One-year mortality in Bordeaux cohort: the value of syndrome classification

To evaluate the usefulness of the International Classification of Epilepsy Syndromes for 1-year mortality in a prospective incidence study of first epileptic seizures.

Date and cause of death from the treating physician in an incidence study of first afebrile seizures collected 15 years ago in Southwest France. Cases were classified by epilepsy syndrome. A total of 804 patients were included: acute symptomatic (n = 277), unprovoked (n = 439), or unclassifiable (n = 88).

One hundred and fifty-one patients died within the first year: none with idiopathic partial or generalized epilepsy, 16/104 with symptomatic (standardized mortality ratio (SMR) 6.4, 95% CI 3.6-10.3), 1/59 with cryptogenic (SMR 1.7, 95% CI 0.1-9.7 CI) partial epilepsy, 1/14 (SMR 28.1, 95% CI 0.4-156.6) with symptomatic/cryptogenic generalized epilepsy, 2/23 with undetermined epilepsy, 1/135 with isolated seizure (SMR 0.6, 95% CI 0.1-3.1), and 90/277 (SMR 10.3, 95% CI 8.3-12.7) with acute symptomatic seizures. Unclassifiable seizures could not be classified as acute symptomatic or unprovoked: associated with alcohol abuse (death: 3/32, SMR 7.7, 95% CI 1.6-22.6), brain tumors (death: 31/39, SMR 41.5, 95% CI 28.2-58.9), and dementia (death: 6/17, SMR 5.4, 95% CI 2.0-11.7). Most deaths were due to progression of underlying disease, only 5.9% were seizure-related.

Although a syndromic diagnosis is important for treatment decisions and some prognostic aspects of seizure disorders, its value in mortality studies is limited. Mortality can be calculated only at the first (partial, generalized, and undetermined epilepsies, and special syndromes) and the second (idiopathic vs. symptomatic or cryptogenic) levels of the International Classification of Epilepsies.