Brain MRI Helpful for Identification of Causal Lesions in Adults with First Suspected Seizure
Abstract & Commentary
By Jennifer A. Best, MD, FACP, FHM, Assistant Professor, University of Washington School of Medicine, Seattle, WA. Dr. Best reports no financial relationships in this field of study.
SOURCE: An Hakami T, et al. MRI-identified pathology in adults with new onset seizures. Neurology 2013;81:920-927.
In settings in which an admitting neurology service is unavailable, hospitalists may be called upon to care for and initiate an appropriate diagnostic evaluation for adult patients with new-onset seizures. Such a workup generally includes metabolic laboratory tests, toxicology screening, electroencephalography (EEG) and brain imaging with CT and/or MRI. The utility of brain MRI in evaluation of patients with first seizure has not been well established, although two clinical guidelines have estimated the yield for epileptogenic lesions to be between 1% and 47%.1
In this study, Hakami and colleagues prospectively evaluated a cohort of Australian patients evaluated in an ambulatory setting following a suspected first seizure event over a 9-year period with the goal of establishing the diagnostic efficacy of brain MRI, determining the incidence and types of identified lesions and evaluating whether specific MRI findings correlate with findings on EEG.
Study patients were referred to the First Seizure Clinic at the Royal Melbourne Hospital from the emergency department or by a primary care practitioner. This clinic was staffed by a consulting neurologist with seizure expertise. Patients with established epilepsy or acute symptomatic seizures were excluded from participation, in contrast to previous studies which have included these populations. EEG (ideally within 24 hours of the event) and MRI ("standard epilepsy" or "temporal lobe epilepsy" protocols, with addition of contrast or MR angiography in patients with history of stroke or tumor) were ordered unless previously obtained; many patients had computed tomography (CT) of the head prior to clinic presentation.
At the time of physician evaluation, each patient’s event was classified as "epileptic" in nature, "nonepileptic" (including psychogenic and syncopal events) or "uncertain." "Epileptic" events were further stratified based on clinical history: "generalized convulsive," "generalized nonconvulsive," "focal convulsive," "focal nonconvulsive" or "unclassified."
MRI reports were reviewed by a neuroradiologist, and a random sample selected for review by a second neuroradiologist to ensure diagnostic congruency. Lack of agreement was identified in only 2% of cases; these files were reviewed by a third neuroradiologist, who established a consensus diagnosis. MRI findings were classified as 1) "epileptogenic" (structural lesions causally or possibly related to epilepsy), 2) "nonepileptogenic" (findings unlikely to be related to epilepsy or findings that were nonspecific) or 3) "normal" (no lesions). Drawing on previously published classification schemes, "epileptogenic" lesions were further categorized by radiologic appearance: "vascular," "developmental," "gliosis/encephalomalacia," "epilepsy-associated brain tumor" or "mesial temporal sclerosis." EEG findings were classified as "focal or generalized epileptiform," "focal or generalized slowing," or "normal."
Finally, the diagnoses suggested by MRI and EEG were correlated with clinical syndrome diagnoses, resulting in four categories: "lesional focal" (focal seizure with focal MRI), "nonlesional focal" (focal seizure with normal or nonepileptogenic lesion), "idiopathic generalized" (if generalized discharge seen on EEG) or "unclassified" (if uncertain type with normal MRI and EEG).
Following exclusions, 993 patients were included in the study cohort. Sixty-one percent were male; mean patient age was 42.2 years. Median time of presentation to clinic following the suspected seizure was 24 days. Seventy-two percent of patients presenting to the clinic were determined to have had an epileptic seizure (18% nonepileptic/10% uncertain) and classification of a specific epileptic syndrome was accomplished in 59%. Seventy-seven percent of study patients underwent MRI; the most common reasons for not completing MRI were failure to show for the appointment or a contraindication, such as indwelling metal. Incidentally, the majority of patients (79%) without MRI did undergo CT scan. Of the 764 patients who underwent MRI, 45% had a positive finding, and 23% of these findings were deemed potentially "epileptogenic." These lesions were most commonly identified as gliosis or encephalomalacia (48%); many of these patients had conditions predisposing to this development, such as trauma or stroke. Tumors were much less common at 15%. Potentially epileptogenic lesions were found to be more common in those who were diagnosed with epileptic events, those who had focal seizures and those with a syndrome diagnosis of focal epilepsy (P<0.00001); there was a nonsignificant trend toward higher incidence in patients greater than 65 years.
EEG was obtained in 94% of patients. Thirty-one percent of these tests were abnormal, with epileptiform discharges noted in 42%. A majority (52%) of patients with epileptogenic lesions had a normal EEG. Eighteen percent of patients had an abnormal MRI and EEG, although there was discordance on seizure foci location in 8% of subjects. Abnormal findings on MRI or EEG led to the start of an antiepileptic regimen in 6% of subjects.
In summary, although it is important to recall that this study was undertaken in a patient population presenting to an outpatient clinic at a single site and findings may not generalize to all settings, this evidence suggests that MRI with an epilepsy protocol (which may vary between institutions) should be added to EEG for evaluation of new-onset seizures in adults, as it has a high diagnostic yield, may supplement EEG which is frequently unrevealing, and may influence treatment planning. Furthermore, it is worth recalling that identification of focal lesions may render certain patients candidates for surgical therapies when medical therapy is ineffective. Neurology consultation is essential in interpreting MRI and EEG findings, adjudicating conflicting results between the two where necessary, and considering the risks and benefits of specific therapies for a given patient.
1. Krumholz A, Wiebe S, Gronseth G et al. Practice parameter: evaluating an apparent unprovoked first seizure in adults (and evidence based review) reports of the Quality Standards Subcommittee of the American Academy of Neurology and the American Epilepsy Society. Neurology 2007;69:1996-2007.