# Do acute postoperative seizures predict epilepsy surgery outcome? a scoping review

## TL;DR

This review explores whether early post-surgery seizures can predict long-term outcomes for epilepsy patients and highlights the need for better classification methods.

## Contribution

The paper identifies key variables influencing postoperative seizure outcomes and emphasizes the need for a multifactorial predictive model.

## Key findings

- Variables like epileptogenic zone size and APOS frequency are linked to long-term seizure outcomes.
- Study heterogeneity and small sample sizes limit the certainty of associations between variables and outcomes.
- Ictal expression and timing of seizure recurrence are less relevant for predicting outcomes.

## Abstract

Acute postoperative seizures (APOS) are common phenomena following resective epilepsy surgery and can be categorized as running-down (RDS) or running-up seizures (RUS). This differentiation is made retrospectively, considering their classification is based on seizure recurrence. However, early differentiation of RDS from RUS may prevent unnecessary escalation of anti-seizure medication or reoperation. This review provides an overview of the available literature on variables influencing the evolution to RDS/RUS in patients exhibiting acute or early postoperative seizures.

A database search was performed addressing studies related to the running-down phenomenon and postoperative seizures in PubMed and Embase. Eligibility required a clear definition of acute or early postoperative seizures. Studies concerning any type of epilepsy surgery or pathology were accepted, excluding those related to high-grade malignancies.

The search yielded a total of n = 1,690 records. We included n = 21 studies with a total of n = 1,496 patients, which examined variables associated with long-term seizure outcome. Interictal epileptiform discharge presence/laterality, epileptogenic zone size, APOS frequency, and history of generalized tonic–clonic seizures, head trauma, or encephalitis were associated with seizure outcome. Ictal expression and timing of seizure recurrence appeared less relevant. However, these associations are uncertain due to conflicting results between studies, likely due to small sample sizes, a limited reporting of secondary variables, and heterogeneity in study population and methodology.

The variability in clinical outcome following APOS highlights the need for a refined classification of postoperative seizures. Future research should focus on constructing and validating a multifactorial model integrating EEG-derived variables, APOS frequency, and medical history to more accurately predict long-term seizure outcome following resective epilepsy surgery.

The online version contains supplementary material available at 10.1007/s00701-025-06486-8.