- Acquired epilepsies can occur as a result of an initiating event such as a brain injury which may include hypoxia that results as a part of cardiac arrest and coma. Currently, it is not possible to predict who will develop seizures after an initial brain injury.
- Studying brain activity to identify biological signals (biomarkers) may help predict who will develop epilepsy following a brain injury. Yet, deciphering changes in brain activity following an injury remains a significant challenge in the field and thus is a critical area of focus for CURE Epilepsy.
- Dr. Edilberto Amorim at the University of California San Francisco received a Taking Flight Award in 2020; his research focuses on patients who experienced a coma after cardiac arrest, a population in which brain activity is routinely monitored via electroencephalography (EEG) immediately after the cardiac event.
- Dr. Amorim’s team identified specific brain activity patterns based on EEG that correlate with good recovery of patients after coma.
- The work of Dr. Amorim and his team has helped to demonstrate the utility of specific brain activity biomarkers in predicting patient outcomes following cardiac arrest. This work provides insights into changes in brain activity that may be extended to people at risk of developing epilepsy after a brain injury.
Acquired epilepsies can result from brain injury such as head trauma or a lack of oxygen (hypoxia) which can occur following a heart attack or cardiac arrest. Often, there is a period of time between the initial injury and the onset of seizures referred to as the “latent” period. During this time, individuals do not experience seizure activity, but a process called epileptogenesis may be at play to change brain activity, making it more hyperexcitable and prone to seizures. This subset of individuals begins to experience spontaneous seizures and develop epilepsy. Knowledge about changes in the brain after the initial injury may provide clues about ways to prevent seizures. A biomarker is a biological factor such as a protein in the blood or brain electrical activity that can be objectively measured and can act as an indicator or even predictor of a normal or an abnormal condition. In epilepsy, a biomarker could provide information on who is at the highest risk of developing seizures following a brain injury.
Research on biomarkers of acquired epilepsies is challenging, however, due to the inherent complexity of different types of epilepsy, the variability that exists between people at risk for epilepsy, and the challenges of monitoring people acutely after an injury that puts them at risk for epilepsy. Knowing the impact that finding biomarkers for the epilepsies could have on patients and families, CURE Epilepsy funds research in this area. Dr. Edilberto Amorim at the University of California, San Francisco practices at the Zuckerberg San Francisco General Hospital, and received a Taking Flight Award in 2020. Taking Flight Awards fund investigators in the field of the epilepsies relatively early in their careers to enable them to develop a research focus and team independent of their mentor(s). Dr. Amorim’s research involves studying brain activity using electroencephalography (EEG) to predict health outcomes in patients who are critically ill. The goal of his research is to develop data-driven approaches to develop personalized therapies. His work builds on previous studies that have identified several EEG patterns that may serve as biomarkers for the epilepsies.
Dr. Amorim’s approach, however, is unique as he studies individuals who are comatose after a cardiac arrest. Since brain activity in these individuals is monitored with EEG soon after the injury, this is an opportune time to study EEG patterns to better understand the evolution of brain activity after an initial injury.[4, 5] It is known that brain activity rapidly change in the first few hours and days following injury ; however, while some individuals with abnormal brain activity do not recover from a coma, others do. Dr. Amorim hypothesized that there are clues in the EEG that will correspond to how well a patient recovers after a coma following cardiac arrest. In earlier, smaller studies, his team used machine learning algorithms and found that EEG patterns change over time after a cardiac arrest and that they may be reflective of the functional status of the patient. Hence, they may be used to accurately predict recovery after a cardiac arrest.[7, 8]
Dr. Amorim’s team conducted a multi-center, international study of 1,038 patients who experienced cardiac arrest and coma and collected more than 50,000 hours of EEG data. By developing machine learning algorithms that can identify patterns from a large amount of data, he focused on three domains of EEG activity:
- Burst suppression ratio, or a pattern of EEG where extremely high-voltage electrical activity is followed by periods of no activity
- Spike frequency, with spikes being discrete events in the EEG signal
- Shannon entropy, which is a measure of uncertainty of a certain event or pattern used to characterize complex processes
To correlate EEG patterns to functional recovery, the team looked at cerebral performance category (CPC), a validated way of categorizing the neurological state of an individual after a cardiac arrest. Analysis of the data showed that certain patterns of EEG (namely lower burst suppression ratio, lower spike frequency, and higher levels of entropy) were associated with better functional outcomes, e.g., independence for activities of daily living. Out of these, high entropy states were most informative, as they were associated with better outcomes even in the presence of less-than-ideal spike frequency and burst suppression rates. This suggests that high entropy states may reflect resilience against brain injury and may herald good recovery following coma. The timing of the appearance of certain patterns in EEG was also informative. People with good recovery from coma had earlier and larger improvements in burst suppression ratio and entropy as compared to those with poor recovery. Hence, the different types of EEG patterns that occur early after a cardiac arrest and coma carry important information about the potential to recover from these health crises.
Dr. Amorim’s team has also launched a data challenge competition called the PhysioNet Challenge 2023 to further the field of prediction of outcomes in cardiac arrest, which given the connection to brain function has implications for epilepsy. It is one of the largest disease-specific EEG databases, with more than 50,000 hours of continuous EEG data and is open-access to anyone interested in using the data. Almost 100 teams from all over the world are participating in the competition; more details about this competition can be found here.
By looking at EEG patterns to predict recovery, Dr. Amorim’s work adds to the existing knowledge about what happens in the brain following an injury. His work provides insights on important parameters of brain function to assess immediately following an injury that may be extended to the study of epileptogenesis, the processes that contribute to the generation of seizures. Dr. Amorim’s work gives us clues as to how detailed and precise EEG analysis after cardiac arrest can provide insights into the mechanisms that make a brain susceptible to seizures and acquired epilepsies. A better understanding of what happens in the brain during the latent period after a brain injury brings us closer to developing preventative strategies and personalized treatments for those at risk of developing an acquired epilepsy.
- Shorvon SD. The etiologic classification of epilepsy Epilepsia. 2011 Jun;52:1052-1057.
- Simonato M, Agoston DV, Brooks-Kayal A, Dulla C, Fureman B, Henshall DC, et al. Identification of clinically relevant biomarkers of epileptogenesis – a strategic roadmap. Nat Rev Neurol. 2021;17:231-242.
- Gallotto S, Seeck M. EEG biomarker candidates for the identification of epilepsy Clin Neurophysiol Pract. 2023;8:32-41.
- Amorim E, Rittenberger JC, Zheng JJ, Westover MB, Baldwin ME, Callaway CW, et al. Continuous EEG monitoring enhances multimodal outcome prediction in hypoxic-ischemic brain injury Resuscitation. 2016 Dec;109:121-126.
- Khazanova D, Douglas VC, Amorim E. A matter of timing: EEG monitoring for neurological prognostication after cardiac arrest in the era of targeted temperature management Minerva Anestesiol. 2021 Jun;87:704-713.
- Hofmeijer J, Beernink TM, Bosch FH, Beishuizen A, Tjepkema-Cloostermans MC, van Putten MJ. Early EEG contributes to multimodal outcome prediction of postanoxic coma Neurology. 2015 Jul 14;85:137-143.
- Ghassemi MM, Amorim E, Alhanai T, Lee JW, Herman ST, Sivaraju A, et al. Quantitative Electroencephalogram Trends Predict Recovery in Hypoxic-Ischemic Encephalopathy Crit Care Med. 2019 Oct;47:1416-1423.
- Amorim E, van der Stoel M, Nagaraj SB, Ghassemi MM, Jing J, O’Reilly UM, et al. Quantitative EEG reactivity and machine learning for prognostication in hypoxic-ischemic brain injury Clin Neurophysiol. 2019 Oct;130:1908-1916.
- Amorim E, Zheng WL, Jing J, Ghassemi MM, Lee JW, Wu O, et al. Neurophysiology State Dynamics Underlying Acute Neurologic Recovery After Cardiac Arrest Neurology. 2023 Aug 29;101:e940-e952.
- Hsu CH, Li J, Cinousis MJ, Sheak KR, Gaieski DF, Abella BS, et al. Cerebral performance category at hospital discharge predicts long-term survival of cardiac arrest survivors receiving targeted temperature management* Crit Care Med. 2014 Dec;42:2575-2581.
- 11. Amorim E, Zheng, W., Lee, J. W., Herman, S., Ghassemi, M., Sivaraju, A., Gaspard, N., Hofmeijer, J., van Putten, M. J. A. M., Reyna, M., Clifford, G., & Westover, B. I-CARE: International Cardiac Arrest REsearch consortium Database (version 2.0). PhysioNet. 2023.