Epileptic Disorders - Editor's Choice Archive
The following are past Editor's Choice articles. Free access may not be available.
Paediatric epilepsy surgery in the posterior cortex: a study of 62 cases
Alexandra Liava, Roberto Mai, Laura Tassi, Massimo Cossu, Ivana Sartori, Lino Nobili, Giorgio Lo Russo, Stefano Francione
Commentary by Pr. Michael Duchowny, Associate Editor
Intractable focal epilepsy that begins in early postnatal life is associated with long-term neuro-cognitive dysfunction. Recognition of this poor prognosis is a decisive factor in the early surgical referral with a goal of definitive cure. However, candidates for surgical therapy in the first decade of life have many different etiologies and multiple regions of involvement that are approached utilizing different anatomic procedures. Despite these challenges, temporal and extra-temporal pediatric epilepsy surgery data in very early life is now widely available and serves as a helpful guide; the experience with MRI-negative patients has also been described (Duchowny et al, 1998; Jayakar et al, 2009). Missing from this picture, however, are large cohort studies focused more specifically on childhood surgery for focal seizures originating in the posterior hemispheric regions.
The recently published cohort study of 62 children with focal epilepsy arising in the posterior cerebral hemispheres (Liava et al, 2014) is therefore a welcome step towards filling this gap. Although the outcome of posterior hemisphere pediatric epilepsy surgery has been reported previously (Fogarisi et al, 2003), additional carefully collected data is extremely valuable. Liava and colleagues (2014) describe their experience with 62 cases undergoing excisional procedures and functional disconnections who had thorough pre-operative investigations. The cohort was primarily MRI-positive- only 3 cases had unremarkable MRI studies, and 24 patients (39%) underwent stereo-EEG recordings. Not surprisingly, the cohort was heterogeneous with regard to etiology, surgical procedure and region(s) of involvement. The occurrence of varying degrees of pre-existing cognitive disturbance in a high proportion of cases is fully consistent with both pediatric surgical series and childhood epilepsy in general. The 29% incidence of hemianopic visual field defect is also not surprising in this selected cohort.
The investigators achieved a very high rate of seizure freedom with 86% achieving Engel Class 1 outcomes. This result is even more remarkable given the authors acknowledgement that complete resection “did not necessarily imply the complete resection of the MRI-detectable lesion”. That this cohort could attain such a high rate of seizure-freedom is welcome news for pediatric epilepsy surgery teams.
Although posterior hemispheric cases constitute a minority pediatric epilepsy surgery caseloads, they are by no means rare. Cortical malformations can occur anywhere within the cerebral cortex and osterior cortical involvement is common. The Milano cohort included a high proportion of multi-lobar cases, many of whom had ill-defined resection margins and involvement of eloquent (visual, somatosensory) cortex. Addressing both issues is a primary pediatric epilepsy surgery challenge and often contributes to compromised surgical success. As there is greater experience in adult epilepsy surgical populations with posterior resections and their challenges, the results of the Milano group offer important reassurance that similar approaches can also be used successfully in very young patients.
An important distinguishing feature of the Liava et al (2014) study is their attention to their patients’ seizure semiology. As the authors correctly note in their introduction, seizures that arise in the posterior cortex may propagate widely via multiple fascicular pathways resulting in clinical features more typical of secondary sites. This issue is especially problematic for seizures that arise in the parietal lobe as electrographic seizure onset is often clinically silent and thus associated with a falsely localizing “pseudo-focus” (Jayakar et al, 1991). With this understanding, the authors were able to identify a specific complex of symptoms and signs that were strongly suggestive of primary occipital lobe seizure origin. These included the presence of a pre-existing visual field defect and lateralized complex visual hallucinations in conjunction with oculogyric or oculo-clonic movement at ictal onset. Equally important, the investigators noted that ictal blindness and eye deviation were much less specific. Fear was a frequent subjective ictal complaint that disappeared in six of 7 patients undergoing parietal resection. However, although the perception of fear is described in patients with parietal lobe seizures, caution should be exercised as ictal fear can occur in patients with seizures arising in the frontal and temporal lobes (Toth et al, 2010; Akiyama et al, 2014).
All three MRI-negative patients had poor outcomes. While the presence of a discrete lesion on imaging is a useful marker of the epileptogenic zone, favorable postoperative outcomes are still possible without it (Jayakar et al, 2009). Functional imaging was not employed in the pre-operative evaluations of the Liava et al (2014) cohort and one is left wondering whether including this information might have made a difference. As electrophysiological information was insufficient for surgical planning in approximately one-third of cases and the scalp interictal EEG was judged to be misleading in approximately one-fifth of cases, it is understandable that accurate localization would be extremely difficult. However, for MRI-positive patients, complete removal of the anatomic lesion was not required for seizure-freedom, and the MRI-negative group, though small, likely failed for different reasons. It is known that resection of either the anatomic lesion or the epileptogenic zone can result in good outcome even if the other is not fully excised (Perry et al, 2010). The epilepsy in the MRI-negative patients was probably secondary to Type 1 focal cortical dysplasia and surgical failure was likely due to incomplete resection of a dysfunctionally activated cortical network. Whether the ability to define dysfunctional epileptic networks in the posterior cerebral hemisphere is comparable to anterior networks remains unknown. It is even less certain whether anatomically diverse networks are functionally similar.
The Liava et al (2014) study findings should be viewed as an important step in our progress towards understanding the surgical implications of lesion-driven posterior hemispheric seizure surgery. Lesionectomy, lobectomy, multi-lobar resection and disconnection, the surgical procedures employed in this report, are all preferable to more radical procedures such as hemispherectomy- the ability to achieve function-sparing seizure-freedom is obvious. The rate of seizure-freedom in this series is certainly comparable to the outcomes reported for frontal lobe cases. The justification for this “less is more” approach will become more secure if longitudinal studies can assess the long-term cognitive, behavioral and social outcomes of early posterior hemisphere surgery, particularly after tailored resection and lesionectomy. Pre-operative and post-operative neuropsychological data was collected by the investigators in this study and will hopefully be reported in a future publication.
Cortical network dysfunction in musicogenic epilepsy reflecting the role of snowballing emotional processes in seizure generation: an fMRI-EEG study.
Volker Diekmann & Anselm C. Hoppner
Comments by Associate Editor Sara Wilson
Within the field of epilepsy research, musicogenic epilepsy has received scant attention, with little known about its underlying mechanisms despite being described over 75 years ago. By contrast, the field of music neuroscience is burgeoning with studies investigating the music networks of the brain, including the potent ability of music to activate reward and emotion processing systems. Musicogenic seizures provide an obvious opportunity to investigate these networks and their dysfunction in people with epilepsy. In the current study this was done in a patient using the gold-standard technique of EEG-fMRI.
An intriguing aspect of musicogenic epilepsy is triggering by highly selective music stimuli, which in this case was a specific type of Russian music as opposed to similar music of different origin. In addition to pink noise, the latter provided an ideal control stimulus, allowing the specificity of network dysfunction to be explored. This revealed activation of the affective network, particularly regions involved in positive emotional processing, at the onset of the Russian music. A clear strength of the study was subsequent exploration of the way this activation changed over time, prior to the emergence of a seizure. This analysis identified maladaptive interactions between components of the cognitive control, autobiographic memory and affective networks, with reduced cognitive down regulation thought to allow ‘snowballing’ activity in emotion-related areas, ultimately triggering the seizure.By drawing on the broader cognitive neuroscience literature investigating network function in healthy individuals combined with careful examination of this instructive case, this study presents a compelling argument that musicogenic epilepsy may not just arise from musicological features of the stimulus per se. Moreover, the researchers propose that similar mechanisms of cognitive dysregulation of emotion processing networks may apply in other cases of reflexive epilepsy, such as reading epilepsy. This interesting hypothesis provides an exciting direction for future research into this poorly understood form of epilepsy.
Generalised electrographic seizures presenting as perioral myoclonia
Jennifer Deraborn and Peter Kaplan
A 41-year-old man had several episodes of rhythmic and intermittent, sometimes lateralized chin twitching lasting over a week for a seven years period. He reported understanding what was said but had difficulty replying due to the chin movement. The EEG during the episode showed paroxysms of polyspike and slow wave activity, maximal over the fronto-central regions, correlating with the chin movements. Brain imaging was normal. On one occasion, the episode was followed by a generalized tonic-clonic seizure. Levetiracetam added to valproate resolved the episode.
The authors along with video presentation guide the reader to the interesting differential diagnosis of this peculiar case with late onset long-lasting repetitive myoclonia localized to the perioral area with preserved consciousness. Published with videosequences
Extreme startle and photomyoclonic response in severe hypocalcaemia
Marcello Moccia, Roberto Erro, Elvira Nicolella, Pasquale Striano, Salvatore Striano
Moccia and colleagues report a case with episodes of sudden generalised muscle contractions resulting in violent falls with injuries and urinary incontinence. The fits appeared to relate to sudden auditory stimuli. During the examination, a sudden and unexpected noise triggered a violent startle with a forceful closure of eyes, cranial muscle contraction, and raising of arms over the head, immediately followed by generalised stiffness for several seconds. Flashing light triggered an excessive startle followed by myogenic potentials. There was no epileptic activity on the EEG. Blood tests revealed severe hypocalcaemia and the symptoms disappeared after calcium supplementation.
This report teaches us that electrolyte disturbances can be associated with abnormal startle responses. Published with videosequences
Confirming an expanded spectrum of SCN2A mutations: a case series
Dena Matalon, Ethan Goldberg, Livija Medne, Eric D. Marsh
Comments by Associate Editor Aristea S. Galanopoulou MD PhD
Mutations in voltage-sensitive sodium channels (SCN) have been increasingly identified in human epilepsy syndromes. These syndromes include both “benign” forms (e.g., benign familial neonatal-infantile epilepsy) and more severe types (e.g., Dravet syndrome, infantile epileptic encephalopathies). Matalon and colleagues here describe the clinical features of 3 girls with 3 different de novo SCN2A mutations that presented with early onset epilepsies, poor developmental outcomes and axial hypotonia. Thinning of the corpus callosum was eventually identified in two of the patients. Two of the girls manifested infantile spasms whereas the third had neonatal tonic seizures. In one of the patients with infantile spasms choreiform movements were also observed.
Of interest, a subsequent independent study by Hackenberg et al (Neuropediatrics, 2014) describes an identical mutation (c.4025T>C) of the SCN2A gene in a different girl who also presented with infantile spasms, choreiform movements, hypersomnia and progressive brain atrophy. These observations lend support to the pathogenicity of this specific mutation but also to its functional impact on both cortical and subcortical structures. Matalon et al nicely review the literature on genotype-phenotype correlations of the known human SCN2A mutations, discussing the inherent heterogeneity but also increasing evidence for a possible involvement of these mutations as modifiers of human epilepsies. Further studies are certainly needed to confirm the functional consequences of these mutations in the developing brain. The need for creation of comprehensive public access genotype-phenotype databases and animal models of the known and emerging epilepsy-related mutations is however becoming increasingly evident, so as to best study their clinical and pathogenic relevance.