Predicting verbal memory decline following temporal lobe resection for epilepsy

Area of Science:

• Neurology
• Neuropsychology
• Surgical Outcomes

Background:

• Epilepsy surgery, specifically temporal lobe resection (TLR), is a common treatment for refractory epilepsy.
• Verbal memory decline is a significant concern following TLR, impacting patients' quality of life.
• Accurate prediction of post-surgical memory decline is crucial for informed patient consent and rehabilitation planning.

Purpose of the Study:

• To develop and validate a predictive model for verbal memory decline after temporal lobe resection (TLR) in epilepsy patients.
• To provide individualized risk assessments for preoperative patient counseling.
• To enhance shared decision-making regarding epilepsy treatment options.

Main Methods:

• Utilized data from 110 epilepsy patients who underwent TLR between 1987 and 2011.
• Employed penalized regression and 20x10-fold cross-validation for model development.
• Included neuropsychological assessments conducted pre-surgery and two years post-surgery.

Main Results:

• The final prediction model incorporated four key predictors: surgical side, hippocampal involvement, preoperative verbal memory, and history of tonic-clonic seizures (TCS).
• The model achieved 82% accuracy in identifying patients with post-operative verbal memory decline.
• Overall model accuracy ranged from 70% to 85%, depending on risk threshold selection.

Conclusions:

• The developed model enables personalized predictions of verbal memory decline risk after TLR.
• Informed patients can make better treatment decisions.
• The model supports epilepsy surgery teams in preparing patients for rehabilitation.