Abstracts

Rationale: The default mode network (DMN) is a network of brain regions that are spontaneously active when an individual is at wakeful rest. The DMN has a central hub of the posterior cingulate cortex which recruits the ventral medial prefrontal cortex, inferior lateral parietal cortex, and the bilateral hippocampi. Frings and colleagues (2009) showed reduced connectivity between posterior cingulate and left hippocampus in patients with left mesial temporal sclerosis (MTS), although this finding did not extend to right-sided disease. Given intersubject variability in the strength of DMN functional connectivity, we propose an interhemispheric difference in hippocampal DMN recruitment in TLE. Specifically, we expect the affected hippocampus to be less strongly incorporated into DMN than the unaffected hippocampus; and this trend to exist in TLE patients but not healthy control participants. Methods: 15 TLE patients (age= 38 -14 years; 7 male; 5 [33%] with unilateral hippocampal sclerosis) and 18 healthy control subjects (age= 36 -12 years; 5 male) underwent a resting state fMRI scan. During the resting state scan, participants were instructed to clear their minds of any specific thoughts and passively view a fixation cross. Regions of interest (ROIs) were defined as 6mm radius spheres for the midline posterior cingulate (PCC), midline medial prefrontal cortex (MPFC), ipsi- and contralateral hippocampus (iHPC, cHPC), and ipsi- and contralateral inferior parietal (iPAR, cPAR). For controls, right hemisphere was arbitrarily picked as ipsilateral. Each ROI s correlation (r) with PCC was calculated and Fisher z-transformed into linear variables. One tailed t-tests determined if ipsilateral ROIs were less correlated with PCC than contralateral ROIs. Results: The difference of [r(PCC-iHPC) minus r(PCC-cHPC)] was significantly less than 0 for TLE patients (p=0.042) but not for control participants (p=0.714)(Figure 1). No other default mode component showed this group difference; the difference of [r(PCC-iPAR) minus r(PCC-cPAR)] was not significant for either group (Figure 2), and r(PCC-MPFC) did not differ between groups (not shown). Boxplots for these correlations by group are depicted in Figures 1 and 2; bars, diamonds and circles respectively indicate median, mean, and outlier values. Conclusions: We report diminished DMN recruitment of HPC in TLE patients. Only one-third of patients presented with MTS, suggesting that diminished HPC recruitment does not solely stem from anatomic irregularities. Our finding of diminished hippocampal connectivity irrespective of hemisphere may stem from our usage of a resting-state task, which we believe is a more ecologically valid measure of DMN functional connectivity than task-related deactivations. Given the absence of hippocampal sclerosis in most of our patients, we believe the reduced PCC-HPC connectivity cannot be attributed to a purely anatomical etiology.