Frontiers in Imaging Markers for Epileptogenesis

The uptake of AMT reflects increased activity in the kynurenine pathway. This is the condition that contributes to seizures observed in persons with tuberous sclerosis.^13,14

The receptor ligands of serotonin (5HT)_1A also exhibit changes in persons with epilepsy; specifically, 5HT_1A binding is decreased in persons with temporal lobe involvement. The research of William H. Theodore, MD, chief of the clinical epilepsy section at the National Institutes of Neurological Disorders and Stroke, suggests that PET with a marker that shows deficits in signaling for serotonin is more sensitive than traditional PET measurements of brain glucose. This can be useful for detecting epileptogenic zones in patients with temporal lobe epilepsy and reduces the need for EEG studies.¹⁵

"We used PET to compare 5HT_1A binding measured with the silent antagonist, ¹⁸FCWAY, with glucose metabolism measured with [¹⁸F]FDG [¹⁸fluorine-fluorodeoxyglucose], in a group of 19 patients who underwent surgery for uncontrolled epilepsy," explained Theodore. "Three patients—including one with a normal MRI—had unrevealing FDG PET, but a clear FCWAY asymmetry. We noticed that patients who were free of seizures tended to have greater FCWAY, but not FDG-A1, than patients with persistent seizures. So, it looks like FCWAY PET may be more sensitive than FDG," Theodore surmised.

If changes detected on PET reflect dynamic disturbances, rather than cell loss, it is possible that labeling with a shorter half-life radionuclide would permit continuous time-lapse studies to track markers of epilepsy.

Functional MRI (fMRI). fMRI can show metabolic changes that occur as a consequence of red spikes. Several investigators have used EEG spike-triggered fMRI averaging to anatomically localize the source of interictal EEG spikes. Data thus far indicate that fMRI images produced by spontaneous spikes originating from primary epileptogenic regions differ from those produced by propagated spikes arising from secondary epileptogenic areas.^10,11

A new application for fMRI involves the use of magnetized nanoparticles (MNPs), which attach to bioactive molecules and are visible on MRI. MNPs can be used to measure localized alterations in neurotransmitter activities that reflect brain excitation and inhibition, cerebral metabolism, immune responses, and drug distribution. The potential for MNP to serve as a marker for epileptogenesis is significant.¹

Transcranial magnetic stimulation (TMS). TMS is capable of measuring cortical excitability, a potential surrogate marker of epileptogenicity.^16,17 This capability also could be applied to follow the patient's response to treatment.

Gene microarray technology. Gene microarray technology examines epileptic disturbances at the genetic level. It follows alterations in gene expression profiles that occur in brain tissue during epileptic episodes. In addition to showing the presence of epilepsy, gene expression profiles reflect the severity of the epileptic condition.¹ A particular advantage of using gene microarray technology as a marker for epileptogenesis is that genetic fingerprints of brain disorders can be measured in peripheral blood; thus, patients could potentially be tested for epilepsy with a finger stick.¹⁸

Optical intrinsic signal imaging. This invasive technique measures changes in brain tissue associated with neuronal activity. Primarily used during surgery, it may eventually be used to follow surrogate markers for epilepsy.¹⁹

MARKERS FOR PREDICTING SEIZURES

The use of epileptiform EEG events to predict the onset of seizures—and to perhaps abort them before they occur—has long been a focus of research. Recent reports indicate that dynamic analysis of ongoing depth-recorded EEG activity can identify changes in neuronal synchronization that precede epileptic seizures.¹

Evidence also indicates that scalp EEG recordings, an essential preliminary test for patients undergoing surgery to correct seizures, could potentially be used to noninvasively predict seizure onset. One problem with this technique has been widespread misconception with respect to the necessary cortical area, synchrony, and amplitude needed to generate EEG tracings that can be recorded at the scalp.

A group from the University of Chicago used simultaneous scalp and intracranial EEG recording to further explore this issue.²⁰ According to James Tao, MD, assistant professor of neurology and lead investigator, "We found that the cortical area needed to record epileptiform discharges is larger than most of us think. We needed a cortical area of at least 10 to 20 cm² before we could generate a recognizable interictal spike or ictal rhythm." So, to reliably use scalp EEG tracings to mark epilepsy, it is essential to have sufficient cortical source area and synchrony. Additional research will yield further insights into using this technique to study markers of epileptogenesis.

THE CHALLENGE AHEAD

While the possibility of using imaging studies to identify markers for epileptogenesis is exciting, this area is not without its challenges. Shlomo Shinnar, MD, PhD, professor of neurology and pediatrics, Hyman Climenko Professor of Neuroscience Research, and director of the Comprehensive Epilepsy Management Center of Montefiore Medical Center at Albert Einstein College of Medicine in New York City, explained: "The imaging techniques that we have today are certainly important and useful, but their application and interpretation really depend on the context of the situation. If you have a patient with a long history of temporal lobe seizures related to hippocampal atrophy, then regular MRI is fine—as long as you still use quantitative volumetrics to do the studies. Other forms of imaging are not that well developed yet."

One of the biggest challenges in the field is differentiating lesions of brain trauma or stroke from those of epilepsy. "Trauma or stroke will produce an abnormal MRI. Now, only a proportion of those patients will go on to have epilepsy, but all will have abnormalities on their MRI," Shinnar pointed out. "The key is learning how to distinguish the changes related to that stroke or trauma from those that reflect epileptogenesis. It is not that easy, considering that it can take 5 to 10 years—usually 5 years—between the initial insult, and the onset of epileptic seizures. If you do find something that suggests epileptogenesis, you have to be prepared to track it and follow its association with seizures, for years."

Shinnar noted that this is just the beginning. "Once that presumed marker has been identified and followed for at least 5 years with serial studies, then it is necessary to conduct clinical trials to test that biomarker and examine its association with epilepsy risk factors and occurrence. If that biomarker stands up to the rigors of clinical trials, then drugs can be developed that work at that level to stop the epilepsy process—not just control seizures," Shinnar observed. "We could potentially shift our attack from merely controlling symptoms, to actually preventing or even reversing seizures."

Even recognizing the need to differentiate changes related to brain trauma and stroke from those related to epilepsy is a big step forward in this process. "This is probably one of the most significant advancements in our thinking to have occurred over the past 5 years or so. In the next few years, look for researchers to be making real strides toward our ultimate goal of preventing or reversing the epilepsy process."