The earlier seizure activity is detected, the better the chance of preventing it. For several years, neurologists Gregory A. Worrell, M.D., Ph.D., and S. Matthew Stead, M.D., Ph.D., and their neurosurgical colleagues W. Richard Marsh, M.D., and Fredric B. Meyer, M.D., at Mayo Clinic in Rochester, Minn., have been working to pinpoint the exact moment and the precise location of seizure generation.
Two years ago, using microelectrodes 40 microns in diameter, or thinner than a human hair, the research team began recording EEG activity from brain regions the size of cortical columns at frequencies beyond the upper and lower limits of standard EEG recordings. Cortical columns, the smallest functional unit in the cortex, are approximately 300 microns across and contain from 1,000 to 7,500 neurons. In perspective, a typical EEG electrode captures the activity of millions of neurons and hundreds of cortical columns.
This research revealed that isolated brain regions the size of cortical columns do, indeed, show evidence of seizure activity (microseizures) in humans. The investigators found that microseizures occurred most often in patients with epilepsy but also occurred, although rarely, in control patients without epilepsy. This latter finding demonstrates that pathologic oscillations can occur even in people who do not have epilepsy.
How do microseizures transition to clinical seizures in patients with epilepsy? The findings of Drs. Stead and Worrell support the hypothesis that in patients with epilepsy, individual microdomains the size of cortical columns generate frequent hypersynchronous discharges, which recruit other columns of neurons.
Similar, noncolumnar groupings of neurons have been identified in the rat hippocampus by Anatol Bragin and colleagues at UCLA, and are known as pathologically interconnected neuron (PIN) clusters. These PIN clusters can be considered microdomains of epileptogenesis or seizure initiation. When a critical volume of microdomain activity is reached, a large-scale seizure is generated. This explanation is sometimes referred to as the sick column hypothesis.
Clinical EEG recordings do not probe the spatial and temporal scales of microdomain activity, which makes early detection difficult. This hypothesis may explain why some patients do not respond to first-generation responsive stimulation devices designed to abort seizures.
Two new implantable epilepsy devices that use electrical stimulation are undergoing multicenter trials for treatment of medically refractory epilepsy.
The initial results from these 2 pivotal multicenter trials have demonstrated evidence for reducing seizure frequency in patients with medically intractable partial epilepsy. Importantly, each of these trials investigating implantable devices has shown good safety.
Not knowing when a seizure will occur is one of the most debilitating aspects of epilepsy. Dr. Worrell notes, "Patients may only have 4 or 5 seizures a month, each lasting a few minutes. But those few minutes mean they can't drive, and they may be afraid to go out in public for fear of having a seizure. These effects can lead to social isolation."
The recent discovery that seizurelike events can occur in pathologic microdomains (ie, microseizures) in humans adds to a growing body of evidence that seizures may begin before they are evident clinically. In the future, devices may be able to provide patients with warnings that a seizure is about to occur or, possibly, to prevent a seizure.
Dr. Marsh, a neurosurgeon who performs epilepsy surgery and stimulator implantation, points out that after the safety of a device is determined, the question of its efficacy and clinical benefit rests on the length of time between seizure warning and seizure onset. Dr. Stead adds that if that period is reasonable (neither too long nor too short), fast-acting drugs or neurostimulation could be triggered to prevent seizures.
The future of seizure prevention rests on understanding the mechanisms underlying seizure generation. As knowledge in these areas advances, future clinical applications include improvements in epilepsy surgery, improvements in the devices delivering neurostimulation, and the possible development of algorithms that identify periods of increased probability of seizures.