News from the American Epilepsy Society Annual Meeting
New DTI Technique Finds Epileptogenic Zone Where Conventional MRI Fails
By Dan Hurley
January 19, 2017
ARTICLE IN BRIEF
A new diagnostic approach using diffusion tensor imaging to locate epileptogenic abnormalities not detected by conventional 3T magnetic resonance imaging was successful in 78 percent of 43 cases studied. Independent experts said the technique is promising, but it needs replication in larger numbers of cases.
HOUSTON—A new diagnostic approach using diffusion tensor imaging to locate epileptogenic abnormalities not detected by conventional 3T MRI was successful in 78 percent of 43 cases studied, according to an abstract presented here at the American Epilepsy Society annual meeting in December.
The technique is already being regularly used to guide implantation of intracranial electrodes at the University of Munich, where the study was conducted. Neurologists familiar with the technique said it is also being adopted at some epilepsy centers in the United States.
“Other centers are using it too,” said Lara Jehi, MD, director of research at the Cleveland Clinic Epilepsy Center. “When patients don't have a clear lesion on standard MRI, clinicians have a hard time narrowing down the specific area of the brain where seizures are suspected to come from. The more guidance we have for where we can place the electrodes, the higher the chances of success.”
She and other neurologists emphasized, however, that the report requires validation in larger, prospective studies, with outcome data to substantiate whether the technique ultimately results in fewer seizures following surgery.
The DTI technique involves quantification of regional U-fibers to identify microstructural abnormalities in patients with cryptogenic focal epilepsy (cFE). U-fibers represent connections between adjacent gyri of the brain, located within the cortex or in the very outer parts of the subcortical white matter.
Although U-fibers are the subject of relatively few published studies in the medical literature, they provide a logical site of scrutiny, said the first author of the paper, Christian Vollmar, MD, PhD, a neurologist at the University of Munich Hospital's epilepsy center and its department of neuroradiology.
“We know from histopathology of resected specimens that the most frequent pathology in cases with cryptogenic focal epilepsy is focal cortical dysplasia (FCD),” Dr. Vollmar said. “FCD is typically located at the gray-white-matter border, typically in the bottom of a sulcus. And this is where U-fibers are located. The density of U-fibers we analyzed is therefore a specific neuroimaging biomarker for the gray-white-matter border and seems to be very sensitive to alterations caused by FCD.”
The study began with 55 cFE patients and 60 healthy controls. Whole-brain tractography by DTI was performed, and tracts representing subcortical U-fibers were selected based on length, curvature, and shape. Tract density images were then created, specifying the number of tracts passing through any voxel. Clusters of significant reductions in U-fiber density were compared with clinical data from presurgical evaluation, including ictal and interictal electroencephalography (EEG), seizure semiology, fluorodeoxyglucose-positron emission tomography (FDG-PET) and ictal single-photo emission computed tomography (SPECT) imaging.
Of the original 55 cFE patients, 12 were excluded because clinical data was inconclusive, thereby preventing comparison to the U-fiber analysis. Among the remaining 43, clusters of significant U-fiber reductions were found in 91 percent of the patients, and 78 percent were consistent with the clinical localization of the epileptogenic zone. Twelve patients underwent intracranial EEG recording with depth electrodes, confirming seizure onset in 74 percent of the detected U-fiber reductions. Eight patients had resective surgery and the most frequent histopathological finding was mild malformation of cortical development, with blurred gray-white-matter border and ectopic neurons in the white matter.
The Munich team's working explanation for their finding is that “ectopic neurons disrupt the microstructural order of subcortical white matter, thereby reducing the number of reconstructed U-fibers in this region.”
Steven C. Karceski, MD, assistant professor of neurology and director of clinical trials at the Weill Cornell Epilepsy Center, said the findings look promising, particularly because DTI is performed using the same MRI machines that epilepsy centers already have.
“This might turn out to be directly applicable,” Dr. Karceski said. He added, however, that the current trial included just 55 people, so it needs to be replicated. “You need to have somebody who's really good at doing DTI mapping. If the parameters are not set in just the right way, you can skew the findings in such a way as to trick yourself into thinking there's an abnormality where there's not,” he said.
Moreover, Dr. Karceski said, “Just finding an abnormality doesn't mean it's the cause of the seizures. The next step is to incorporate this into studies to see if DTI mapping improves surgical outcomes.”
Dr. Jehi said the new technology would be a welcome addition to the tools already available for pinpointing an epileptogenic region in drug-resistant patients.
“The challenge with patients who don't have a clear lesion on their MRI is that we have a hard time narrowing down the specific area of the brain we suspect seizures to come from,” she told Neurology Today. “The more guidance we have for where we can place these invasive EEG electrodes, the more targeted the evaluation will be and the higher chance of success.”
Lily Wong-Kisiel, MD, assistant professor of neurology with a specialty in pediatric epilepsy at the Mayo Clinic, said that about a quarter of drug-resistant patients who present for epilepsy surgery evaluation have normal MRI.
“That's a significant number of patients where there is no visible target,” she said. “Patients who have an imaged lesion have better seizure-free outcomes following surgery. Having a technique like DTI mapping, when it is validated for improving epilepsy surgery outcomes, could be helpful for our patients.”
Dr. Vollmar told Neurology Today that his group in Munich has already acquired and analyzed DTI data in many more cases.
“No single diagnostic modality alone is ever good enough to base surgical decisions on in case of cryptogenic focal epilepsy,” he said. “But for us it is definitely good enough to base further investigations on, particularly to plan the implantation of invasive electrodes, and to give it the same weight that other established methods like ictal SPECT or FDG PET have. We encourage anyone treating these patients to consider our approach.”
EXPERTS: ON U-FIBER IMAGING FOR EPILEPTOGENIC ABNORMALITIES