We define intractable in the first 5 years of epilepsy treatment as an
average of at least one seizure every 2 months. For the longer term,
we define intractable as at least one seizure per year. Population stu
dies from Chicago, IL, U.S.A., Finland, and Nova Scotia, Canada indica
te that with long follow-up, many children with intractable epilepsy e
ventually have remission of their seizure disorder. Epilepsy is no lon
ger intractable when the seizures stop completely. How often does a ne
w antiepileptic drug (AED) render a child seizure-free when one or mor
e AEDs have failed? Literature on adults with epilepsy suggests that f
ew with chronic epilepsy who have not achieved seizure control with se
veral AEDs will achieve complete seizure control with additional AEDs.
The Nova Scotia study suggests that if a child's seizure fails to be
controlled with a first AED, there is an increased risk of intractable
epilepsy. Nonetheless, the chance of eventual, complete remission of
epilepsy (seizure-free without AED treatment) is approximately 40%. We
conclude that intractability should not be considered until there has
been failure of at least three first-line AEDs. Intractable epilepsy
is rare. Careful definition of the characteristics of children with in
tractable epilepsy who do respond completely to new AEDs will likely p
rovide the only rational approach to treatment of children with three
drug failures. Collaboration by multiple epilepsy centers will be requ
ired to gain this information.