• Cecilie Risager Kennedy
5. Term (Master thesis), Medicine, Master (Master Programme)
Introduction: An epileptic seizure is a transient state caused by hypersynchronous or excessive neuronal activity usually self-limiting within minutes, but sometimes persistent and needing medical intervention. Diagnosing epilepsy is often challenging as it is a clinical diagnosis largely based on information given by random people present at the times of seizures and supplements to the existing diagnostic methods are needed. An important differential diagnosis is the functional disorder, PNES (psychogenic nonepileptic seizures), with symptoms that can easily be mistaken for an epileptic seizure, but not involving changes in the electrical activity in the brain. Delay and imprecision in the diagnosis of epilepsy and PNES can potentially bring serious and life-threatening consequences for the patient and economic consequences for society. In some cases of epileptic seizures transient abnormalities on MRI (magnetic resonance imaging) can be seen, which are considered to be caused by oedema due to the metabolic processes during epileptic activity. The purpose of the current study is to investigate the occurrence and characteristics of these postictal abnormalities on MRI in patients hospitalized due to epileptic seizures or PNES. One of the points of discussion is if better use of MRI in the early phase of seizures will lead to a faster and more precise diagnosis.

Methods: A retrospective examination of medical journals and MRIs of inpatients at Aalborg University Hospital admitted in 2015 due to status epilepticus, single or clusters of seizures and PNES. Patients diagnosed with seizures related to alcohol or withdrawal in the current admission, patients with seizures related to known primary or secondary brain tumor and patients with new intracranial pathology occurring within 90 days before admission were excluded. Exclusion was based on the fact that these conditions potentially can lead to abnormal MRI changes that cannot be differentiated from postictal abnormalities. 149 patients from a database of 347 admissions due to seizures had an MRI within 14 days after admission. In cases with acute MRI abnormalities not attributable to a known neurological disorder, localisation, characteristics and reversibility of the abnormalities were noted along with type and duration of seizures. In addition, it was found out if the MRI abnormalities were attributed to another pathology at a later point.

Results: Acute MRI abnormalities were found in 12 of 116 (10.3 %) cases of epileptic seizures. 2 of 12 (16.7 %) were given another diagnosis at a later point that could explain the abnormalities seen on MRI (posterior reversible encephalopathy syndrome and sinus thrombosis respectively). Thus, 10 of 114 (8.8 %) patients acutely admitted with epileptic seizures had MRI abnormalities not subsequently attributed to another pathology. They were scanned between 0 and 86 hours after the time of admission, while the patients with epileptic seizures without postictal abnormalities were scanned between 0 and 330 hours after the time of admission. 6 of 10 patients with postictal MRI abnormalities had a control scan between 4 and 237 days after admission. In all cases the abnormalities had disappeared on the control scan. As expected, no acute MRI abnormalities were found among the 8 patients with PNES.

Conclusion: 8.8 % of the included patients admitted in 2015 to Aalborg University Hospital due to epileptic seizure were shown to have postictal MRI abnormalities without other explanation for the abnormalities. In all subjects scanned afterwards, the abnormalities showed complete reversibility. I contrast, no acute MRI abnormalities were found among patients with PNES. It is possible that a better use of MRI in the early phase of seizures will lead to a faster and more precise diagnosis of epileptic seizures, but further research is needed. Future studies of postictal MRI abnormalities should be prospective, focus on the acute phase after a seizure and systemize the use of control scans to avoid some of the methodological issues encountered in the current study.
Publication date2016
ID: 246727960