April 25, 2001
In Ground-Breaking Study, Scientists
at Penn Track Epileptic Seizures
Temporal Lobe Epilepsy Seizures Appear to Begin Hours
Prior to Clinical Symptoms
Scientists studying epilepsy
at the University of Pennsylvania Medical Center are
finding a pattern of human brain activity that indicates
the conditions triggering seizures can take hours to
develop.
The work points toward a method to short-circuit epileptic
seizures and convulsions before they strike, through
the use of implantable brain devices and medications,
according to Brian Litt, MD, author of the study. It
appears tomorrow (Thursday) in the April issue of the
journal Neuron.
"This study is part of a large collaborative effort
to control symptoms of a condition that dominates the
lives of otherwise healthy individuals by its dramatic
unpredictability," Litt said. "The potential
to use our findings to help people with poorly controlled
seizures world-wide is enormous."
About 50 million people throughout the world suffer
from epilepsy. Almost 25 percent have seizures that
are not controlled by any available therapy.
Although epilepsy is the most common neurologic disease
after stroke, its cause cannot be identified in a large
percentage of cases. Recently, scientists looking for
changes in the brain that predict seizures have had
some success using mathematically-based chaos theory.
But those studies have generally been limited in scope
-- they concentrate on a period of minutes prior to
seizures -- and are further limited by the difficulty
of applying an abstract theory to what actually happens
inside the brain.
Litt and his colleagues, on the other hand, relied on
the traditional method of measuring brain activity through
EEG (electroencephalograpy) readings for five epilepsy
patients who were being evaluated for surgery and had
therefore stopped taking anti-seizure medication. The
EEG readings tracked the patients for periods ranging
from four to 14 days.
Using electrodes implanted in both sides of the brain,
the scientists examined "a continuous stream of
data for reproducible patterns associated with seizures,
and found a chain of events that predicted that seizures
were going to occur," Litt said.
The researchers discovered cycles of abnormal brain
activity -- epileptic discharges -- lasting 15 to 30
minutes. The discharges became more frequent over a
period of hours as they led to brief, asymptomatic seizures
at specific points in the brain. Those smaller seizures
triggered a steady increase in activity that spread
across the brain and culminated in clinical seizures.
Litt likened this cascade of events to a match striking
over and over, lighting and re-lighting a fuse in the
affected part of the brain. The fuse goes out and re-ignites
more and more frequently, until finally it ignites the
energy that leads to clinical seizures. In some of the
study patients, the process lasted up to seven hours.
"This information provides a real opportunity to
stop abnormal activity in epileptic brain regions before
seizures develop," Litt said. Although substantial
work remains before the study findings can be put to
clinical use, he believes scientists may eventually
be able to implant devices in the brain that will abort
seizures by reacting to, and diffusing, the cycle of
increasing abnormal brain activity.
Litt collaborated in the study with Rosanna Esteller;
Javier Echauz, PhD; Maryann D'Alessandro; Rachel Shor;
and George Vachtsevanos, PhD, all of the Georgia Institute
of Technology, along with Thomas Henry, MD; Page Pennell,
MD; Roy Bakay, MD, and Charles Epstein, MD, of Emory
University. Marc Dichter, MD, PhD, of Penn also collaborated
in the study.
Along with Vachtsevanos, Echauz and Esteller, Litt is
co-founder of IntelliMedix, a company devoted to this
research.
The study was funded by IntelliMedix (in which Drs.
Litt, Vachtsevanos, Echauz and Esteller have a financial
interest), the Epilepsy Foundation, the Whittaker Foundation,
the American Epilepsy Society, the National Institutes
of Health and the University of Pennsylvania Research
Foundation.
# # #
|
