LOS ANGELES — Magnetic resonance imaging is useful for identifying the etiology of new-onset afebrile seizures in infancy and, when available, should be included as a part of the standard diagnostic evaluation in this population, according to Dr. William S. Benko.
Of 103 MRIs performed in 144 infants presenting to the emergency department of Children's National Medical Center in Washington between January 2001 and January 2005 with new-onset afebrile seizures (NOAS), 40% detected diagnostic abnormalities related to the seizures, Dr. Benko said in a presentation at the annual meeting of the Child Neurology Society.
By comparison, in 136 of the 144 infants who underwent CT, abnormalities were detected in 42% of the scans; however, 15% of these were deemed incidental findings.
Fifteen of the infants with normal CT scans had abnormal MRIs. “In the majority of these CT-negative, MRI-positive patients, the CT scans did not identify focal abnormalities,” said Dr. Benko of the National Institute of Neurological Disorders and Stroke in Bethesda, Md.
According to practice parameters published in 2000 by the Quality Standards Subcommittee of the American Academy of Neurology, the Child Neurology Society, and the American Epilepsy Society, the evaluation of a first nonfebrile seizure in a child should include EEG as a routine part of the diagnostic evaluation to predict the risk of recurrence and to classify the seizure type and epilepsy syndrome (Neurology 2000;55:616–23).
There was insufficient evidence for the routine use of other studies, such as lumbar puncture and neuroimaging, so these were deemed warranted only under specific circumstances and at a neurologist's discretion, according to the parameters.
More studies with large, well characterized samples were needed, according to the parameter, before neuroimaging could be considered for routine evaluation of NOAS.
Toward that end, Dr. Benko and colleagues investigated a prospective cohort of 1,189 patients presenting to the hospital's emergency department identified with possible NOAS.
Out of the entire cohort, 144 patients were infants. By the time of presentation to the emergency department, 82% of the infants had experienced two or more seizures. All of the patients were evaluated by a child neurologist on the basis of the patient history: laboratory values, including complete blood count, electrolytes, urinalysis, and toxicology screen; 24-hour observation; electroencephalography; and CT, MRI, and/or lumbar puncture at the clinician's discretion.
“In this afebrile population, we found that CBC, urinalysis, and toxicology screens were not at all useful or contributory,” Dr. Benko said. Of 59 infants who were given lumbar punctures, 90% had normal results and 5 had evidence of pleocytosis.
Treatable electrolyte abnormalities, including hypocalcemia, hyponatremia, and hypoglycemia were detected in five of the children. Electroencephalography showed abnormal results in 62% of patients, including focal abnormalities, hypsarrhythmia, and generalized spikes.
By comparison, neuroimaging results were particularly revealing, Dr. Benko said. With CT, congenital malformations were identified in 12% of the patients imaged, evidence of trauma was seen in 5%, and atrophy was evident in 3%.
Among the patients scanned by MRI, dysplasia was seen in 14%. These abnormalities included focal diffuse involvement, dysgenesis, heterotopias, Aicardi syndrome, and Dandy Walker malformation.
MRI also detected vascular events in 8% of patients. These included new and old cerebrovascular accidents, hemorrhage, subdural hematoma, and Sturge-Weber syndrome.
The abnormalities noted on MRI but missed on CT included dysplasia, mesial temporal sclerosis, cerebrovascular accidents, and tuberous sclerosis, he said. “The yield of CT-negative, MRI-positive findings was highest in patients with focal neurologic exam, neurodevelopmental delay, focal EEG, focal seizure, [and MRI evidence of right mesial temporal sclerosis].”
The decision to perform MRI was made by the treating neurologist, raising the possibility of a selection bias, Dr. Benko said.
In addition, the findings are limited by the fact that not all the patients who underwent CT scans also underwent MRIs, concluded Dr. Benko, who conducted the investigation during his pediatric neurology fellowship at Children's National Medical Center.
This MRI sagittal image shows a single subependymal lesion not seen on CT.
A proton density axial image shows tubers of tuberous sclerosis complex. Photos courtesy Dr. William S. Benko