# Long‐term memory trajectories in seizure‐free patients following epilepsy surgery for hippocampal sclerosis

## TL;DR

This study tracks memory changes over 10-15 years in patients who had successful epilepsy surgery for hippocampal sclerosis.

## Contribution

It reveals that memory improvements after surgery are often not sustained long-term, with disease progression outpacing brain plasticity.

## Key findings

- Patients with normal preoperative memory showed significant memory decline at 10-15 years post-surgery.
- Dominant-side surgery and younger age were strong predictors of long-term memory decline.
- The SAH surgical technique was linked to better verbal learning outcomes compared to ATL.

## Abstract

This study was undertaken to study long‐term memory trajectories over the years in patients with temporal lobe epilepsy and unilateral hippocampal sclerosis (TLE/HS) seizure‐free since surgery.

This cross‐sectional study included patients with TLE/HS from a single‐center epilepsy surgery program who had been seizure‐free for at least 10 years following anterior temporal lobectomy (ATL) or selective amygdalohippocampectomy (SAH). Memory performance was evaluated preoperatively (T1), 1–4 years postoperatively (T2), and 10–15 years after surgery (T3). Logistic regression evaluated variables correlated with memory function at each point in time. A reliable change index was performed to identify changes in individual measures.

A total of 54 patients were included, of whom 36 (66%) were male and 52 (96%) right‐handed. Patients with left HS and normal preoperative Rey Auditory Verbal Learning Test or Wechsler Memory Scale–Revised (WMS‐R) logical memory showed worsening at T2 (13% × 52%, p = .029; 0 × 31%, p < .015, respectively) and T3 (27% × 63%, p = .045; 22% × 81%, p < .001, respectively). Visual reproduction (WMS‐R) following nondominant surgery also deteriorated at T3 for patients who improved or sustained normal performance between T1 and T2 (33% × 50%, p = .64). The predictive factors for memory decline included normal preoperative memory function (odds ratio [OR] = 15, 95% confidence interval [CI] = 4.03–55.9, p < .001 for logical memory; OR = 1.5, 95% CI = 1.12–2.01, p = .007 for visual reproduction), younger age (OR = 1.2, 95% CI = 1.12–1.28, p < .001), dominant‐side surgery (OR = 3.66, 95% CI = 1.49–8.95, p < .01), and lower education level (OR = 8.74, 95% CI 1.77–43.2, p = .008). The SAH technique was associated with better long‐term verbal learning outcomes compared to ATL (OR = 3.02, 95% CI = 1.17–7.81, p = .02).

Memory preservation or improvement in the first few postoperative years is usually not sustained in the long term, suggesting that disease progression surpasses plasticity over the years.