BY JOSEPH R. MADSEN, M.D.

Nikki, now a college sophomore, was born with a very large fluid collection in the right side of her brain, causing pressure and seizures. When she was 14 months old, the pressure in the cyst -which was like a water balloon full of cerebrospinal fluid - was treated with an implanted drain to divert the fluid; the cyst decreased in size and virtually went away.

Her seizure problem did not end, however; although free of grand mal convulsions, she continued to have frequent staring spells for 30 to 60 seconds at a time. Like many patients with a seizure disorder, or epilepsy, she was able to live a generally normal life between spells.

Although she would sometimes have seizures when running around the bases during a softball game, she kept playing. Unlike most afflicted patients, however, her seizures could never be completely stopped with medicines.

In February of 1992, when Nikki was 15, the worst nearly happened. While swimming in the deep end of a pool, she had a seizure and nearly drowned. She was immediately pulled out of the water and resuscitated from a full cardiac arrest, and, fortunately, recovered from this with no additional neurological problems. She and her parents became very interested in more drastic measures to control her epilepsy.

Several investigations, including brain wave studies while she was monitored by video cameras, suggested that the right temporal area, where the brain tissue had grown back into the large fluid space, was the major culprit in her seizure disorder. The decision eventually made by the surgeon, neurologists, family, and Nikki herself was to expose the brain, remove the wall of the cyst so that she would not need a drainage shunt any more, and implant electrodes onto the surface of the cortex (the brain's folded outer layer) to try determine precisely where the seizures were coming from.

After three operations, culminating in aggressive removal of the offending temporal lobe, she stopped having her seizures and the premonitions (or "auras") of seizures entirely. She is an honor student in college and safely engages in swimming, surfing, and biking.

Nikki's situation illustrates a potential for cure of epilepsy, but the major problem remains: When is it safe to remove all of the abnormal tissue, and what should be done if the functional areas, such as those responsible for speech or movement, overlap with the regions generating the epileptic spikes?

The range of seizure operations today|
Even though surgical treatment for epilepsy is more invasive and traumatic than treatment with medicines (which is far more common) the strategic thinking behind the surgery is often quite simple. If an area of brain can be identified that is the sole or major source - known as the "focus," or foci when there is more than one - of the electrical disturbances of seizures and if this area can be safely removed without causing damage to the patient, the opportunity exists for a surgical cure of this disabling problem.

The most common region in the brain for surgically removable sites of epilepsy is the inner, or medial, part of one temporal lobe, an area located just above the ear. The major area in question is called the hippocampus, or "sea horse," because it is shaped like this animal. Seizures originating from this area often cause lapses in awareness and behavior, called partial complex seizures. When these spells can be proved to originate in only one temporal lobe, surgical cure is frequently possible.

Abnormal sites in other brain can also be removed, and, in some rare instances, one entire hemisphere of the brain is removed for control of severe intractable epilepsy. This very dramatic operation, called a hemispherectomy, is reserved for cases where the abnormal or damaged tissue in the hemisphere is found to be quite widespread. Because of this damage, the patients almost always have weakness on one side of the body before the surgery. What is remarkable is that, in young patients, up through five or six years of age, despite surgical removal of the entire motor cortex, which controls voluntary movement, leg strength (usually including the ability to walk), arm movements (but not fine finger dexterity), and the ability to speak generally recover. This is a striking example of "functional plasticity" - the power of nerve cells to reorganize their functional connections to take over for lost tissue - in the brain of the young child

Frequently in pediatric epilepsy, no definable focus of seizure generating tissue can be isolated for removal. One surgical option available in such cases is to decrease the total number of nerve fiber connections between one side of the brain and the other, by splitting the midline of the brain in an operation called a "corpus callosotomy." Patients who have undergone this operation may have difficulties transmitting sensory information from one side of the body to the other, but they can lead remarkably normal lives. The type of seizures most helped by corpus callosotomy is the "drop attack," where patients suddenly and unexpectedly become unconscious and fall to the floor. Many children who come in for this operation have many cuts and abrasions on their heads from these violent falls, which can be brought under good control with this unusual operation.

What limits the use of surgical treatment?
For the majority of patients, drugs may remain the best choice. But severely intractable cases of epilepsy continue to seize and many are never evaluated for surgery treatments. In some cases, the limiting factor may be lack of access to a surgical center; for others, the advantages of surgery seem small because the seizures do not seem to originate from a small discrete focus. Alternatively, the area that apparently would need to be removed may overlap with important functional areas. The situation is similar to the problem of determining boundaries between countries or real estate plots: Lack of overlap is key.

To consider possible scientific advances for the future, we can concentrate on this one general problem of overlap between an area where seizures are presumed to arise (called an "epileptogenic" area) and a functionally eloquent area. A schematic representation of such a situation is shown in the drawings on page 4. The surgeon's strategy in planning the operation depends on the prospects for drawing a boundary line between what is to be removed and what is to be left in place.

Solving the overlap problem
At least three types of strategies may be employed to respond to the problem of overlapping "functional" and "epileptogenic" tissue.

The first is to better define the boundaries on the map between the areas which must be saved and those which may be removed. Refinement of surgical techniques will require more sophisticated data collection, with the hope that precision in mapping alone may resolve some of the boundary disputes. Such techniques may reveal that there is less overlap than the less precise mapping procedure suggests, and such a straightforward improvement in the resolution of the domains of the tissue could sometimes reveal that surgery is less apt to cause functional damage than it first appeared.

The second possibility is to exploit changes in the brain itself which may occur following a tissue removal, and may allow functional activity to re-emerge with an altered map afterward. Thus, it may prove "safe" to remove functional tissue, if it can be expected the brain's "functional plasticity" will compensate for loss of the tissue in question.

The third approach is to use surgical strategies short of actual removal of neurons to diminish the likelihood of seizure generation from an epileptogenic region.

Sharper maps
All of our maps of functional areas, and in most cases our maps of "hot" areas of seizure focus, are necessarily approximations. Whether a given area of brain is necessary for a particular task may be difficult to determine, and may depend on the technique used to generate the cortical map. For example, most maps of cortical function (in patients who can cooperate with this testing) involve electrical stimulation of small regions of exposed brain during surgery on the awake patient. The electrical current "turns off" small areas, allowing the mapping of functions.

Alternatively, functional areas have been mapped by sampling electrical activity of individual neurons during surgery, a technique called single-unit recording, which is like listening in on the electrical signals from single neurons without stimulating them. The signal can be correlated with behavior, giving another map. Single unit recording maps vary tremendously from maps in the same patient made by the technique of cortical stimulation, as shown by Dr. George Ojemann and colleagues at the University of Washington. It is intriguing that these maps are different.

Optical imaging is an even newer technique of mapping during surgery. It uses high-speed, computer enhanced photography to capture extremely subtle changes in the brain's reflection of colored light as nerve cells become active. Dr. Michael Haglund, now at the Duke University, developed this technique while he was a research fellow in the laboratory of Dr. Gary Blasdell at Harvard Medical School, then later applied the technique to humans with Ojemann in Seattle. Preliminary results with this kind of imaging have been very encouraging that even better maps may eventually be made. Functional magnetic resonance imaging may yield a non-operative means for sharpening the map boundaries as well.

Exploit plasticity
Alternatively, a better understanding of functional plasticity in the brain could allow us to remove tissue in the immature brain when we might expect restitution of function. What are the limits of functional reorganization, especially in the immature brain? It is clear from studies of young animals that some functions can return after lesions that would be expected to devastate adult animals. For example, complete hemispherectomy can be followed by partial behavioral recovery of visual fields in kittens, only if done at the right age.

The critical information for determining what surgery is safe or not safe in children would be a detailed understanding of the limits of this kind of functional recovery. In practice, clinical guidelines for optimal upper limit of age to allow motor recovery after hemispherectomy, for example, are based on experience but not fundamental understanding of functional plasticity. Epilepsy neurosurgery in children would be a more common treatment if individualized, predictive measures could be made of the functions able to survive the removal of apparently functionally important tissue.

Alternative surgical strategies
Even with optimal mapping and the most careful estimates of functional recovery, there will still be situations where the zone of epileptogenesis overlaps significantly with regions where removing brain tissue would seem likely to cause a deficit. Frank Morrell, of Rush Presbyterian St. Luke's Medical Center in Chicago, pioneered a procedure involving multiple small cuts into the cortex to deal with this problem. It appears that this procedure can prevent lateral spread of the seizure over the cortex and leave function intact. By isolating regions of the cortex from one another, the effect is of "firewalls" that restrain the generation of the critical amount of epileptic activity to cause a spreading seizure.

Finally, a general strategy would be to globally affect nerve cell communication of seizure signals and decrease the overall tendency to seize in an area without actually removing it or lesioning it at all. Obviously this is what happens when antileptic drugs are employed. But surgical attempts to control seizures may potentially use the same strategy. For example, electrical stimulation of the vagal nerve, an approach now in clinical trials and shows some degree of seizure relief to a minority of patients with seizures, seems to work this way, very much like an antiepileptic drug.

While repetitive electrical stimulation may change oscillatory circuits in various ways, a more refined approach has recently been suggested by Dr. Steve Schiff and colleagues at National Children's Hospital in Washington, D.C. If the sites where seizures arise are considered to be chaotic electrical generators, it may be possible to control the chaotic activity with precisely timed single impulses. This has been done in laboratory experiments, but the implications for the control of chaotic electrical behavior may extend to otherwise inoperable seizure sites. Obviously, this type of approach would require considerable theoretical and practical advances before application in children.

Clearly, basic research in brain functional anatomy and brain plasticity has been of great value in the treatment of epilepsy in children. Future work should pay off more handsomely in the future. These gains require surgical imagination, significant resources and some risk. But a lifetime of pain and distress for a youngster with intractable epilepsy is also risky, expensive and, one can hope, avoidable. *

Dr. Madsen is a neurosurgeon at Children's Hospital in Boston.