J Neurosurg Pediatrics 12:288–292, 2013 ©AANS, 2013 Cerebellar seizures Report of 2 cases Sarah Boop, B.A.,1 James Wheless, M.D.,1–4 Katherine Van Poppel, M.D.,1–4 Amy McGregor, M.D.,1–4 and Frederick A. Boop, M.D.1,2,4–6 Le Bonheur Children’s Hospital; 3Division of Pediatric Neurology, 4Comprehensive Epilepsy Program, and Division of Neurosurgery, Le Bonheur Children’s Hospital; and Departments of 2Pediatrics and 6Pediatric Neurosurgery, the University of Tennessee Health Science Center, Memphis, Tennessee 1 5 Epilepsy, especially with refractory seizures, is thought to arise only from cortical lesions or substrate. The authors report on 2 patients with refractory epilepsy and cerebellar lesions. Depth electrodes were placed within the cerebellar lesions in both patients, and intracranial electroencephalographic recordings showed seizure origin from the cerebellar lesions. One patient eventually attained seizure control with antiepileptic drugs. The other case involved a child with generalized myoclonic epilepsy associated with a pilocytic astrocytoma of the cerebellum. This patient obtained seizure control following gross-total resection of the tumor. (http://thejns.org/doi/abs/10.3171/2013.5.PEDS1394) C Key Words • depth electrode cerebellar seizure • • epilepsy surgery lassic epilepsy teachings dictate that epilepsy arises only from the cerebral cortex, not subcortical structures like the cerebellum.10,12 As detailed by McCrory et al.,12 John Hughlings Jackson, who has been credited with developing the modern concept of epilepsy, observed and described cerebellar seizures in the late 1800s and early 1900s. His clinical concept of cerebellar seizures is thought to have been overshadowed by studies of decerebrate rigidity. Later, Penfield and Erickson postulated that “it is conceivable that some seizures characterized by stereotypic movements, or slow postural changes, are due to a cerebellar discharge … but we have never found clinical confirmation of such a conclusion.”15 Since then there have been a number of cases demonstrating that cerebellar seizures can occur.1–11,13,14,16,19 In this article we document 2 such patients. The first is a child with a hypodense lesion on MRI deep in the cerebellum resulting in complex partial seizures that were eventually medically controlled. The second case is a child who had generalized epilepsy associated with a pilocytic astrocytoma of the cerebellum. In both cases we used depth electrodes to monitor seizure activity from the lesions, verifying that these abnormalities were, in fact, the origin of the seizures. Case 1 Case Reports History and Examination. This 8-year-old boy preAbbreviations used in this paper: EEG = electroencephalography; LGG = low-grade glioma. 288 pilocytic astrocytoma • low-grade glioma • sented with intractable, symptomatic partial-onset seizures. He had no family history of epilepsy. His seizures started at 3 years of age and had previously been treated with oxcarbazepine, topiramate, valproic acid, clonazepam, phenytoin, levetiracetam, carbamazepine, pregabalin, and felbamate. At the time of examination he was experiencing approximately 20 complex partial seizures per month. Magnetic resonance imaging demonstrated a hypo­ dense, nonenhancing abnormality deep in the right cerebellum (Fig. 1). In 2 previous biopsies it was concluded that the abnormality was consistent with either gliosis or an LGG. His video-EEG monitoring with scalp electrodes showed no interictal epileptiform activity. Multiple simple and complex partial seizures were recorded in this patient. Electrographically, there were no changes seen with the simple partial seizures. His complex partial seizures lateralized to the left hemisphere and his clinical features had a consistent semiology with right body motor activity. Magnetoencephalography and magnetic source imaging were ordered to aid in the localization of spontaneous epileptiform discharges, receptive language areas, and somatosensory cortex prior to surgery. During the magnetoencephalography the patient had frequent epileptiform discharges that localized to the left temporal and frontal lobes. Language localization showed bilateral activation. Procedure. Based on these data, the decision was made to perform a repeat biopsy and insert a 4-contact depth electrode within the lesion for ictal clarification. This article contains some figures that are displayed in color on­line but in black-and-white in the print edition. J Neurosurg: Pediatrics / Volume 12 / September 2013 Two cases of cerebellar seizures Fig. 1. Case 1. Axial T1-weighted noncontrasted images demonstrating the cerebellar expansile lesion with a 4-contact depth electrode placed into the stereotactic biopsy tract for ictal recording. Note the hyperintense signal from the previous biopsy tract as well. The biopsy and electrode placement were performed using MRI-guided frameless stereotaxy, without complication (Fig. 1 right). Outcome. Monitoring from the depth electrode indi- cated rare spikes arising from the right cerebellar lesion. The patient had 2 secondarily generalized tonic-clonic seizures characterized by turning from his side to a supine position with flexing of both legs and the right arm. He had subtle clonic jerks of the right arm followed by postictal Todd paresis. Onset of these events was clearly recorded by the cerebellar depth electrode. The EEG recording showed onset with rhythmic theta frequencies from the right cerebellar depth electrode in contacts 1–4 (Fig. 2). After the event was over, he had attenuation in all 4 contacts. Although the seizures were found to be arising from the cerebellar lesion, the family decided to continue with medication management because of their concern for possible postoperative deficits. Postoperatively, the patient’s seizures have been fairly well controlled on lamotrigine and carbamazepine. He has 1 brief complex partial seizure (approximately 20–30 seconds in duration) every 3–4 months, and has gone as long as 1 year free of seizures. Follow-up MRI studies have shown that the lesion has remained stable for 4 years. At the moment no further surgery has been entertained. His third biopsy result was consistent with gliosis, but LGG remains a concern. Case 2 History and Examination. This patient was a 3-yearold boy with no family history of seizures or brain tumors who presented with right hand tremor at the age of 4 years. Neuroimaging showed a right cerebellar mass. He underwent partial resection at another institution, which revealed a Grade I cerebellar pilocytic astrocytoma. He did well for 1 year, and then had onset of generalized seizures. The initial event was a generalized tonic-clonic seizure, but he then developed refractory generalized myoclonic seizures, accompanied by vocalization; he had more than 20 events per hour. These episodes were initially not thought to be related to the cerebellar tumor and did not J Neurosurg: Pediatrics / Volume 12 / September 2013 Fig. 2. Case 1. This EEG reading demonstrates the onset of the cerebellar seizure as recorded from the 4-contact depth electrode. The first 3 channels are bipolar recordings linking electrodes 1 to 2, 2 to 3, and 3 to 4. The second 4 channels are referential recordings linked to a surface electrode. Note the change in background rhythm to a rhythmic high-voltage theta at the onset of the seizure. respond to medical treatment with zonisamide, lamotrigine, phenobarbital, diazepam, valproic acid, levetiracetam, phenytoin, clonazepam, lacosamide, oxcarbazepine, or topiramate. Phase I evaluation was performed, and MRI showed a residual enhancing mass in the right cerebellum near the midline on the superior aspect of his prior surgery site (Fig. 3A and B). The patient underwent prolonged videoEEG monitoring with scalp electrodes. The interictal EEG study showed generalized spike and wave complexes, and he had multiple generalized myoclonic and myoclonictonic seizures with generalized epileptiform activity seen on EEG recordings. The decision was made to proceed to a Phase II evaluation with placement of depth electrodes into the cerebellar tumor. Procedure. Frameless stereotactic MRI-guided placement of 2 posterior fossa depth electrodes was performed (Fig. 3C). A 6-contact depth electrode was placed in the superior right cerebellum in the lesion, and a 4-contact inferior cerebellar electrode was placed just inferior to the lesion. Outcome. After placement of the depth electrodes, the patient underwent prolonged video-EEG monitoring performed with his intracranial electrodes and scalp electrodes. Interictal EEG recordings showed essentially continuous spikes seen from the superior cerebellar depth electrode. Occasionally these spikes were associated with a generalized burst of spike and slow wave complex or bifrontal epileptiform discharges seen on the simultaneous scalp EEG studies (Fig. 4A). Many of the cerebellar discharges were not associated with any activity on the scalp EEG recording. During ictal events, the EEG study showed bilateral spike and slow wave complexes recorded by the depth electrode. The changes were recorded by the depth electrode 20–50 msec before changes were recorded on the scalp EEG study (Fig. 4B). After monitoring, the patient was taken back to surgery for tumor resection and depth electrode removal. Fol­ lowing resection, an intraoperative MRI study was 289 S. Boop et al. Fig. 3. Case 2. Sagittal (A) and coronal (B) T1-weighted Gd-enhanced images demonstrating residual partially enhancing tumor within the right cerebellar hemisphere and extending to the tentorial incisura. A postoperative anteroposterior skull x-ray (C) was obtained following Stage I surgery, demonstrating 2 depth electrodes in place, a 6-contact depth electrode at the superior aspect of the tumor, and a 4-contact depth electrode at the inferior aspect of the tumor. per­formed, showing a gross-total resection. Postoperatively, the patient experienced 3 typical seizures the night of surgery, but was then seizure free. He was discharged home on lamotrigine. As an outpatient, he later required addition of ethosuximide. At this time, he has been seizure free for the 1.5 years since surgery, and his interictal scalp EEG findings are now normal. Follow-up MRI studies have shown no evidence of residual or recurrent tumor (Fig. 5). Fig. 4. Case 2. A: Scalp-recorded EEG study showing generalized epileptiform bursts and a generalized burst followed by generalized attenuation associated clinically with a myoclonic seizure (arrow) characterized by a brief nod of the head and vocalization. B: An EEG recording from depth and scalp electrodes capturing a similar myoclonic seizure. The burst clearly begins near the depth electrode (arrow) and spreads over 20 msec to the rest of the electrodes. 290 Discussion Subcortical structures are classically thought to have only ancillary effects on cerebral seizure activity. However, a number of cases found in the literature indicate otherwise (Table 1). Epileptic activity of cerebellar origin is typically associated with cerebellar dysplasia or glio- Fig. 5. Case 2. A T1-weighted nonenhanced coronal MRI study obtained 1 year after resection of the pilocytic astrocytoma, at a point in time when the child had remained seizure free—but while he still needed treatment with anticonvulsant medications (lamotrigine and ethosuximide)—demonstrating postsurgical changes within the resection cavity with no residual tumor. J Neurosurg: Pediatrics / Volume 12 / September 2013 Two cases of cerebellar seizures TABLE 1: Literature review of cases published between 1976 and the present* Authors & Year Patient Age, Scalp EEG Sex Results Neuroimaging Findings Pathology Invasive Monitoring Sz Control Langston & Tharp, 1976 1.5 mos, M abnormal mass bulging into lt superior 4th vent ganglioglioma no Jayakar & Seshia, 1987 12 mos, M normal CT: mass at lt cerebellar hemisphere no Flüeler et al., 1990 1 wk, F 10 mos, F normal ND no no not controlled not controlled Bills & Hanieh, 1991 3 wks, M ND ganglioglioma no al-Shahwan et al., 1994 1 day, F normal MRI: mass at rt MCP & SCP MRI: mass extending into rt MCP & ICP & compressing 4th vent CT/MRI: mass at lt SCP compressing 4th vent MRI: mass at rt MCP & SCP low-grade astrocytoma none none PR; szs resumed at lower frequency PR ganglioglioma no Harvey et al., 1996 1 day, M normal ganglioglioma yes Chae et al., 2001 Delande et al., 2001 Mesiwala et al., 2002 4 mos, M 36 mos, F 36 mos, F 1 day, M normal abnormal normal abnormal ganglioglioma hamartoma hamartoma ganglioglioma yes yes no yes PR, no szs 3 mos out PR; szs resumed at lower frequency CR, no szs w/o meds PR, then CR resection resection PR, then CR Strazzer et al., 2006 32 mos, F abnormal no Kulkarni et al., 2007 Park et al., 2009 48 mos, M 3 mos, M abnormal ND pilocytic astrocytoma ganglioglioma hamartoma no yes PR + AEDs, but szs persist AEDs PR, then CR Hanai et al., 2010 Yagyu et al., 2011 7 mos, M 20 mos, M normal abnormal ganglioglioma gangliocytoma no yes CR PR present study 96 mos, M 36 mos, M abnormal abnormal gliosis or LGG pilocytic astrocytoma yes yes AEDs PR, then CR MRI: mass at lt MCP & SCP effacing 4th vent lesion, lt SCP lesion, floor 4th vent lesion, floor 4th vent MRI: mass at lt MCP & SCP displacing 4th vent lesion, rt deep median-paramedian cerebellar peduncle lesion, rt cerebellum/4th vent lesion, floor 4th vent w/ extended cerebellar peduncle & hemisphere MRI: lesion, lt MCP MRI: dysplastic neuronal lesion, floor 4th vent lesion, rt deep cerebellum lesion, rt medial cerebellar hemisphere * AEDs = antiepileptic drugs; CR = complete resection; ICP = inferior cerebellar peduncle; MCP = middle cerebellar peduncle; ND = not done; PR = partial resection; SCP = superior cerebellar peduncle; Sz = seizure; vent = ventricle. neuronal tumors. It is usually seen in young children and is rarely controlled with antiepileptic drugs. Our first case report presents a child with either gliosis or an LGG. The EEG recordings from the depth electrodes placed within the lesion showed that the tumor was the origin of his seizure activity. Concerns by the family that surgery presented the possibility of new neurological deficit made us pursue medical therapy rather than further surgery. This case is unique because it is the only case in the literature describing cerebellar seizures originating from a suspected LGG that is controlled by medical therapy without resection. Most of the literature suggests that complete resection of cerebellar glioneuronal tumors is necessary for seizure control. There is 1 case of cerebellar seizures stemming from a low-grade astrocytoma that was reported in 1987 by Jayakar and Seshia.8 This case reported that seizures were controlled after only partial resection. In 2006, Strazzer et al.16 reported a case of a child who underwent resection of a pilocytic astrocytoma at 32 months of age for cerebellar J Neurosurg: Pediatrics / Volume 12 / September 2013 seizures. This resulted in reduction of seizures with periods of seizure control, although initiation of anticonvulsant therapy was necessary. Follow-up MRI studies in this patient revealed a small area of residual tumor. Our second patient with pilocytic astrocytoma is an interesting demonstration of a focal lesion causing generalized epilepsy. During monitoring, the patient’s scalp EEG recordings demonstrated generalized epileptiform discharges interictally, and also in association with the patient’s typical seizures, which were generalized myoclonic seizures. Although scalp EEG changes were noted, the EEG recording from the depth electrodes clearly demonstrated that the lesion was the origin of the epileptic discharges. That, coupled with the fact that the seizure activity resolved after complete resection of the tumor, suggests that the lesion itself was the cause of the epileptic activity. Focal lesions causing generalized epilepsy syndromes and generalized EEG changes have been described previously. The original cases noted were infants with congenital focal lesions and infantile spasms associated with a hypsar291 S. Boop et al. rhythmic EEG background.9,17 More recently, older children and young adults with congenital or early acquired cortical lesions and generalized EEG findings have been shown to be surgical candidates in some instances.18 Our report is unusual in demonstrating that focal cerebellar lesions can also lead to generalized EEG abnormalities and even a seizure semiology, which appears to be generalized in nature. A similar case of myoclonic cerebellar seizures associated with a pilocytic astrocytoma has been reported, but depth electrode monitoring was not used to prove that the lesion was the origin of the seizure activity.16 The patient reported on by Strazzer et al. did have a reduction of seizures following subtotal tumor resection. Conclusions These case reports, in conjunction with the reviewed literature, lead to the conclusion that in patients with cerebellar structural lesions and seizures a causative relationship should be considered. We have demonstrated, as have others, that surface EEG recordings may not be diagnostic, and invasive monitoring is necessary to confirm that a cerebellar lesion is the origin of the epilepsy. Findings on PET or SPECT studies may also be supportive. Medications and partial resections are often ineffective in controlling the seizures associated with cerebellar lesions, whereas complete resection of the lesion is generally curative. As shown in our first case, medical therapy may be effective when surgical options could lead to further neurological deficit; however, this is distinctly uncommon. Disclosure The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper. Author contributions to the study and manuscript preparation include the following. Conception and design: FA Boop, Wheless. Acquisition of data: S Boop. Drafting the article: S Boop, Van Poppel. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: FA Boop. References 1. al-Shahwan SA, Singh B, Riela AR, Roach ES: Hemisomatic spasms in children. Neurology 44:1332–1333, 1994 2. Bills DC, Hanieh A: Hemifacial spasm in an infant due to fourth ventricular ganglioglioma. Case report. J Neurosurg 75:134–137, 1991 3. Chae JH, Kim SK, Wang KC, Kim KJ, Hwang YS, Cho BK: Hemifacial seizure of cerebellar ganglioglioma origin: seizure control by tumor resection. Epilepsia 42:1204–1207, 2001 4. Delande O, Rodriguez D, Chiron C, Fohlen M: Successful sur- 292 gical relief of seizures associated with hamartoma of the floor of the fourth ventricle in children: report of two cases. Neurosurgery 49:726–731, 2001 5. Flüeler U, Taylor D, Hing S, Kendall B, Finn JP, Brett E: Hemifacial spasm in infancy. Arch Ophthalmol 108:812–815, 1990 6. Hanai S, Okazaki K, Fujikawa Y, Nakagawa E, Sugai K, Sasaki M, et al: Hemifacial seizures due to ganglioglioma of cerebellum. Brain Dev 32:499–501, 2010 7. Harvey AS, Jayakar P, Duchowny M, Resnick T, Prats A, Altman N, et al: Hemifacial seizures and cerebellar ganglioglioma: an epilepsy syndrome of infancy with seizures of cerebellar origin. Ann Neurol 40:91–98, 1996 8. Jayakar PB, Seshia SS: Involuntary movements with cerebellar tumour. Can J Neurol Sci 14:306–308, 1987 9. Kramer U, Sue WC, Mikati MA: Focal features in West syndrome indicating candidacy for surgery. Pediatr Neurol 16: 213–217, 1997 10. Kulkarni S, Hegde A, Shah KN: Hemifacial seizures and cerebellar tumor: a rare co-existence. Indian Pediatr 44:378–379, 2007 11. Langston JW, Tharp BR: Infantile hemifacial spasm. Arch Neurol 33:302–303, 1976 12. McCrory PR, Bladin PF, Berkovic SF: The cerebellar seizures of Hughlings Jackson. Neurology 52:1888–1890, 1999 13. Mesiwala AH, Kuratani JD, Avellino AM, Roberts TS, Sotero MA, Ellenbogen RG: Focal motor seizures with secondary generalization arising in the cerebellum. Case report and review of the literature. J Neurosurg 97:190–196, 2002 14. Park YS, Oh MC, Kim HD, Kim DS: Early surgery of hamartoma of the floor of the fourth ventricle: a case report. Brain Dev 31:347–351, 2009 15. Penfield W, Erickson TC: Epilepsy and Cerebral Localization. Springfield, IL: Charles Thomas, 1941 16. Strazzer S, Zucca C, Fiocchi I, Genitori L, Castelli E: Epilepsy and neuropsychologic deficit in a child with cerebellar astrocytoma. J Child Neurol 21:817–820, 2006 17. Wyllie E, Comair YG, Kotagal P, Raja S, Ruggieri P: Epilepsy surgery in infants. Epilepsia 37:625–637, 1996 18. Wyllie E, Lachhwani DK, Gupta A, Chirla A, Cosmo G, Worley S, et al: Successful surgery for epilepsy due to early brain lesions despite generalized EEG findings. Neurology 69:389– 397, 2007 19. Yagyu K, Sueda K, Shiraishi H, Asahina N, Sakurai K, Kohsaka S, et al: Direct correlation between the facial nerve nucleus and hemifacial seizures associated with a gangliocytoma of the floor of the fourth ventricle: a case report. Epilepsia 52: e204–e206, 2011 Manuscript submitted February 25, 2013. Accepted May 30, 2013. Portions of this work were presented in poster form at the annual meeting of the Amer­ican Epilepsy Society in San Diego, California, on December 2, 2012. Please include this information when citing this paper: published online June 28, 2013; DOI: 10.3171/.2013.5.PEDS1394 Address correspondence to: Frederick A. Boop, M.D., SemmesMurphey Clinic, 6325 Humphreys Blvd., Memphis, TN 38120. email: FABoop@aol.com. J Neurosurg: Pediatrics / Volume 12 / September 2013