Epilepsy is a common, treatable neurologic disorder; however, 20 to 30% of patients with epilepsy have seizures that cannot be completely controlled by antiepileptic drugs. l As many as half of these patients are potential candidates for surgical treatment. 2 A recent advance in surgical treatment for epilepsy is the recognition that among these medically refractory patients are identifiable subgroups who have surgically remediable~ yndromes.~ These syndromes have a well-defined pathophysiology and natural history, can be identified by noninvasive diagnostic evaluation, and are resistant to pharmacotherapy, but have a high surgical cure rate. Because seizures and their behavioral consequences are progressive, early surgical intervention may prevent irreversible disability. The prototypic surgically remediable syndrome is mesial temporal lobe epilepsy (MTLE), the form of TLE associated with hippocampal~ clerosis.~ With modern techniques, 80 to 90% of patients with MTLE become seizure free after anteromesial temporal 10bectomy.~~~ Traditional arguments that antiepileptic drugs are usually ineffective, that the condition becomes worse over time, and that surgical treatment is underused because MTLE is a common disorder and surgery is performed relatively rarely are based predominantly on anecdotal evidence derived from patients referred for epilepsy surgery. In this issue, Semah et aL6 provide important new data on the prevalence and prognosis of specific epileptic disorders, including MTLE, encountered over a 7-year period in the epilepsy outpatient clinic of La Salp&tri&re Hospital in Paris. Their study of 2,200 patients, ages 16 years and older (median, 33), included all outpatient visits from 1990 through 1997. Patients underwent a detailed evaluation including high-resolution MRI, which by 1990 was capable of revealing abnormalities characteristic of hippocampal~ clerosis.~ J It is important to remember, however, that this was a hospital-based study, not a population-based study, and that most patients were tertiary referrals for difficult-to-treat epilepsy. Only 45% of the total patient population had been seizure free for the previous year, which is comparable to results obtained from other epilepsy centers. 8 Nevertheless, the data provide important new comparative information on the prevalence and prognosis of different types of epilepsy, particularly with respect to MRI-demonstrated abnormalities, and also support a view that some are progressive.

By recalculating the authors’ figures, among patients whose epilepsy could be so classified, 74% had localization-related epilepsy and only 26% had generalized epilepsy. This perhaps reflects the fact that most patients with benign idiopathic generalized epilepsies did not need to be referred to a specialized epilepsy unit. Among those with localization-related epilepsy, 66% had TLE, making it the most prevalent diagnosis in this cohort. Again recalculating the authors’ numbers, among patients for whom adequate MRI data existed, 35% of those with localization-related epilepsy (53% with TLE) had evidence of hippocampal sclerosis, making it the most prevalent lesion encountered in the population. Consequently, over a fourth of the total patients with MRI data had MTLE and undoubtedly there were additional patients whose hippocampal sclerosis was not apparent on MRI. These findings confirm the perception that MTLE is a very common epileptic syndrome and the one most often encountered in epilepsy centers.