PURPOSE: Cryptogenic focal epilepsy (CFE) is a heterogeneous clinical disorder including patients with severe refractory forms and patients with a fairly good prognosis. Predictors of prognosis in CFE are poorly understood. The aim of this retrospective study is to identify long-term (5-year) prognostic predictors in patients with newly diagnosed CFE. METHODS: Subjects with cryptogenic focal epilepsy (CFE) seen from April 1987 to September 2011 in two twin Epilepsy Centres located in Reggio Calabria and Catanzaro, Calabria, Southern Italy, were screened. Patients were excluded if they had psychogenic seizures, major psychiatric disorders presence of brain lesions except for non-specific white matter T2-hyperintensities, short follow-up (less than five years) or for having received the diagnosis of CFE elsewhere. One hundred and eighty-six patients, firstly diagnosed in our Centres, constituted the study sample. Survival curves were generated according to the Kaplan-Meier method and compared with the log-rank test. The endpoint was the cumulative time-dependent chance of 5-year remission after treatment start. Independent predictors of remission were tested by multivariate analysis using Cox proportional hazards function models. The accuracy of the resulting model was tested with Receiver Operating Characteristics (ROC) curve analysis. KEY FINDINGS: The cumulative incidence of remission was 23%. At Kaplan-Meier analysis, the only factor predicting remission was family history of epilepsy or febrile seizures (FS; p=0.02). At Cox regression, family history and frontal lobe epilepsy showed to be independent predictors of outcome (p=0.02 and 0.03, respectively). The accuracy of these predictors was good (area under ROC curve 0.648, 95% CI 0.575-0.716). Interestingly, we also found a considerable (7 years) diagnostic delay that did not result in a worse prognosis. SIGNIFICANCE: About one quarter of subjects with newly diagnosed CFE attains 5-year seizure remission during follow-up. Family history of epilepsy or FS and frontal localization are independent prognostic predictors.
PURPOSE: Cryptogenic focal epilepsy (CFE) is a heterogeneous clinical disorder including patients with severe refractory forms and patients with a fairly good prognosis. Predictors of prognosis in CFE are poorly understood. The aim of this retrospective study is to identify long-term (5-year) prognostic predictors in patients with newly diagnosed CFE. METHODS: Subjects with cryptogenic focal epilepsy (CFE) seen from April 1987 to September 2011 in two twin Epilepsy Centres located in Reggio Calabria and Catanzaro, Calabria, Southern Italy, were screened. Patients were excluded if they had psychogenic seizures, major psychiatric disorders presence of brain lesions except for non-specific white matter T2-hyperintensities, short follow-up (less than five years) or for having received the diagnosis of CFE elsewhere. One hundred and eighty-six patients, firstly diagnosed in our Centres, constituted the study sample. Survival curves were generated according to the Kaplan-Meier method and compared with the log-rank test. The endpoint was the cumulative time-dependent chance of 5-year remission after treatment start. Independent predictors of remission were tested by multivariate analysis using Cox proportional hazards function models. The accuracy of the resulting model was tested with Receiver Operating Characteristics (ROC) curve analysis. KEY FINDINGS: The cumulative incidence of remission was 23%. At Kaplan-Meier analysis, the only factor predicting remission was family history of epilepsy or febrile seizures (FS; p=0.02). At Cox regression, family history and frontal lobe epilepsy showed to be independent predictors of outcome (p=0.02 and 0.03, respectively). The accuracy of these predictors was good (area under ROC curve 0.648, 95% CI 0.575-0.716). Interestingly, we also found a considerable (7 years) diagnostic delay that did not result in a worse prognosis. SIGNIFICANCE: About one quarter of subjects with newly diagnosed CFE attains 5-year seizure remission during follow-up. Family history of epilepsy or FS and frontal localization are independent prognostic predictors.