To determine the consistency of semiology over time in new-onset focal (NOF) and refractory focal epilepsy (RFE).

Semiology, the subjective description of signs/symptoms of a seizure, is an important tool to lateralize and localize seizures. Changes in semiology may be attributed to dynamic changes in epileptogenic networks due to disease progression, anti-seizure medication effects, or shifts in the patterns of seizure propagation networks. Understanding how semiology changes over time in patients with NOF and RFE can be insightful for epilepsy management and surgical workup. We hypothesized that RFE would exhibit a more complex semiology that would change over time compared to NOF.

We utilized data from the Human Epilepsy Projects (HEP) 1 and 2, prospective cohort studies of NOF and RFE, respectively. Baseline demographics, etiology, and semiology were examined. Participant-reported changes in seizure descriptions over time were collected and analyzed. Chi-square, Wilcoxon, and logistic regression tests, with significance at p=0.05, were used to analyze changes in semiology and its risk factors.

Among the 658 participants, the median age was 34.0 (IQR; 24.0-45.0) years, 58.2% were female, 78.4% White, 21.1% unemployed, and 74.5% had NOF. At baseline, 63.3% of NOF and 81.3% of RFE had ≥2 described semiology types, while 19.1% of NOF and 49.5% of RFE had ≥3 types. Among those with persistent seizures at follow-up, consistency of semiology was similar for both NOF and RFE (57.1% vs. 64.3%, p>0.05), with no significant time effect (P_{interaction (time,epilepsy type)}=0.930). Males were more likely to have a consistent semiology (OR:2.26, 95%CI: 1.04-4.93). This effect disappeared after age was added, with no significant interaction (P_{interaction (age,sex)} =0.514).

 We found that semiology was consistent at similar rates for NOF and RFE participants over time.  While males were more likely to have a consistent semiology, this effect disappeared when age was added to the model.