Trudy, a 34-year-old bank employee, had been suffering from epilepsy for more than 18 years. She had tried all the usual medications, with little success. Typically she would feel nauseated before an oncoming seizure, then lose consciousness. A few minutes later she would wake up, exhausted. According to her husband, she would smack her lips during her seizures and fumble with her hands.

When it got to the point that Trudy was experiencing two to three seizures a week, she decided to contact the Epilepsy Clinic at Bonn University, Germany, which she had heard about in a television report. Several weeks later she had her first outpatient appointment. After a detailed discussion of her medical history, physicians took blood samples, an electroencephalogram (EEG) of her brain and magnetic resonance images (MRI) of her head. Within days a doctor called to tell Trudy that surgery was recommended and that she should come in for an inpatient workup. Trudy was glad--and scared.

More than 2.5 million people in the U.S. and 600,000 in Germany have epilepsy. About two thirds of them are freed of seizures with drug therapy, but for the rest, surgery is the only other option. Although the operations carry risks, 60 percent of adults and 70 percent of children remain free of seizures afterward. As physicians perform more procedures, the presurgical testing and the final outcomes are helping researchers learn more about the condition and the workings of the human brain.

Many Names

Seizures occur when groups of overexcited neurons suddenly fire in unusual synchrony. The chorus rings out from a limited region of the brain--the onset zone. Given Trudys symptoms, the clinics specialists suspected her seizures were initiated in the temporal lobe. Sure enough, radiologists reading the MRI scans detected peculiarities in the left hippocampus--a C-shaped structure deep within the temporal lobe--which seemed to have resulted from scarring. Back when Trudy experienced her first seizures, the minute structural changes would not have been visible; only since the debut of high-resolution MRI technology in the mid-1990s has such damage been discernible.

The sacred disease of the Greeks, Valentines sickness, falling sickness--the illness has afflicted human beings since time immemorial. During a seizure, the victim appears to be obeying an alien will. That is why epilepsy was often assumed to have supernatural causes. Followers of the ancient Tibetan Bon religion believed that epileptics were chosen people, but the ancient Jewish and later Christian traditions viewed the condition as Gods punishment or the work of demons. Not everyone was convinced, however. The legendary Greek physician Hippocrates observed in the fifth century B.C. that head injuries in soldiers and gladiators sometimes led to seizures that mimicked those of his own patients. He concluded that the brain caused the affliction, but scientific research did not make substantial gains until the 20th century.

Today physicians distinguish more than 30 forms of epilepsy. One of the best known is characterized by the petit mal seizure, sometimes called absence seizure--minor episodes of 10 to 20 seconds during which victims eyeballs may roll upward and they no longer respond to stimuli. A grand mal seizure has a generalized pattern whereby the person first loses consciousness, then becomes rigid, often falling. The arms and legs convulse and twitch uncontrollably, for as long as two minutes. At least half of all cases are symptomatic--attributed to abnormal changes in brain structure caused by tumors, trauma or inflammation. In other cases, however, no such peculiarities are detectable.

Neither a single nor several random seizures necessarily lead to epilepsy. High fever occasionally causes convulsions in infants, and sleep deprivation can trigger a seizure in children and teenagers. Approximately 10 percent of all people will experience at least one seizure during their lifetime. Our brains differ primarily in the partially genetic individual threshold for an episode, beyond which groups of neurons fire spontaneously and synchronously.

Exorcising the Demon

Our understanding of neuronal behavior during a seizure advanced markedly after 1924, when neuropsychiatrist Hans Berger developed electroencephalography. Electrodes affixed to the scalp register fluctuations in the electromagnetic fields created by neuronal activity. In modern instruments the signals are electronically enhanced, digitized and stored. It is impressive how an EEG changes at the moment a grand mal seizure begins: just a second earlier the pens that trace activity on graph paper--like a seismograph--draw fine, mildly undulating lines. Suddenly, one electrode pen jumps. Then, within seconds, they all show the same sharp peaks and valleys, as thousands of neurons discharge in lockstep. It seems as if a neuronal bandleader suddenly gave the orchestra the direction, All together now!

The strong, rhythmic neuronal activity explains why the behavior and experience of the patient change so abruptly. The symptoms depend on where the epileptic onset zone, or focus, is located in the brain, how far the activity spreads from that focus, and what the involved brain areas normally do. One patient may barely notice a transient seizure, whereas another is plunged into violent, unconscious convulsions.

What determines when neurons will fire synchronously remains a riddle. Transitional phases, such as waking up, relaxing, or getting angry or stressed, appear to be precarious. Some people react to flickering light or acoustic stimuli. Sleep deprivation, physical exhaustion and alcohol all enhance the chance of seizures.

Patients understandably want to get rid of their epilepsy at all costs. Medication is the method of choice, and there are a dozen approved substances. Several new anticonvulsants (also known as antiepileptics) are in clinical trials. Yet all anticonvulsants have a drawback: they do not actually cure the cause. They only suppress the neuronal hyperactivity, thereby preventing seizures. Patients must therefore take these tablets every day. The drugs also have mild to severe side effects, including weight changes, fatigue and concentration problems.

Two primary surgical strategies exist: resection and transection. In resection, surgeons attempt to remove all of the onset region, but only that region. The prerequisite, naturally, is that there be a single onset zone that can be precisely localized. If this is not the case, then transection is done, to disconnect distant tracts of neurons that seem to fire together. Although this procedure does not lessen the frequency of seizures, it prevents their spread, minimizing the intensity of symptoms.

A third treatment, approved by the U.S. Food and Drug Administration in 1997, is vagus nerve stimulation. An impulse generator, much like a cardiac pacemaker, is implanted in the chest wall. It emits electrical signals that stimulate the vagus nerve in the left side of the neck. For some patients, this action reduces hyperreactivity in the brain after a few months, by a mechanism that researchers have not yet elucidated. Fewer than 5 percent of patients enjoy complete freedom from subsequent seizures, however.

Probing the Onset Zone

The current therapeutic rule of thumb is that if a patient continues to experience seizures after trying several medications for two to three years, he or she should be evaluated for surgery. Too many patients, or their doctors, wait too long, unnecessarily prolonging a fractured quality of life, as Trudy had. Soon after her initial outpatient exams, Trudy returned to the Bonn clinic for her inpatient stay.

Neuropsychologists first performed a variety of mental tests [see box above]. Like many patients with temporal lobe epilepsy, Trudy apparently suffered from slight verbal memory impairment, which is known to occur more frequently when the seizure onset zone is in the left hippocampus. But whether her seizures truly originated there could be determined only during an actual seizure recorded by an EEG.

So Trudy checked into her private room at the centers monitoring unit, which trains video cameras on its guests 24 hours a day [see illustration on pages 62 and 63]. She wore a bonnet that held electrodes against her scalp, wired to a bedside EEG machine. Technicians who monitor the video in a separate room spring into action as soon as they notice telltale peculiarities that signify an oncoming seizure. When this occurs, they help patients and conduct short behavioral tests during the episode, which enable them to characterize the seizure more precisely.

Trudy did not have to wait long. Her first seizure ignited after only six hours. Three more followed in the first two days. Nevertheless, the electrode tracings did not enable physicians to locate the onset zone precisely; they could not tell conclusively whether the spark occurred in the left or right hippocampus. For some patients, the EEG signals become too distorted or damped as they travel through the brain tissue and skull. Better accuracy would require that electrodes be implanted under the cranium.

Without hesitation, Trudy decided to undergo the procedure. Surgeons drilled two small holes in the cranium, above the left and right temporal lobes, and placed electrodes almost directly on the cerebral cortex. They then carefully slipped in a fine electrode at a precalculated angle from the occipital lobe deep into both the right and left temporal lobes, to measure activity along the length of the amygdala at the core of the brain all the way to the hippocampus.

Two days later Trudy had two seizures identical to the previous ones. Each time, the epileptic activity began deep within the area by the left electrode. All findings pointed to the left hippocampus as the sole onset zone. Trudy was a good candidate for epilepsy surgery.

Limiting Risk

The surgeons explained to Trudy that she was lucky the onset zone was limited to the hippocampus in one hemisphere. Under no circumstances can both hippocampi be removed. They told her about the famous case of an epileptic man known as H.M. at the Montreal Neurological Institute who had both hippocampi removed in 1953. Although the 27-year-old patient became largely free of seizures, the procedure robbed him of his memory. H.M. could access memories that had been stored prior to the operation, but he forgot every subsequent experience within five minutes. At least one functional hippocampus is indispensable for our ability to write our ongoing autobiography.

Epilepsy patients from H.M.s era through today have supplied scientists with important knowledge. The implanted electrodes offer a unique opportunity to measure brain activity in real time directly--something that even functional MRI cannot do, because the images lag the actual processes by several seconds. Like many patients do, Trudy had consented to undergo a few experiments while the electrodes were in her head. She performed computerized exercises for 30 to 40 minutes, and the data showed precisely what brain activity correlated with a particular cognitive event, such as recognizing a word when it appeared on the computer screen. Based on such experiments, the Bonn researchers have since explained which brain regions are involved, and in what order, when we perceive a word, as well as when we later recall that word--a possible boon to understanding language impairments.

To minimize the risk of neurological damage, neurosurgeons remove as little brain tissue as possible. Today no one would resect two thirds of the temporal lobe, the way doctors did years ago. Much more common is to limit removal to the amygdala and hippocampus on one side, leaving the rest of that hemispheres temporal lobe untouched. The Bonn doctors recommended this form of surgery for Trudy.

Such standardized surgical routines have been worked out for many types of epilepsy. But if the onset zone is located in the frontal lobe or the parietal lobe, it is hard to predict whether the operation will damage regions crucial to motor functions and speech, among other abilities. In these cases, doctors implant electrodes during the presurgical diagnostic workup. The electrodes record epileptic brain activity but also allow surgeons to apply electrical impulses to specific brain tissue. This procedure essentially simulates the consequences of the planned surgery. A neurologist will systematically activate each electrode with currents of various strengths and frequencies, while asking the patient to perform certain actions, such as counting out loud.

Stimulating a motor area might cause a patients finger to start to twitch, for example, whereas pulsing the association cortex usually leads to speech deficits. More uncommon phenomena may occur, such as intense emotions or a sudden flashback to a long-lost memory. In one well-known case a woman began laughing hysterically, telling the doctors they were just too comical, the way they were standing around her. Changes in visual and spatial perception can also be fascinating. One patient, whose neurologists were examining the gyrus angularis in the parietal lobe, suddenly felt as if she were floating above the bed, observing herself lying there--an out-of-body experience. Her response indicates that, perhaps, supernatural experiences stem from odd brain processes.

The main purpose of such stimulation, before surgery, is to create a functional map of the individuals brain that will guide surgeons as they decide what and how much tissue to cut or remove.

Surgery Day

Trudys surgery date finally arrived. The neurosurgeons needed half an hour to work through the large fissure that separates the temporal lobe from the frontal lobe and reach deep into the left temporal lobe. They were guided by MRI pictures of Trudys brain and a microscope that enlarges the view of the surgical field. It is important to damage as few blood vessels as possible and to exert minimal pressure on the tissue. The next step was to carefully remove the amygdala and hippocampus in the left hemisphere. The operation took four hours.

Before surgery, Trudy had agreed that tissue taken from her brain could be used for research. Once extracted, it was immediately placed in a nutrient solution so electrophysiologists could test it even after 20 to 30 hours; most work would occur right away, though, given this chance to investigate living neurons that only a short time earlier had performed real functions in a human brain. The researchers hoped to gain insight into what causes synchronized firing among thousands of neurons.

What the neuropathologists found was a markedly reduced number of nerve cells in certain lower hippocampus regions, confirming the suspected scarring. They also discovered a peculiarity in individual nerve cells: a larger than normal proportion spontaneously exhibited so-called burst discharges--the firing of three action potentials in uncommonly rapid succession. This propensity to discharge could have a number of causes, such as changes in ion channels or neurotransmitter receptors in the cell membrane. Such peculiarities may result from injury or be genetic.

Trudy recovered quickly from the operation.Her seizures, initially, seem to have stopped, and she is delighted. Still, she knows that the verbal memory deficits she already had--because of too many years of inadequate treatment--could worsen. More time will be needed to assess permanent changes, good and bad. But Trudy is optimistic. Her personal and work lives should be easier. And she will be less likely to be socially shunned, as if she were cursed.

The rest of us owe a debt of thanks to epilepsy patients such as Trudy who, during their own anxious medical ordeals, unselfishly participate in studies that help scientists--and ultimately all of us--better understand the human brain. The best thank-you society could offer in return would be greater understanding and acceptance of people suffering from this condition.