasks. Others were EEG-fMRI studies). thalamus during GSWD as well as interictally in JME IGE.100 However, a later autopsy study comparing the his- patients.98 No cortical abnormalities were reported in this tology of brain from people who had IGE with healthy con- study. In this study, EEG monitoring was done before, dur- trols found no significant difference in microdysgenesis ing, and after the isotope injection. In contrast, another features between the two groups.101 study did not find any difference between JME patients and healthy controls in resting (interictal) FDG-PET scans.99 Focal features in IGE: what does it mean? However, EEG monitoring was not done, and the interictal There are four possible explanations for the presence of state was decided on clinical observation, thereby limiting focal features in studies on IGE; (1) misdiagnosis of focal the validity of data. epilepsy as IGE due to poor diagnostic criteria, (2) focal fea- A major drawback in FDG-PET is its limited temporal tures as an integral component of IGE compatible with that resolution due to delayed tracer uptake.98 It is of limited diagnosis, (3) coexistence of both IGE and focal epilepsy in value in evaluating focal abnormalities in IGE. the same patient, and (4) generalized seizures in IGE having Among the studies discussed, only one has used 1989 a focal onset. ILAE diagnostic criteria in patient selection.98 Misdiagnosis of focal epilepsy as IGE Neuropathology Poor case selection with imprecise diagnostic criteria Microdysgenesis refers to specific microscopic abnor- may result in the inclusion of focal epilepsy cases in IGE malities of brain reported in people with epilepsy as well as study cohorts and one might interpret focal features on that normal subjects. Early neuropathologic studies indicated basis. However, most studies from tertiary centers have that microdysgenesis was found in most patients with included patients based on established ILAE criteria. Hence,

Epilepsia, 55(8):1157–1169, 2014 doi: 10.1111/epi.12688 1163 Focal Features in IGE this does not appear to be a major confounding factor. Yet, IGE. It would be useful to look at different hypotheses of this aspect needs attention in the interpretation of study IGE pathophysiology and relevant animal and human data results. in this context. Cellular and network mechanisms of GSWD are com- Focal features as an integral component of IGE plex. There are several seizure types and EEG markers in In general, studies indicate that the presence of focal fea- IGE. Underlying pathophysiologic mechanisms of all dif- tures is not incompatible with the diagnosis of IGE, when ferent types may not be the same. What has been mostly standard ILAE criteria are fulfilled. This is illustrated in studied is the typical GSWDs of absence seizure, the elec- Figure 1, demonstrating both generalized and focal EEG troclinical prototype of IGE. abnormalities in a 22-year-old patient with JME. The under- Several hypotheses have been proposed to explain the lying mechanism of focal abnormalities on neuropsycholog- GSWDs of absence seizures. Historically, the “centrence- ical testing and neuroimaging are unclear. Whether such phalic theory” was the first mechanism proposed, which focal abnormalities are caused by the impact of repetitive suggested origin of GSWDs from subcortical midline struc- seizure activity, effect of antiepileptic medications, or tures, in particular thalamus and brainstem.104 Buzsaki105 genetic factors remain speculative. refined this theory further by localizing the pacemaker to reticular thalamic nucleus (“thalamic clock theory”). The Coexistence of both IGE and focal epilepsy in the same use of intracarotid and intravertebral injections of amobar- patient bital sodium and pentamethylenetetrazol in patients with There are case reports and case series describing coexis- generalized seizures106 as well as the advent of feline gener- tence of proven focal epilepsy and well-characterized IGE alized penicillin epilepsy (FGPE) model107 lead to the “cor- in the same patient.102,103 However, this phenomenon ticoreticular theory” describing the genesis of GSWD in a appears to be rare, accounting for <1% of the IGE popula- thalamocortical circuitry. This theory postulated a key role tion in the studies.102,103 Given the tertiary center referral played by the cortex in a more diffuse manner rather than a bias, the true prevalence is likely to be even smaller. Hence, focal abnormality. According to “corticoreticular theory,” this phenomenon is unlikely to be responsible for all the the cortex is in a hyperexcitable state so that thalamic sleep focal features of IGE reported in the studies. spindles reaching the cortex are transformed into GSWD.108 More recently, the “cortical focus theory” was proposed to Focal onset of generalized seizures in IGE indicate the origin of GSWD in a frontal cortical Whether seizures of IGE have focal onset is a highly con- focus.109,110 Using a rat genetic model of absence epilepsy, tentious subject. It is not possible to draw any conclusions Meeren et al.110 demonstrated that a cortical focus within based on the data from studies discussed in this article. the perioral region of the somatosensory cortex lead the However, despite some inconsistencies and methodologic thalamus by a mean time of 8.1 msec during the first limitations, one common trend seen across most of the neu- 500 msec of an absence seizure. According to this hypothe- roimaging studies is the involvement of the thalamus and sis, GSWDs originate from a focus in the somatosensory frontal cortex. Further evaluation of these findings may cortex followed by rapid generalization over the cortex improve our understanding on the pathophysiologic basis of bilaterally via corticocortical networks and paroxysmal

C A B Figure 1. EEG of a patient with typical juvenile myoclonic epilepsy demonstrating both generalized and focal abnormalities. (A) Typical generalized spike-wave and polyspike-wave discharges. (B) Focal bitemporal sharp-wave discharges (X and Y) captured in sleep. (C) The magnetic resonance imaging of the brain does not reveal any focal structural abnormality. Epilepsia ILAE

Epilepsia, 55(8):1157–1169, 2014 doi: 10.1111/epi.12688 1164 U. Seneviratne et al. oscillations through thalamocortical loops. At the onset the More recently, data from advanced neuroimaging studies cortex drives the thalamus, but subsequently the coupling provide further evidence in this area. Most studies show becomes bidirectional, with both structures driving each involvement of both the thalamus and cortex in relation to other. This mechanism provides a circuitry to amplify and GSWDs, although the relative importance of each compo- sustain GSWDs.109,110 It is important to distinguish cortical nent is less clear. At least two EEG-fMRI studies have onset of GSWD in IGE from focal frontal lobe epilepsy with shown BOLD activity in the frontal cortex before thalamic rapid generalization, based on a thorough assessment activation, which could be interpreted as evidence support- including clinical details, neuroimaging, and EEG.111,112 ing the “cortical focus theory.”83,92 There has been a growing body of literature supporting However, some studies are not in favor of cortical focus the theory of GSWD origin from a cortical, in particular theory. In the rat model, some studies have shown that tha- frontal, focus. Most of the data come from experiments lamic activity precedes the cortical activation at the onset of involving animal models including FGPE and genetic rat GSWDs.128,129 Some of the early experiments on anesthe- models of absence epilepsy such as Genetic Absence Epi- tized cats demonstrated generation of 3-Hz bilateral spike- lepsy Rats from Strasbourg (GAERS) and Wistar Albino wave discharges over the anterior cortical regions with mid- Glaxo/Rijswijk (WAG/Rij). line thalamic stimulation supporting centrencephalic the- Using nonlinear association analysis, Meeren et al.110 ory.130 One depth electrode study reported capturing 3-Hz have demonstrated a focus of GSWD origin within the per- GSWDs synchronously from the scalp surface electrode and ioral region of somatosensory cortex in the WAG/Rij thalamic depth electrodes.131 model. Concordant findings have been reported from the The crucial role played by the reticular thalamic nucleus GAERS model as well.113 Several other animal studies have (RTN) has been demonstrated with lesional studies resulting supported this hypothesis. Upregulation of sodium channel in abolition of GSWDs in both the GAERS132 and the expression114 and increased excitability on electrical stimu- WAG/Rij models.133 Based on the lesional studies, lation115 have been demonstrated in the same region. In the researchers have postulated that even though GSWDs are WAG/Rij model, focal microinfusions of phenytoin116 and initiated in the cortex, RTN is necessary to self-sustain the ethosuximide117,118 into the somatosensory cortex abol- discharges.133 ished GSWD, but intrathalamic microinfusion failed to No further conclusions can be drawn from the data pre- achieve this result.117,118 Further evidence for the impor- sented with regard to the pathophysiology of IGE. Yet, it is tance of cortex in the generation of GSWD has been demon- important to emphasize that even if there is a focal onset as strated in the FGPE model. In decorticated cats, penicillin indicated in some studies discussed earlier, IGE cannot be failed to trigger GSWD in the thalamus.119 Furthermore, interpreted as a form of focal epilepsy. The key feature is systemic injection and bilateral diffuse cortical application, “origin within and rapid engagement of bilateral networks” but not thalamic injection of penicillin triggered GSWD.107 in generalized seizures as opposed to “origin within net- In cats and monkeys, creation of acute bilateral cortical sei- works limited to one hemisphere” in focal seizures as con- zure foci by chemicals and freezing generated typical ceptualized in a recent ILAE report.2 GSWD.120 Human data from invasive EEG recordings are scarce Misdiagnosis of focal epilepsy as IGE and a few studies support the cortical focus theory. In a case The converse problem would be the misdiagnosis of focal series of four patients with GSWD on scalp EEG, depth epilepsy as IGE with potential treatment implications such electrode records demonstrated focal frontal onset in as denial of epilepsy surgery. However, in this review, we two.121 Another depth electrode study in patients with gen- have confined our literature search and discussion only to eralized seizures revealed cortical onset of GSWD.122 With focal features of IGE and its implications. stereo-EEG implanted in 10 patients, Bancaud et al.123 were able to induce typical GSWD accompanied by clinical Clinical implications of focal abnormalities in IGE absences and generalized tonic–clonic seizures by stimulat- ing the mesial frontal cortex. Tukel and Jasper124 observed Prognosis GSWD among 26 of 31 patients with parasagittal lesions. The value of focal EEG abnormalities as a predictor of However, those cases could be focal frontal lobe epilepsy prognosis in IGE is not well defined, with studies reporting with secondary bilateral synchrony rather than IGE. variable and conflicting results.27,83,134–140 Some studies Some evidence supporting focal onset of GSWD has been have reported focal EEG abnormalities as a predictor of reported from studies based on noninvasive scalp EEG. In poor prognosis in IGE, although contradicted by others dense-array EEG, source analysis localized the ictal onset (Table 4). All were hospital-based retrospective studies, of absence seizures to dorsolateral, orbital, and mesial fron- except for a single community based cohort.27 Most studies tal regions.125 Source analysis of conventional EEG yielded involved patients diagnosed with absence epilepsy, but similar results.126 Electrical field mapping of GSWD local- there appears to be an equal mix of syndromes in the two ized the field maximum to the frontal region.127 groups (poor prognosis vs. good prognosis).

Epilepsia, 55(8):1157–1169, 2014 doi: 10.1111/epi.12688 1165 Focal Features in IGE

Table 4. Studies on prognostic value of EEG focal abnormalities Age range FA is a predictor References (mean) year Total number Cohort of poor prognosis 27 ≤15 97 AE Yes 83 (34.2) 267 IGE Yes 134 NS 80 CAE, JAE No 135 (39.9 Æ 9.5) 48 JME No 136 14–27 (23.1) 119 CAE Yes 137 (23.2) 28 JME No 138 NS 32 JME Yes 139 NS 962 IGE No 140 NS 119 AE No

AE, absence epilepsy; CAE, childhood absence epilepsy; FA, focal abnormalities; IGE, idiopathic generalized epilepsy; JAE, juvenile absence epilepsy; JME, juvenile myoclonic epilepsy; NS, not speciﬁed.

Many variables including the recording technique, acti- one19 included patients based on diagnostic criteria consis- vation procedures, circadian rhythm, and length of record- tent with 1989 ILAE guidelines. ing affect the yield of EEG abnormalities in IGE.141 This fact along with heterogeneity of studies, particularly in Study limitations methodology and diagnostic criteria, should be taken into There is a growing body of literature on various focal fea- account in evaluating the prognostic value of focal EEG tures of IGE in different domains. However, certain study changes in IGE. limitations compromise the value of those data. First, almost There are no data available on the prognostic value of all studies have been conducted in tertiary centers, introduc- other focal features in IGE. ing a potential for selection bias. It is possible that more complex IGE patients with atypical features such as focal Misdiagnosis and delayed diagnosis abnormalities are overrepresented in such centers. Second, Focal abnormalities could potentially lead to delayed there is bias toward JME, with other IGE syndromes being diagnosis and misdiagnosis of IGE as focal epilepsy with less well studied. Hence, the results may not be extrapolated inappropriate AED choice as a result. These pitfalls have to all IGE syndromes. This is most likely because JME is been highlighted in several studies involving JME patients the most common IGE syndrome. Third, there is wide heter- (Table 5).5,18,19,142–146 In those studies, the reported mean ogeneity in study methodology, which could explain the diagnostic delay ranged from 5.9 to 15 years. A fair propor- variable and sometimes conflicting results across different tion of patients had been originally misdiagnosed as having studies. In particular, inclusion and diagnostic criteria are focal epilepsy (range 5.3–91.4%). As a result, many had likely to affect the results. Less stringent diagnostic criteria been on inappropriate AEDs prior to the correct diagnosis of IGE may result in inclusion of patients with focal epi- was established, resulting in seizure exacerbation. Focal lepsy in the cohort contaminating the results. It should be and asymmetric EEG and semiologic features were major noted, as detailed in previous sections, diagnostic criteria contributory factors responsible for delayed diagnosis and are not clearly defined in some studies. Finally, there is misdiagnosis. It should be noted that all studies except potential reporting and publication bias with selective

Table 5. Studies on misdiagnosis and diagnostic delay due to focal features of IGE Mean diagnostic Misdiagnosis as References Cohort Design delay (year) focal epilepsy (%) Use of inappropriate AED 5 JME 37 R 15 46 59.4% 18 JME 131 P 6.8 Æ 6.3 5.3 NS 19 JME 85 R 9.5 19 PHT most frequently used (number NS) 142 JME 76 R 5.9 15.8 52.6% 143 JME 15 R 14.5 26.7 33% 144 JME 90 R 9 17.8 58.8% (PHT), 16.6% (CBZ) 145 JME 63 R 13.3 52.4 NS 146 JME 70 P 8.3 Æ 5.5 91.4 NS

AED, antiepileptic drugs; CBZ, carbamazepine; IGE, idiopathic generalized epilepsy; JME, juvenile myoclonic epilepsy; NS, not speciﬁed; P, prospective; PHT, phenytoin; R, retrospective.

Epilepsia, 55(8):1157–1169, 2014 doi: 10.1111/epi.12688 1166 U. Seneviratne et al. reporting and publication of positive studies of focal fea- electroencephalographic classification of epileptic seizures.. – tures. It is evident in the publications discussed in this article Epilepsia 1981;22:489 501. 4. Boylan LS, Labovitz DL, Jackson SC, et al. Auras are frequent in that there are very few negative studies. Hence, it is possible idiopathic generalized epilepsy. Neurology 2006;67:343–345. that focal features in IGE may be overreported and over- 5. Vazquez B, Devinsky O, Luciano D, et al. Juvenile myoclonic – emphasized. epilepsy clinical features and factors related to misdiagnosis. J Epilepsy 1993;6:233–238. 6. Nakken KO, Solaas MH, Kjeldsen MJ, et al. The occurrence and characteristics of auras in a large epilepsy cohort. Acta Neurol Scand Conclusions 2009;119:88–93. 7. Taylor I, Marini C, Johnson MR, et al. Juvenile myoclonic epilepsy There is a large body of literature describing various and idiopathic photosensitive occipital lobe epilepsy: is there focal abnormalities in IGE. Some studies shed light on overlap? Brain 2004;127:1878–1886. underpinning pathophysiologic mechanisms of GSWD. 8. Aguglia U, Gambardella A, Le Piane E, et al. Idiopathic generalized epilepsies with versive or circling seizures. Acta Neurol Scand Whether the onset of GSWD is in the cortex or the thala- 1999;99:219–224. mus is controversial, but more recently cortical focus 9. Casaubon L, Pohlmann-Eden B, Khosravani H, et al. Video-EEG theory, mostly backed by animal data, has received much evidence of lateralized clinical features in primary generalized epilepsy with tonic-clonic seizures. Epileptic Disord 2003;5:149– emphasis. EEG-fMRI studies have shown activation of 156. thalamus and frontal cortex with spike-wave activity 10. Chin PS, Miller JW. Ictal head version in generalized epilepsy. – supporting the hypothesis of both cortical and thalamic Neurology 2004;63:370 372. 11. Kiley MA, Smith SJ, Sander JW. Idiopathic generalized epilepsy involvement in IGE. There are conflicting data on focal presenting with hemiconvulsive seizures. Epilepsia 2000;41:1633– and asymmetric EEG abnormalities as a predictor of 1636. prognosis. No information is available on the prognostic 12. Leutmezer F, Lurger S, Baumgartner C. Focal features in patients with idiopathic generalized epilepsy. Epilepsy Res 2002;50:293–300. relevance of other focal features in IGE. Focal features, 13. Niaz FE, Abou-Khalil B, Fakhoury T. The generalized tonic-clonic particularly in EEG and semiology, may result in misdiag- seizure in partial versus generalized epilepsy: semiologic – nosis and delayed diagnosis of IGE. As a result, the patient differences. Epilepsia 1999;40:1664 1666. 14. Park KI, Lee SK, Chu K, et al. The value of video-EEG monitoring may receive an inappropriate AED, leading to poor seizure to diagnose juvenile myoclonic epilepsy. Seizure 2009;18:94–99. control. This perhaps is the most important practical impli- 15. Usui N, Kotagal P, Matsumoto R, et al. Focal semiologic and cation of focal features in IGE. Clinicians need to be aware electroencephalographic features in patients with juvenile myoclonic epilepsy. Epilepsia 2005;46:1668–1676. of this pitfall. 16. Walser G, Unterberger I, Dobesberger J, et al. Asymmetric seizure termination in primary and secondary generalized tonic-clonic seizures. Epilepsia 2009;50:2035–2039. Conflict of Interest 17. Oguni H, Mukahira K, Oguni M, et al. Video-polygraphic analysis of myoclonic seizures in juvenile myoclonic epilepsy. Epilepsia A/Professor D’Souza has received travel, investigator-initiated, and 1994;35:307–316. speaker honoraria from UCB Pharma; educational grants from Novartis 18. Murthy JM, Rao CM, Meena AK. Clinical observations of juvenile Pharmaceuticals and Pfizer Pharmaceuticals; educational, travel, and fel- myoclonic epilepsy in 131 patients: a study in South India. Seizure lowship grants from GlaxoSmithKline Neurology Australia; and honoraria 1998;7:43–47. from SciGen Pharmaceuticals. Professor Cook has received speaker hono- 19. Lancman ME, Asconape JJ, Penry JK. Clinical and EEG raria from UCB Pharma and Sanofi Australia and travel honoraria from asymmetries in juvenile myoclonic epilepsy. Epilepsia UCB Pharma and SciGen. He has received research grants from National 1994;35:302–306. Health and Medical Research Council (Australia), and Australian Research 20. Sadleir LG, Scheffer IE, Smith S, et al. Automatisms in absence Council. He has also received a Science, Technology, and Innovation Grant seizures in children with idiopathic generalized epilepsy. Arch from the State Government of Victoria, Australia. Dr. Seneviratne has Neurol 2009;66:729–734. received an educational grant and speaker honoraria from UCB Pharma. 21. O’Brien JL, Goldensohn ES, Hoefer PF. Electroencephalographic We confirm that we have read the Journal’s position on issues involved in abnormalities in addition to bilaterally synchronous 3 per second ethical publication and affirm that this report is consistent with those guide- spike and wave activity in petit mal. Electroencephalogr Clin lines. Neurophysiol 1959;11:747–761. 22. Lombroso CT. Consistent EEG focalities detected in subjects with primary generalized epilepsies monitored for two decades. Epilepsia Funding 1997;38:797–812. 23. Aliberti V, Grunewald RA, Panayiotopoulos CP, et al. Focal None. electroencephalographic abnormalities in juvenile myoclonic epilepsy. Epilepsia 1994;35:297–301. 24. Jayalakshmi SS, Srinivasa Rao B, Sailaja S. Focal clinical and References electroencephalographic features in patients with juvenile myoclonic epilepsy. Acta Neurol Scand 2010;122:115–123. 1. Commission on Classification and Terminology of the International 25. Panayiotopoulos CP, Obeid T, Tahan AR. Juvenile myoclonic League Against Epilepsy. Proposal for revised classification of epilepsy: a 5-year prospective study. Epilepsia 1994;35:285–296. epilepsies and epileptic syndromes. Epilepsia 1989;30:389–399. 26. Covanis A, Skiadas K, Loli N, et al. Absence epilepsy: early 2. Berg AT, Berkovic SF, Brodie MJ, et al. Revised terminology and prognostic signs. Seizure 1992;1:281–289. concepts for organization of seizures and epilepsies: report of the 27. Hedstrom A, Olsson I. Epidemiology of absence epilepsy: EEG ILAE Commission on Classification and Terminology, 2005–2009. findings and their predictive value. Pediatr Neurol 1991;7:100–104. Epilepsia 2010;51:676–685. 28. Matur Z, Baykan B, Bebek N, et al. The evaluation of interictal focal 3. Commission on Classification and Terminology of the International EEG findings in adult patients with absence seizures. Seizure League Against Epilepsy. Proposal for revised clinical and 2009;18:352–358.

Epilepsia, 55(8):1157–1169, 2014 doi: 10.1111/epi.12688 1167 Focal Features in IGE

29. Tezer FI, Sahin G, Ciger A, et al. Focal EEG findings in juvenile 53. Betting LE, Li LM, Lopes-Cendes I, et al. Correlation between absence syndrome and the effect of antiepileptic drugs. Clin EEG quantitative EEG and MRI in idiopathic generalized epilepsy. Hum Neurosci 2008;39:33–38. Brain Mapp 2010;31:1327–1338. 30. Vierck E, Cauley R, Kugler SL, et al. Polyspike and waves do not 54. Betting LE, Mory SB, Lopes-Cendes I, et al. MRI volumetry shows predict generalized tonic-clonic seizures in childhood absence increased anterior thalamic volumes in patients with absence epilepsy. J Child Neurol 2010;25:475–481. seizures. Epilepsy Behav 2006;8:575–580. 31. Gelisse P, Serafini A, Velizarova R, et al. Temporal intermittent 55. Tae WS, Hong SB, Joo EY, et al. Structural brain abnormalities in delta activity: a marker of juvenile absence epilepsy? Seizure juvenile myoclonic epilepsy patients: volumetry and voxel-based 2011;20:38–41. morphometry. Korean J Radiol 2006;7:162–172. 32. Koutroumanidis M, Tsiptsios D, Kokkinos V, et al. Focal and 56. Caplan R, Levitt J, Siddarth P, et al. Frontal and temporal volumes generalized EEG paroxysms in childhood absence epilepsy: in Childhood Absence Epilepsy. Epilepsia 2009;50:2466–2472. topographic associations and distinctive behaviors during the first 57. Betting LE, Mory SB, Li LM, et al. Voxel-based morphometry in cycle of non-REM sleep. Epilepsia 2012;53:840–849. patients with idiopathic generalized epilepsies. Neuroimage 33. Stuss DT, Levine B. Adult clinical neuropsychology: lessons from 2006;32:498–502. studies of the frontal lobes. Annu Rev Psychol 2002;53:401–433. 58. Chan CH, Briellmann RS, Pell GS, et al. Thalamic atrophy in 34. Swartz BE, Halgren E, Simpkins F, et al. Primary memory in childhood absence epilepsy. Epilepsia 2006;47:399–405. patients with frontal and primary generalized epilepsy. J Epilepsy 59. Helms G, Ciumas C, Kyaga S, et al. Increased thalamus levels of 1994;7:232–241. glutamate and glutamine (Glx) in patients with idiopathic 35. Iqbal N, Caswell HL, Hare DJ, et al. Neuropsychological profiles of generalised epilepsy. J Neurol Neurosurg Psychiatry 2006;77:489– patients with juvenile myoclonic epilepsy and their siblings: a 494. preliminary controlled experimental video-EEG case series. 60. Kim JH, Lee JK, Koh SB, et al. Regional grey matter abnormalities Epilepsy Behav 2009;14:516–521. in juvenile myoclonic epilepsy: a voxel-based morphometry study. 36. Kim JH, Suh SI, Park SY, et al. Microstructural white matter Neuroimage 2007;37:1132–1137. abnormality and frontal cognitive dysfunctions in juvenile 61. Lin K, Jackowski AP, Carrete H Jr, et al. Voxel-based morphometry myoclonic epilepsy. Epilepsia 2012;53:1371–1378. evaluation of patients with photosensitive juvenile myoclonic 37. Moschetta SP, Valente KD. Juvenile myoclonic epilepsy: the impact epilepsy. Epilepsy Res 2009;86:138–145. of clinical variables and psychiatric disorders on executive profile 62. Pardoe H, Pell GS, Abbott DF, et al. Multi-site voxel-based assessed with a comprehensive neuropsychological battery. Epilepsy morphometry: methods and a feasibility demonstration with Behav 2012;25:682–686. childhood absence epilepsy. Neuroimage 2008;42:611–616. 38. O’Muircheartaigh J, Vollmar C, Barker GJ, et al. Focal structural 63. Woermann FG, Free SL, Koepp MJ, et al. Abnormal cerebral changes and cognitive dysfunction in juvenile myoclonic epilepsy. structure in juvenile myoclonic epilepsy demonstrated with voxel- Neurology 2011;76:34–40. based analysis of MRI. Brain 1999;122:2101–2108. 39. Pascalicchio TF, de Araujo Filho GM, da Silva Noffs MH, et al. 64. Pell GS, Briellmann RS, Chan CH, et al. Selection of the control Neuropsychological profile of patients with juvenile myoclonic group for VBM analysis: influence of covariates, matching and epilepsy: a controlled study of 50 patients. Epilepsy Behav sample size. Neuroimage 2008;41:1324–1335. 2007;10:263–267. 65. Yasuda CL, Betting LE, Cendes F. Voxel-based morphometry and 40. Pulsipher DT, Seidenberg M, Guidotti L, et al. Thalamofrontal epilepsy. Expert Rev Neurother 2010;10:975–984. circuitry and executive dysfunction in recent-onset juvenile 66. Ioannidis JP. Excess significance bias in the literature on brain myoclonic epilepsy. Epilepsia 2009;50:1210–1219. volume abnormalities. Arch Gen Psychiatry 2011;68:773–780. 41. Roebling R, Scheerer N, Uttner I, et al. Evaluation of cognition, 67. Richardson M. Update on neuroimaging in epilepsy. Expert Rev structural, and functional MRI in juvenile myoclonic epilepsy. Neurother 2010;10:961–973. Epilepsia 2009;50:2456–2465. 68. Bernasconi A, Bernasconi N, Natsume J, et al. Magnetic resonance 42. Sonmez F, Atakli D, Sari H, et al. Cognitive function in juvenile spectroscopy and imaging of the thalamus in idiopathic generalized myoclonic epilepsy. Epilepsy Behav 2004;5:329–336. epilepsy. Brain 2003;126:2447–2454. 43. Piazzini A, Turner K, Vignoli A, et al. Frontal cognitive dysfunction 69. de Araujo Filho GM, Lin K, Lin J, et al. Are personality traits of in juvenile myoclonic epilepsy. Epilepsia 2008;49:657–662. juvenile myoclonic epilepsy related to frontal lobe dysfunctions? A 44. Wandschneider B, Kopp UA, Kliegel M, et al. Prospective memory proton MRS study. Epilepsia 2009;50:1201–1209. in patients with juvenile myoclonic epilepsy and their healthy 70. Doelken MT, Mennecke A, Stadlbauer A, et al. Multi-voxel siblings. Neurology 2010;75:2161–2167. magnetic resonance spectroscopy at 3 T in patients with idiopathic 45. Zamarian L, Hofler J, Kuchukhidze G, et al. Decision making in generalised epilepsy. Seizure 2010;19:485–492. juvenile myoclonic epilepsy. J Neurol 2013;260:839–846. 71. Fojtikova D, Brazdil M, Horky J, et al. Magnetic resonance 46. Wandschneider B, Centeno M, Vollmar C, et al. Risk-taking spectroscopy of the thalamus in patients with typical absence behavior in juvenile myoclonic epilepsy. Epilepsia 2013;54:2158– epilepsy. Seizure 2006;15:533–540. 2165. 72. Haki C, Gumustas OG, Bora I, et al. Proton magnetic resonance 47. Hommet C, Billard C, Motte J, et al. Cognitive function in spectroscopy study of bilateral thalamus in juvenile myoclonic adolescents and young adults in complete remission from benign epilepsy. Seizure 2007;16:287–295. childhood epilepsy with centro-temporal spikes. Epileptic Disord 73. Kabay SC, Gumustas OG, Karaman HO, et al. A proton magnetic 2001;3:207–216. resonance spectroscopic study in juvenile absence epilepsy in early 48. Pavone P, Bianchini R, Trifiletti RR, et al. Neuropsychological stages. Eur J Paediatr Neurol 2010;14:224–228. assessment in children with absence epilepsy. Neurology 74. Lin K, Carrete H Jr, Lin J, et al. Magnetic resonance spectroscopy 2001;56:1047–1051. reveals an epileptic network in juvenile myoclonic epilepsy. 49. Ciumas C, Savic I. Structural changes in patients with primary Epilepsia 2009;50:1191–1200. generalized tonic and clonic seizures. Neurology 2006;67:683–686. 75. Mory SB, Li LM, Guerreiro CA, et al. Thalamic dysfunction in 50. Mory SB, Betting LE, Fernandes PT, et al. Structural abnormalities juvenile myoclonic epilepsy: a proton MRS study. Epilepsia of the thalamus in juvenile myoclonic epilepsy. Epilepsy Behav 2003;44:1402–1405. 2011;21:407–411. 76. Savic I, Lekvall A, Greitz D, et al. MR spectroscopy shows reduced 51. Natsume J, Bernasconi N, Andermann F, et al. MRI volumetry of the frontal lobe concentrations of N-acetyl aspartate in patients with thalamus in temporal, extratemporal, and idiopathic generalized juvenile myoclonic epilepsy. Epilepsia 2000;41:290–296. epilepsy. Neurology 2003;60:1296–1300. 77. Savic I, Osterman Y, Helms G. MRS shows syndrome differentiated 52. Seeck M, Dreifuss S, Lantz G, et al. Subcortical nuclei volumetry in metabolite changes in human-generalized epilepsies. Neuroimage idiopathic generalized epilepsy. Epilepsia 2005;46:1642–1645. 2004;21:163–172.

Epilepsia, 55(8):1157–1169, 2014 doi: 10.1111/epi.12688 1168 U. Seneviratne et al.

78. Simister RJ, McLean MA, Barker GJ, et al. A proton magnetic 102. Jeha LE, Morris HH, Burgess RC. Coexistence of focal and resonance spectroscopy study of metabolites in the occipital lobes in idiopathic generalized epilepsy in the same patient population. epilepsy. Epilepsia 2003;44:550–558. Seizure 2006;15:28–34. 79. Simister RJ, McLean MA, Barker GJ, et al. Proton MRS reveals 103. Nicolson A, Chadwick DW, Smith DF. The coexistence of frontal lobe metabolite abnormalities in idiopathic generalized idiopathic generalized epilepsy and partial epilepsy. Epilepsia epilepsy. Neurology 2003;61:897–902. 2004;45:682–685. 80. Gotman J, Grova C, Bagshaw A, et al. Generalized epileptic 104. Penfield WG, Jasper HH. Epilepsy and functional anatomy of the discharges show thalamocortical activation and suspension of the human brain. Boston, MA: Little Brown & Co; 1954. default state of the brain. Proc Natl Acad Sci U S A 2005;102:15236– 105. Buzsaki G. The thalamic clock: emergent network properties. 15240. Neuroscience 1991;41:351–364. 81. Aghakhani Y, Bagshaw AP, Benar CG, et al. fMRI activation during 106. Gloor P. Generalized cortico-reticular epilepsies. Some spike and wave discharges in idiopathic generalized epilepsy. Brain considerations on the pathophysiology of generalized bilaterally 2004;127:1127–1144. synchronous spike and wave discharge. Epilepsia 1968;9:249–263. 82. Archer JS, Abbott DF, Waites AB, et al. fMRI “deactivation” of the 107. Quesney LF, Gloor P. Generalized penicillin epilepsy in the cat: posterior cingulate during generalized spike and wave. Neuroimage correlation between electrophysiological data and distribution of 2003;20:1915–1922. 14C-penicillin in the brain. Epilepsia 1978;19:35–45. 83. Bai X, Vestal M, Berman R, et al. Dynamic time course of typical 108. Kostopoulos GK. Spike-and-wave discharges of absence seizures as childhood absence seizures: EEG, behavior, and functional magnetic a transformation of sleep spindles: the continuing development of a resonance imaging. J Neurosci 2010;30:5884–5893. hypothesis. Clin Neurophysiol 2000;111 (Suppl. 2):S27–S38. 84. Berman R, Negishi M, Vestal M, et al. Simultaneous EEG, fMRI, 109. Meeren H, van Luijtelaar G, Lopes da Silva F, et al. Evolving and behavior in typical childhood absence seizures. Epilepsia concepts on the pathophysiology of absence seizures: the cortical 2010;51:2011–2022. focus theory. Arch Neurol 2005;62:371–376. 85. Hamandi K, Salek-Haddadi A, Laufs H, et al. EEG-fMRI of 110. Meeren HK, Pijn JP, Van Luijtelaar EL, et al. Cortical focus drives idiopathic and secondarily generalized epilepsies. Neuroimage widespread corticothalamic networks during spontaneous absence 2006;31:1700–1710. seizures in rats. J Neurosci 2002;22:1480–1495. 86. Liu Y, Yang T, Liao W, et al. EEG-fMRI study of the ictal and 111. Niedermeyer E. Primary (idiopathic) generalized epilepsy and interictal epileptic activity in patients with eyelid myoclonia with underlying mechanisms. Clin Electroencephalogr 1996;27:1–21. absences. Epilepsia 2008;49:2078–2086. 112. Pavone A, Niedermeyer E. Absence seizures and the frontal lobe. 87. Moeller F, LeVan P, Muhle H, et al. Absence seizures: individual Clin Electroencephalogr 2000;31:153–156. patterns revealed by EEG-fMRI. Epilepsia 2010;51:2000–2010. 113. Polack PO, Guillemain I, Hu E, et al. Deep layer somatosensory 88. Moeller F, Maneshi M, Pittau F, et al. Functional connectivity in cortical neurons initiate spike-and-wave discharges in a genetic patients with idiopathic generalized epilepsy. Epilepsia model of absence seizures. J Neurosci 2007;27:6590–6599. 2011;52:515–522. 114. Klein JP, Khera DS, Nersesyan H, et al. Dysregulation of sodium 89. Moeller F, Siebner HR, Ahlgrimm N, et al. fMRI activation during channel expression in cortical neurons in a rodent model of absence spike and wave discharges evoked by photic stimulation. epilepsy. Brain Res 2004;1000:102–109. Neuroimage 2009;48:682–695. 115. Luttjohann A, Zhang S, de Peijper R, et al. Electrical stimulation of 90. Moeller F, Siebner HR, Wolff S, et al. Changes in activity of striato- the epileptic focus in absence epileptic WAG/Rij rats: assessment of thalamo-cortical network precede generalized spike wave local and network excitability. Neuroscience 2011;188:125–134. discharges. Neuroimage 2008;39:1839–1849. 116. Gurbanova AA, Aker R, Berkman K, et al. Effect of systemic and 91. Moeller F, Siebner HR, Wolff S, et al. Simultaneous EEG-fMRI in intracortical administration of phenytoin in two genetic models of drug-naive children with newly diagnosed absence epilepsy. absence epilepsy. Br J Pharmacol 2006;148:1076–1082. Epilepsia 2008;49:1510–1519. 117. Manning JP, Richards DA, Leresche N, et al. Cortical-area specific 92. Szaflarski JP, DiFrancesco M, Hirschauer T, et al. Cortical and block of genetically determined absence seizures by ethosuximide. subcortical contributions to absence seizure onset examined with Neuroscience 2004;123:5–9. EEG/fMRI. Epilepsy Behav 2010;18:404–413. 118. Richards DA, Manning JP, Barnes D, et al. Targeting thalamic 93. Szaflarski JP, Kay B, Gotman J, et al. The relationship between the nuclei is not sufficient for the full anti-absence action of localization of the generalized spike and wave discharge generators ethosuximide in a rat model of absence epilepsy. Epilepsy Res and the response to valproate. Epilepsia 2013;54:471–480. 2003;54:97–107. 94. Vollmar C, O’Muircheartaigh J, Barker GJ, et al. Motor system 119. Avoli M, Gloor P. Interaction of cortex and thalamus in spike and hyperconnectivity in juvenile myoclonic epilepsy: a cognitive wave discharges of feline generalized penicillin epilepsy. Exp functional magnetic resonance imaging study. Brain 2011;134: Neurol 1982;76:196–217. 1710–1719. 120. Marcus EM, Watson CW, Simon SA. An experimental model of 95. Engel J Jr, Lubens P, Kuhl DE, et al. Local cerebral metabolic rate some varieties of petit mal epilepsy. Electrical-behavioral for glucose during petit mal absences. Ann Neurol 1985;17:121–128. correlations of acute bilateral epileptogenic foci in cerebral cortex. 96. Ochs RF, Gloor P, Tyler JL, et al. Effect of generalized spike-and- Epilepsia 1968;9:233–248. wave discharge on glucose metabolism measured by positron 121. Niedermeyer E, Laws ER Jr, Walker EA. Depth EEG findings in emission tomography. Ann Neurol 1987;21:458–464. epileptics with generalized spike-wave complexes. Arch Neurol 97. Swartz BE, Simpkins F, Halgren E, et al. Visual working memory in 1969;21:51–58. primary generalized epilepsy: an 18FDG-PET study. Neurology 122. Hayne RA, Belinson L, Gibbs FA. Electrical activity of subcortical 1996;47:1203–1212. areas in epilepsy. Electroencephalogr Clin Neurophysiol 1949; 98. Kim JH, Im KC, Kim JS, et al. Correlation of interictal spike-wave 1:437–445. with thalamic glucose metabolism in juvenile myoclonic epilepsy. 123. Bancaud J, Talairach J, Morel P, et al. “Generalized” epileptic Neuroreport 2005;16:1151–1155. seizures elicited by electrical stimulation of the frontal lobe in man. 99. McDonald CR, Swartz BE, Halgren E, et al. The relationship of Electroencephalogr Clin Neurophysiol 1974;37:275–282. regional frontal hypometabolism to executive function: a resting 124. Tukel K, Jasper H. The electroencephalogram in parasagittal lesions. fluorodeoxyglucose PET study of patients with epilepsy and healthy Electroencephalogr Clin Neurophysiol 1952;4:481–494. controls. Epilepsy Behav 2006;9:58–67. 125. Holmes MD, Brown M, Tucker DM. Are “generalized” seizures 100. Meencke HJ, Janz D. The significance of microdysgenesia in truly generalized? Evidence of localized mesial frontal and primary generalized epilepsy: an answer to the considerations of frontopolar discharges in absence. Epilepsia 2004;45:1568–1579. Lyon and Gastaut. Epilepsia 1985;26:368–371. 126. Rodin EA, Rodin MK, Thompson JA. Source analysis of generalized 101. Opeskin K, Kalnins RM, Halliday G, et al. Idiopathic generalized spike-wave complexes. Brain Topogr 1994;7:113–119. epilepsy: lack of significant microdysgenesis. Neurology 2000;55: 127. Rodin E, Ancheta O. Cerebral electrical fields during petit mal 1101–1106. absences. Electroencephalogr Clin Neurophysiol 1987;66:457–466.

Epilepsia, 55(8):1157–1169, 2014 doi: 10.1111/epi.12688 1169 Focal Features in IGE

128. Inoue M, Duysens J, Vossen JM, et al. Thalamic multiple-unit 137. Letourneau K, Cieuta-Walti C, Deacon C. Epileptiform asymetries activity underlying spike-wave discharges in anesthetized rats. Brain and treatment response in juvenile myoclonic epilepsy. Can J Neurol Res 1993;612:35–40. Sci 2010;37:826–830. 129. Vergnes M, Marescaux C, Depaulis A, et al. Spontaneous spike and 138. Matsuoka H. The seizure prognosis of juvenile myoclonic epilepsy. wave discharges in thalamus and cortex in a rat model of genetic Jpn J Psychiatry Neurol 1992;46:293–296. petit mal-like seizures. Exp Neurol 1987;96:127–136. 139. Nicolson A, Appleton RE, Chadwick DW, et al. The relationship 130. Pollen DA, Perot P, Reid KH. Experimental bilateral wave and spike between treatment with valproate, lamotrigine, and topiramate and from thalamic stimulation in relation to level of arousal. the prognosis of the idiopathic generalised epilepsies. J Neurol Electroencephalogr Clin Neurophysiol 1963;15:1017–1028. Neurosurg Psychiatry 2004;75:75–79. 131. Laws E, Niedermeyer E, Walker AE. Depth EEG findings in 140. Sinclair DB, Unwala H. Absence epilepsy in childhood: epileptics with generalized spike-wave complexes. electroencephalography (EEG) does not predict outcome. J Child Electroencephalogr Clin Neurophysiol 1970;28:94–95. Neurol 2007;22:799–802. 132. Avanzini G, Vergnes M, Spreafico R, et al. Calcium-dependent 141. Seneviratne U, Cook M, D’Souza W. The electroencephalogram of regulation of genetically determined spike and waves by the reticular idiopathic generalized epilepsy. Epilepsia 2012;53:234–248. thalamic nucleus of rats. Epilepsia 1993;34:1–7. 142. Atakli D, Sozuer D, Atay T, et al. Misdiagnosis and treatment in 133. Meeren HK, Veening JG, Moderscheim TA, et al. Thalamic juvenile myoclonic epilepsy. Seizure 1998;7:63–66. lesions in a genetic rat model of absence epilepsy: dissociation 143. Grunewald RA, Chroni E, Panayiotopoulos CP. Delayed diagnosis between spike-wave discharges and sleep spindles. Exp Neurol of juvenile myoclonic epilepsy. J Neurol Neurosurg Psychiatry 2009;217:25–37. 1992;55:497–499. 134. Bartolomei F, Roger J, Bureau M, et al. Prognostic factors for 144. Lancman ME, Asconape JJ, Brotherton T, et al. Juvenile childhood and juvenile absence epilepsies. Eur Neurol myoclonic epilepsy – an underdiagnosed syndrome. J Epilepsy 1997;37:169–175. 1995;8:215–218. 135. Baykan B, Altindag EA, Bebek N, et al. Myoclonic seizures subside 145. Montalenti E, Imperiale D, Rovera A, et al. Clinical features, EEG in the fourth decade in juvenile myoclonic epilepsy. Neurology findings and diagnostic pitfalls in juvenile myoclonic epilepsy: a 2008;70:2123–2129. series of 63 patients. J Neurol Sci 2001;184:65–70. 136. Grosso S, Galimberti D, Vezzosi P, et al. Childhood absence 146. Panayiotopoulos CP, Tahan R, Obeid T. Juvenile myoclonic epilepsy: evolution and prognostic factors. Epilepsia epilepsy: factors of error involved in the diagnosis and treatment. 2005;46:1796–1801. Epilepsia 1991;32:672–676.

Epilepsia, 55(8):1157–1169, 2014 doi: 10.1111/epi.12688