Abstracts

Rationale: Invasive surgery is often the only curative option for patients with drug-resistant epilepsy, but predicting post-surgical seizure freedom remains challenging. Factors such as the procedure type, brain region, pathology and demographics must be considered. Here, we conducted a meta-analysis of the literature to quantify the rates of surgical success in a large dataset of patients with drug-resistant epilepsy.


Methods: We screened 3705 publications to identify relevant patient cohorts for this meta-analysis (described in Fig. 1). A predefined protocol was used to identify and eliminate duplicates, resulting in a dataset of 7026 unique patients. This dataset was used to determine the chances of seizure freedom (defined as Good Outcome: Engel I or ILAE 1-2, or Seizure-Free: Engel 1A or ILAE1) at the lobar level. Results are reported as the percentage of patients from the entire cohort who meet both the Good Outcome and Seizure-Free definitions. Odds ratios (OR) and post hoc statistical tests are provided for specific regions and variables. Statistics between groups were performed using parametric t-tests, and effect sizes were calculated using Cohen’s d.

Results: Our cohort had a balanced sex distribution (F=2682, M=2833, not reported=1511, p >0.05) and a wide age range from 6 months to 80 years old (mean age=24.09±15.63; not reported=701). Overall, 62.2% (4370/7026) achieved good outcomes and 40.9% (2556/5717) were seizure-free with a median follow up time of 24 [2-463] months. Intracranial electroencephalogram (iEEG) investigations were performed in 1654 (24%) patients, with 54% chance of good outcomes and 41% chance of being seizure-free for stereo-EEG (964), and 54% chance of good outcome and 44% chance of being seizure-free for electrocorticography (690), both lower than the overall cohort.

Focusing on patients that had unilobar surgical treatment, 52.2% (1463/2800) had good outcomes and 36.1% (1006/2784) were seizure-free. The chances of good outcomes and seizure freedom were: 55% and 36% for temporal lobe epilepsy, 43% and 37% for frontal lobe epilepsy, 42% and 30% for parietal lobe epilepsy, 35% and 26% for occipital lobe epilepsy, and 65% and 54% for insular lobe epilepsy. No lobe showed an interaction with sex or age (p >0.05), although younger patients had a slightly higher chance of seizure freedom (p< 0.01, d=0.18 and 0.08). A subset of patients underwent minimally invasive surgery (n=340), mainly in the temporal lobe, with 45% chance of good outcomes and 31% chance of seizure freedom (p >0.05) (Fig. 2B).

When considering pathology, patients with hippocampal sclerosis had a 64% chance of good outcomes and 48% chance of being seizure-free, compared to 63% and 47% for tumors, 55% and 50% for vascular, as well as 43% and 37% for malformation of cortical development (Fig. 2C-E).


Conclusions: The rate of post-surgical success for patients with drug-resistant epilepsy varies based on the brain region, surgical approach, and pathology. By conducting this meta-analysis, clinicians and patients can acquire a more data-driven understanding of epilepsy surgery outcomes.

Funding: Savoy Foundation, CIHR