Advanced MR imaging in epilepsy assists presurgical evaluation

Two different advanced MRI techniques can help clinicians choose epilepsy patients who will favorably respond to surgical intervention, according to a pair of studies published in the October 2005 issue of Neurology.

One technique uses 3T phased-array MRI to gain finer resolution, translating into detection of lesions not visible on 1.5T. The other is a 3D mapping method that uses volume differences to improve surgical prognosis.

About 15% of patients with focal epilepsy are not adequately managed with anti-epileptic drugs and may be candidates for surgical intervention. MRI has been used to detect lesions in the epileptogenic area of the brain.

When MRI and electroencephalogram findings are concordant, 72% of patients are more likely to be seizure-free after surgery, compared with 41% of patients whose MRI/EEG findings are not concordant.

Dr. Susanne Knake and colleagues from Massachusetts General Hospital and the University of Marburg in Germany reasoned that more sensitive MRI technology is needed. They compared 1.5T MR scans with 3T phased-array MR scans in 40 patients with medically intractable focal epilepsy (Neurology 2005;65:1026-1031). All 40 patients were considered candidates for epilepsy surgery.

Using 3T phased-array MRI, the researchers detected new lesions that had not been seen with 1.5T MRI in 15 patients. The new data changed clinical management in eight patients who had focal cortical dysplasias, lesions that are typically difficult to detect with 1.5T MRI. The 3T results also influenced the surgical procedure in four of these patients, who were seizure-free after 10 to 16 months of follow-up.

Another 15 patients had abnormal studies using both 1.5T and 3T, but the 3T images contributed additional relevant clinical information in five of these patients. A larger area of damage was identified in two of the five. Both have undergone surgery, and one has experienced a significant reduction in seizures, while the other is seizure-free after 12 months follow-up.

In another of these 15 patients, 3T imaging revealed bilateral hippocampal sclerosis and a left hemisphere stroke. As a result, the patient was considered a poor surgical candidate. Nevertheless, a right hippocampal-amygdala resection was performed, but the patient has had no significant improvement in seizure frequency.

Of the remaining patients, two had normal 1.5T studies and indeterminate 3T studies. The 3T studies raised the suspicion of a focal cortical lesion. Neither patient has proceeded to invasive monitoring or surgical resection, and the significance of the 3T findings is unclear, according to the authors. Eight other patients had studies that were normal with both 1.5T and 3T. None has progressed to surgery.

"The proper identification of abnormal tissue, which may otherwise be unrecognized, is crucial for identifying a potential epileptic focus and thereby increases the likelihood of a better surgical outcome," said Dr. Fernando Cendes, an associate professor of neurology in Brazil's University of Campinas, in an accompanying commentary.

In the second study, Dr. Jack J. Lin and colleagues from the University of California, Los Angeles showed that analysis of hippocampal volume differences using 3D surface meshes of MR scans can predict which patients with temporal lobe epilepsy will respond adequately to surgical intervention (Neurology 2005;65:1094-1097).

Patients with mesial temporal lobe epilepsy and hippocampal sclerosis are more likely to be helped by surgical intervention than with anti-epileptic drugs. Some patients continue to experience seizures postsurgery, however. Predictors of which patients will be seizure-free have been elusive.

The team developed a surface modeling technique to compare hippocampal anatomy in patients free of seizures postoperatively with those who continue to have seizures. These mapping techniques were applied to isolate profiles of hippocampal deficits associated with better or poorer outcomes after surgery.

Using this technology, the researchers examined the hippocampi of 40 patients who had received surgical intervention for mesial temporal lobe epilepsy. They found that non-seizure-free patients had more diffuse atrophy in the area of the hippocampus ipsilateral to seizure activity, compared with seizure-free patients. They hypothesized that the greater atrophy could be attributed to disease progression.

The researchers concluded that surface-based anatomic mapping can detect prognostic features of hippocampal anatomy.

In his commentary, Cendes said the two studies raise additional questions: Should general neurologists be arranging for their patients to receive these sophisticated MRI techniques? Might these tests point the way to more restricted surgical approaches, such as a lesionectomy?

More studies are needed to answer these questions, Cendes said.

"As our imaging becomes more refined, so will our ability to ask and answer more complex and subtle questions about the management of medically resistant partial epilepsy. This may be the greatest argument in favor of promoting newer imaging techniques," he said.

For more information from the Diagnostic Imaging archives:

MRS gains indications, but still fights for reimbursement

MR imaging reveals signs of temporal lobe epilepsy

Intraoperative high-field MR imaging revamps neurosurgery practice

MR explores hippocampal abnormalities of epilepsy