Deciding whether an accepted treatment makes a difference in the long run is a tricky business. Does this treatment actually work? There are a number of examples where this basic question has been raised in relation to commonly used, and accepted, treatments, such as the use of stents for coronary artery disease, or antidepressant agents. Today, I would like to discuss surgery for temporal lobe epilepsy.
People with epilepsy have seizures—the uncontrolled firing of nerve cells, or neurons. These seizures often start at a particular site in the brain, a “seizure focus,” and then spread to recruit other groups of nerve cells. The clinical manifestations depend on what part of the brain is involved: it may be sensory, with funny feelings travelling up the arm and then into the face; motor, with shaking of one side of the body; consciousness, with the sudden loss of awareness and falling to the floor, or abnormal smells, sounds, or recurring thoughts. It's common for the seizures to start in the temporal lobe. (If you think of the one half of the brain as shaped like a boxing glove, the temporal lobe would be the thumb.) Temporal lobe epilepsy accounts for about 40 percent of all epilepsy, and it's this form that has been a common target for surgery.
Surgery for epilepsy has a long history, popularized by Wilder Penfield and Herbert Jasper of the Montreal Neurological Institute in the 1940’s and 50’s. Despite its presumed efficacy, many still consider it a last resort; some epilepsy experts think it is severely underutilized. It is suggested that each year only 1,500 candidates, out of nearly 100,000 eligible patients, undergo surgery each year. Further, those who do have surgery are often well into their disease before surgery is performed, an average of 22 years from onset and 10 years after it is obvious that medical management is not working.
To counteract this reluctance of physicians and patients to choose surgery, it was necessary to design studies comparing surgery to medical management. There have been only two such studies, the first by a Canadian group published in 2001. This study involved 40 patients with surgery and 40 patients with continuing medical management. After one year, 58 percent of the surgery patients were without seizures, as opposed to only 8 percent in the non-surgical group. You would think that this data would tip the balance toward the greater use of surgery, but it is not clear that it has.
The second study was recently published in JAMA and is featured in a New York Times article (“Surgery for Epilepsy Gains Urgency in Trial). This study, the “ERSET trial” (Early Randomized Surgical Epilepsy Trial), was designed to operate on temporal lobe epilepsy patients within two years after it was clear that they had failed to respond to medical therapy. Led by Jerome Engel at UCLA, the investigators appeared to do everything right. They recruited 16 United States epilepsy surgical centers, restricted surgery to experienced epilepsy surgeons, used up-to-date imaging and monitoring techniques, and agreed on criteria for inclusion and for outcome. The surgery was carried out in a standardized manner. The only problem was that they couldn’t gather enough patients. The original goal was to have 200, but they stopped at 38 (23 with medical treatment, and 15 with surgery). Despite these small numbers, the results were impressive. In the second year after enrollment, 11 of the 15 surgery patients were without seizures; none in the medicine group were seizure-free.
Taken together, these two studies provide strong evidence that for those with temporal lobe epilepsy who have become unresponsive to medications, surgery should be used, and much sooner than has been the current practice.
Removing brain tissue through surgery to treat epilepsy is not the only option. In a future column, I will discuss alternatives, particularly responsive neurostimulation (RNS) and Gamma Knife radiosurgery. In the RNS system, electrodes are placed near the seizure focus. When the nerve cells in this focus begin to fire, the electrodes pick up this abnormal activity and send it to a monitoring device within the skull. The device then sends an electrical discharge back down to the same electrodes to break up the seizure activity before it can spread. It is a similar idea to what an implantable defibrillator does for the heart. Radiosurgery uses focused beams of radiation to deliver radiation to a seizure focus with surgical precision.
Non-medical approaches to controlling epilepsy are part of a very dynamic field, overlapping between engineering and neuroscience. Hang onto your hats, the field is just getting warmed up.
I received assistance in the preparation of this article from Drs. Brian Litt, a developer of responsive neurostimulation from the University of Pennsylvania, and my son, Guy McKhann II, an epilepsy surgeon at Columbia.