Cluster analysis: Predicting the seizure outcome in temporal lobe epilepsy

Objectives: We applied the Two-Step cluster analysis on a large cohort of patients with temporal lobe epilepsy (TLE). We hypothesized that there are distinct clusters of patients with TLE based on their clinical characteristics and these clusters may predict their seizure outcome. Methods: This was a longitudinal study of a prospectively developed database. All patients with a diagnosis of TLE were studied at the outpatient epilepsy clinic, Shiraz, Iran, from 2008 until 2021. The Two Step cluster analysis (Schwarz's Bayesian Criterion: BIC) was applied to the whole dataset considering the demographic data, clinical characteristics, imaging, and electroencephalography data. The seizure outcome was compared between the clusters of patients. Results: Three hundred and seventy-four patients had the inclusion criteria and were studied. The Two Step cluster analysis showed that there were two distinct clusters of patients with TLE. The most important clinical predictors were the presence (or absence) of focal impaired awareness seizures or focal to bilateral tonic-clonic seizures, aura with seizures, and the brain imaging findings. The seizure outcomes were significantly different between these two clusters (p = 0.008). Conclusion: The Two-Step cluster analysis could identify two distinct clusters of patients with TLE; these data are helpful in providing prognosis and counseling for patients and their care-givers. These data may also be used to develop a practical outcome prediction tool for patients with TLE. (c) 2021 Elsevier Inc. All rights reserved.

Automated summary

By applying Two-Step cluster analysis on a cohort of patients with temporal lobe epilepsy, it was found that there are two distinct clusters of patients with different seizure outcomes based on their clinical characteristics. These findings can be used to develop a practical outcome prediction tool for patients with temporal lobe epilepsy.