Leveraging the resting brain to predict memory decline after temporal lobectomy

Abstract

ObjectivesAnterior temporal lobectomy as a treatment for temporal lobe epilepsy is associated with a variable degree of postoperative memory decline, and estimating this decline for individual patients is a critical step of preoperative planning. Presently, predicting memory morbidity relies on indices of preoperative temporal lobe structural and functional integrity. However, epilepsy is increasingly understood as a network disorder, and memory a network phenomenon. We aimed to assess the utility of functional network measures to predict postoperative memory changes. MethodsPatients with left and right temporal lobe epilepsy (TLE) were recruited from an epilepsy clinic. Patients underwent preoperative resting-state fMRI (rs-fMRI) and pre- and postoperative neuropsychological assessment approximately one year after surgery. We compared functional connectivity throughout the memory network of each patient to a healthy control template based on 19 individuals to identify differences in global organization. A second metric indicated the degree of integration of the to-be-resected temporal lobe with the rest of the memory network. We included these measures in a linear regression model alongside standard clinical and demographic variables as predictors of memory change after surgery. ResultsSeventy-two adults with TLE were included in this study (37 left/35 right). Left TLE patients with more abnormal memory networks, and with greater functional integration of the to-be-resected region with the rest of the memory network preoperatively, experienced the greatest decline in verbal memory after surgery. Together, these two measures explained 44% of variance in verbal memory change (F(2,31)=12.01, p=0.0001), outperforming standard clinical and demographic variables. None of the variables examined in this study were associated with visuospatial memory change in patients with right TLE. ConclusionResting-state connectivity provides valuable information concerning both the integrity of to-be-resected tissue as well as functional reserve across memory-relevant regions outside of the to-be-resected tissue. Intrinsic functional connectivity has the potential to be useful for clinical decision-making regarding memory outcomes in left TLE, and more work is needed to identify the factors responsible for differences seen in right TLE.