Unihemispheric Burst Suppression Nisms.2 Current Clinical Practice Is Based on the Interpretation of BS As a Non-Epileptic Phe- Correspondence: Edward C

Neurology International 2014; volume 6:5487

Unihemispheric burst suppression nisms.2 Current clinical practice is based on the interpretation of BS as a non-epileptic phe- Correspondence: Edward C. Mader, Jr., Epilepsy nomenon. Center of Excellence, Louisiana State University Edward C. Mader Jr., Nicole R. Bihemispheric BS has occasionally been School of Medicine, 1542 Tulane Avenue, New Villemarette-Pittman, Cornel T. Rogers, observed to be asymmetric and/or asynchro- Orleans, LA 70112, USA. Frank Torres-Delgado, Piotr W. nous. In patients with unihemispheric lesions Tel.: +1.504.568.4080 - Fax: +1.504.617.60974. Olejniczak, John D. England (e.g. hemimegalencephaly) asymmetric BS E-mail: [email protected] Epilepsy Center of Excellence, Louisiana may occur spontaneously and disappear in Key words: burst suppression, unihemispheric, State University Health Sciences Center, infancy or it may persist into adult life.3 electroencephalogram, seizure, status epilepti- New Orleans, LA, USA Anesthesia-induced asymmetric and asynchro- cus, propofol. nous BS in patients with lesions of the corpus callosum indicates that this structure plays a Acknowledgments: the authors are grateful to crucial role in interhemispheric synchroniza- their electroencephalogram technologists Abstract tion of normal and abnormal cortical electrical Thomas B. Miller, Terri L. Ware, and Sheryl A. activity.4,5 On the other hand, true unihemi- Wagamotte. Burst suppression (BS) consists of bursts of spheric BS is rare and its clinical significance high-voltage slow and sharp wave activity is poorly understood. In this paper, we describe Contributions: ECM, CTR, FTD, direct patient care, EEG interpretation, and preparation of man- alternating with periods of background sup- two patients with unihemispheric BS while being treated with propofol for epileptic uscript; NRVP, PWO, JDE, literature search, pression in the electroencephalogram (EEG). review, editing, and finalizing manuscript. When induced by deep anesthesia or seizures. encephalopathy, BS is bihemispheric and is Conflict of interests: the authors declare no often viewed as a non-epileptic phenomenon. potential conflict of interests. In contrast, unihemispheric BS is rare and its clinical significance is poorly understood. We Case Report #1 Received for publication: 21 May 2014. describe here two cases of unihemispheric BS. Accepted for publication: 7 July 2014. The first patient is a 56-year-old woman with a Patient A is a 56-year-old woman with a left This work is licensed under a Creative Commons temporoparietal solitary metastatic brain left temporoparietal tumor who presented in Attribution NonCommercial 3.0 License (CC BY- convulsive status epilepticus. EEG showed left lesion (s/p resection) who presented in convul- NC 3.0). hemispheric BS after clinical seizure termina- sive status epilepticus. She was taking leve- tion with lorazepam and propofol. The second tiracetam 1000 mg/day and phenytoin 450 ©Copyright E.C. Mader Jr. et al., 2014 patient is a 39-year-old woman with multiple mg/day at home. Clinical seizure termination Licensee PAGEPress, Italy medical problems and a vague history of was achieved with fosphenytoin 1000 mg, Neurology International 2014; 6:5487 seizures. After abdominal surgery, she experi- lorazepam 4 mg, and propofol 2 mg/kg load doi:10.4081/ni.2014.5487 enced a co nvulsive seizure prompting treat- then 2 mg/kg/hr. EEG recording 45 minutes ment with propofol. Her EEG also showed left after seizure termination showed unihemi- BS on the left but not on the right (Figure 2). hemispheric BS. In both cases, increasing the spheric BS on the left and background slowing EEG monitoring continued to demonstrate left propofol infusion rate resulted in disappear- on the right (Figure 1). Unihemispheric BS unihemispheric BS with propofol infusion at 2 ance of unihemispheric BS and clinical persisted when the propofol infusion rate was mg/kg/hr. Unihemispheric BS disappeared 2 improvement. The prevailing view that typical 2 mg/kg/hr. Its disappearance 4 hours later hours later at a time when the propofol drip bihemispheric BS is non-epileptic should not coincided with a propofol drip rate of 5 be extrapolated automatically to unihemi- mg/kg/hr. Propofol was discontinued after 12 rate was 5 mg/kg/hr. Propofol was discontinued spheric BS. The fact that unihemispheric BS hours with no recurrence of seizure or uni- after 18 hours with no recurrence of seizure was associated with clinical seizure and hemispheric BS. Brain magnetic resonance and unihemispheric BS. Brain computed resolved with propofol suggests that, in both imaging showed left anterior parietal tomography scan revealed left hemispheric cases, an epileptic mechanism was responsi- encephalomalacia with surrounding edema encephalomalacia with predominant involve- ble for unihemispheric BS. and mass effect prompting treatment with dex- ment of the left posterior parietal lobe. Follow- amethasone. Prior to discharge, follow-up EEG up EEG showed complete resolution of uni- showed left hemispheric slowing with no hemispheric BS and normalization of the EEG. recurrence of unihemispheric BS. Introduction

Burst suppression (BS) is an abnormal pat- Discussion tern in the electroencephalogram (EEG) where Case Report #2 bursts of high-voltage slow waves and sharp Clinically, BS is almost always bihemispher- waves alternate with periods of depressed Patient B is a 39-year-old woman whose past ic and is often bilaterally synchronous and background activity.1 The bursts are generally seizures were attributed to alcohol withdrawal; symmetric. As in the two cases described in bihemispheric and more or less bisynchronous she was not taking antiepileptic drugs at this paper, BS can also be restricted to one and bisymmetric. Bihemispheric BS is invari- home. She had a generalized tonic-clonic hemisphere. However, the dearth of published ably associated with coma, usually in the set- seizure after undergoing surgery for a bleed- data on unihemispheric BS suggests that this ting of deep anesthesia, drug intoxication, ing duodenal ulcer. Clinical seizure termina- entity is rare, frequently ignored, or poorly hypothermia, and cerebral anoxia. Although tion was achieved with intravenous levetirac- understood. The pathophysiological basis of BS has been studied extensively at the EEG etam 1000 mg and propofol 2 mg/kg load then BS is not well understood. Two mechanisms level, only sparse information is available in 2 mg/kg/hr. EEG recording 30 minutes after have been implicated through research in this regards to its neurophysiological mecha- seizure termination showed unihemispheric area: cerebral hypometabolism and thalamo-

[Neurology International 2014; 6:5487] [page 53] Case Report

Figure 1. Electroencephalogram of Patient A showing left-sided Figure 2. Electroencephalogram of Patient B showing left-sided burst suppression 45 minutes after status epilepticus termination burst suppression 30 minutes after clinical seizure termination with fosphenytoin 1000 mg, lorazepam 4 mg, and propofol 2 with intravenous levetiracetam 1000 mg and propofol 2 mg/kg mg/kg load. The occasional right-sided sharp waves that stand load. Fragments of burst activity spilling over the right hemi- out against the low-amplitude slow background activity are most sphere are most likely due to volume conduction from the left likely the result of volume conduction from the left hemisphere hemisphere source. Unihemispheric burst suppression was pres- source. Unihemispheric burst suppre ssion persisted as propofol ent during propofol infusion at 2 mg/kg/hr. It disappeared 2 was being infused at 2 mg/kg/hr and disappeared 4 hours later hours later when the propofol infusion rate was 5 mg/kg/hr. when the infusion rate was 5 mg/kg/hr. Propofol was discontin- Propofol was discontinued after 18 hours with no recurrence of ued after 12 hours with no recurrence of seizure activity and uni- seizure and unihemispheric burst suppression. hemispheric burst suppression.

cortical hyperexcitability. citability has also been verified by intracranial that the anticonvulsant propofol can act as a The hypometabolism theory considers brain recordings during anesthesia with BS-induc- proconvulsant depending on infusion rate and hypometabolism during EEG suppression as ing barbiturate methohexital.14 individual susceptibility.17-19 An infusion rate the fundamental physiological disturbance in Traditionally, BS has been viewed as a glob- of 2-3 mg/kg/hr may have facilitated, and a rate BS.6 The burst of synchronous activity is sim- al state with widespread synchronous and of 4-5 mg/kg/hr may have suppressed, uni- ply viewed as a reaction of the brain to prevent homogenous cortical activation during the hemispheric BS in both patients. The occur- membrane potential collapse during the burst episodes and inactivation duri ng the rence of unihemispheric BS on the same side hypometabolic state. This theory is consistent periods of suppression.15 This assumption is of a focal lesion after termination of a clinical with the fact that BS can be induced by a vari- the basis for the clinical induction of BS to seizure and its subsequent suppression with ety of conditions that re duce cerebral metabol- reduce global brain activity. In reality, the propofol alludes to cortical hyperexcitability as ic rate, including hypothermia,7 hypoxic- extent and spatial homogeneity of BS has not the primary mechanism underlying unihemi- ischemic encephalopathy,8 and general anes- yet been fully explored because of the difficul- spheric BS. It also supports th e notion that BS thesia.9,10 Moreover, classical BS is consistent- ty of recording multiple cortical sites simulta- is a product of abnormal local thalamocortical ly induced with gamma-aminobutyric acid neously. Thus, the dynamics of large-scale cor- dynamics and that different networks have dif- enhancing drugs that significantly reduce tical circuits during BS are not well under- ferent thresholds for generating BS. It is cerebral metabolic rate, but not with ketamine stood. There is mounting evidence that BS is a provocative to think that unihemispheric BS and other anesthetic agents that do not local phenomenon. For example, intracranial shares some of the basic mechanisms of peri- decrease cerebral metabolic rate at clinically EEG from patients under propofol anesthesia odic lateralized epileptiform discharges and relevant doses.10 showed that BS can be asynchronous between other periodic EEG discharges.20 It is not clear The hyperexcitability theory emphasizes cortical regions and can occur in limited corti- whether the above arguments about unihemi- abnormal cortical and thalamic bursting activ- cal areas while other areas exhibit continuous spheric BS can be generalized to bihemispher- ity as the basic mechanism driving BS.11 activity.16 This implies that local cortical ic BS. Nevertheless, the most important ques- Intracellular recording of cortical neurons dur- dynamics are not homogeneous even during tion is not really whether cerebral hypometab- ing the EEG burst revealed the presence of significant brain inactivation. It has been sug- olism or cortical hyperexcitability is the funda- depolarizing synaptic potentials with a crown gested that cortical and subcor tical circuits mental mechanism in BS. The two mecha- of action potentials reminiscent of the parox- have different sensitivities to anesthesia lev- nisms are not mutually exclusive and there is ysmal depolarization shift during interictal els and that such a hierarchy governs how the some evidence that both are present in spikes. The absence of cortical synaptic activi- brain enters BS during anesthesia.16 bihemispheric BS. It is also highly probable ty during EEG suppression has been attributed In both of our patients, unihemispheric BS that both mechanisms are fundamentally to extracellular calcium depletion from prior was detected 30-45 minutes after clinical essential to generate BS of any type. Whats i bursting activity.12 The cycle repeats as calci- seizure termination with propofol and other more important to know is whether BS-related um balance is restored by neuron pump activi- anticonvulsants, persisted for 2-4 hours during thalamocortical overactivity constitutes an ty. Earlier studies also showed 1-4 Hz rhythmic propofol infusion at a rate of 2-3 mg/kg/hr, and epileptic phenomenon and whether BS-related thalamic oscillations during cortical silence disappeared when the rate was increased to 4- bursting activity is associated with restricted periods and a propensity for thalamocortical 5 mg/kg/hr. It is conceivable that unihemi- hypermetabolism exposing certain brain struc- networks to produce burst activity even with spheric BS occurred because of pre-existing tures to the risk of additional brain injury. We subliminal stimuli.13 The fact that ictal-like cortical hyperexcitability. Patient A had a hope that researchers will give us the answers motor activity, such as myoclonus, may occur metastatic tumor resected from the left tem- to these questions in the near future. during EEG burst episodes lend support to the poroparietal region and Patient B was found to hyperexcitability theory. Cortical hyperex- have a left parietal lobe lesion. It is also known

[page 54] [Neurology International 2014; 6:5487] Case Report

Asyn chronous and asymmetric burst-sup- 12. Amzica F. Basic p hysiology of burst-sup- Conclusions pression in a patient with a corpus callo- pression. Epilepsia 2009;50:38-9. sum lesion. Clin Neurophysiol 13. Kroeger D, Amzica F. Hypersensitivity of Burst suppression is typically bihemispheric 1999;110:103-5. the anesthesia-induced comatose brain. J and often bilaterally synchronous and symmet- 5. Lazar LM, Milrod LM, Solomon GE, Labar Neurosci 2007;27:10597-607. ric. Unihemispheric BS is rare but, as in the DR. Asynchronous pentobarbital-induced 14. Wennberg R, Quesney F, Olivier A, Dubeau two cases presented, it can occur in associa- burst suppression with corpus callosum F. Induction of burst-suppression and acti- tion with seizures, propofol anesthesia, or hemorrhage. Clin Neurophysiol vation of epileptiform activity after metho- both. It is not clear how unihemispheric BS is 1999;110:1036-40. hexital and selective amygdalo-hippocam- related to bihemispheric BS. The occurrence of 6. Ching S, Purdon PL, Vijayan S, et al. A neu- pectomy. Electroencephalogr Clin unihemispheric BS with seizures and propofol rophysiological-metabolic model for burst Neurophysiol 1997;102:443-51. anesthesia suggests that an epileptic mecha- suppression. Proc Natl Acad Sci USA 15.C lark DL, Rosner BS. Neurophysiologic nism is responsible for its generation. 2012;109:3095-100. effects of general anesthetics. I. The elec- Whether an epileptic mechanism exposes cer- 7. Edgren E, Enblad P, Grenvik A, et al. troencephalogram and sensory evoked tain brain structures to the risk of additional Cerebral blood flow and metabolism after responses in man. Anesthesiology brain injury during unihemispheric BS, cardiopulmonary resuscitation A patho- 1973;38:564-82. bihemispheric BS, or both is an important physiologic and prognostic positron emis- 16. Lewis LD, Ching S, Weiner VS, et al. Local question that must be answered by future sion tomography pilot study. Resuscitation cortical dynamics of burst suppression in research. 2003;57:161-70. the anaesthetized brain. Brain 2013;136: 8. Levy WJ, Pantin E, Mehta S, McGarvey M. 2727-37. Hypothermia and the approximate entropy 17. Walder B, Tramer MR, Seeck M. Seizure of the electroencephalogram. like phenomenon and propofol: a system- References Anesthesiology 2003;98:53-7. atic review. Neurology 2002;58:1327-32. 9. Hirsch N, Taylor C. Pharmacological and 18. San-juan D, Chiappa KH, Cole AJ. Propofol 1. Schaul N. The fundamental neural mecha- pathological modulation of cerebral physi- and the electroencephalogram. Clin nisms of electroencephalography. ology. Anaesth Intens Care Med 2010;11: Neurophysiol 2010;121:998-1006. Electroencephalogr Clin Neurophysiol 349-54. 19. Cochran D, Price W, Gwinnutt CL. 1998;106:101-7. 10. Barash PG, ed. Clinical anesthesia. 6th ed. Unilateral convulsion after induction of 2. Niedermeyer E. The burst-suppression Philadelphia: Lippincott Williams & anaesthesia with propofol. Br J Anaesth electroencephalogram. Am J Electroneu- Wilkins; 2009. 1996;76:570-2 rodiagnostic Technol 2009;49:333-41. 11. Steriade M, Amzica F, Contreras D. 20. Kalamangalam GP, Diehl B, Burgess RC. 3. Ohtsuka Y, Ohno S, Oka E. Electroclinical Cortical and thalamic cellular correlates of Neuroimaging and neurophysiology of characteristics of hemimegalencephaly. electroencephalographic burst-suppres- periodic lateralized epileptiform dis- Pediatr Neurol 1999;20:390-3. sion. Electroencephalogr Clin Neuro - charges: observations and hypotheses. 4. Lambrakis CC, Lancman ME, Romano C. physiol 1994;90:1-16. Epilepsia 2007;48:1396-405.

[Neurology International 2014; 6:5487] [page 55]