Abstracts

Rationale: There is an extensive individual variation in memory outcome after temporal lobe resection (TLR) for epilepsy. Considerable efforts have been made to identify specific risk factors for postoperative memory decline. The aims of this study were to describe individual verbal memory changes after TLR and to explore risk factors for postoperative early and late verbal memory decline. Methods: Fifty patients who had undergone TLR (23 in the speech dominant temporal lobe, DTL; 27 in the non-dominant temporal lobe, NDTL) were tested preoperatively, 2 and 10 years postoperatively. Twenty-three healthy controls were assessed at baseline and at corresponding test intervals. Learning and delayed recall of a wordlist and immediate and delayed recall of word-pairs were used for assessing verbal memory. On the basis of how many of the four verbal memory variables that had changed according to cut-off scores of the reliable change indices (RCI) 90% confidence interval (CI) of the controls, individual changes were categorized as follows: 'Decline?1 (?1 negative and no positive change); Decline?2 (?2 negative and no positive changes); Improvement?1 (?1 positive and no negative change); Improvement?2 (?2 positive and no negative changes). Results: In the whole patient group fewer patients had Decline?2 at 10-year (14%) compared with 2-year follow-up (28%). This held true both for the DTL (10-year: 26.1%, 2-year: 43.5%) and the NDTL group (10-year: 3.7%, 2-year: 14.8%). A similar pattern was seen for the NDTL group regarding Improvement?2 (10-year: 29.6%, 2-year: 18.5%). Five potential prediction variables were selected based on earlier findings and availability of data in the patient series: laterality; baseline verbal memory; baseline verbal IQ; baseline chronological age and cortical dysgenesis in the resected tissue. Intact baseline verbal memory (p=0.004), DTL resection (p=0.009), and older age at baseline (p=0.017) were identified as predictors of verbal memory Decline?1 2 years after surgery, while only DTL resection was a predictor at 10 years (p=0.001). Conclusions: Our findings indicate an extensive individual variability in verbal memory outcome after TLR, but also a partial recovery for some patients in verbal memory function between 2 and 10 years after surgery. Factors that are important predictors for verbal memory decline at earlier stages may be less important at long-term follow-up. The only remaining risk factor for impaired verbal memory at the long-term follow-up was DTL resection.