Physiologic and cortical response to acute psychosocial stress in left temporal lobe epilepsy - A pilot cross-sectional fMRI study

•Following acute psychosocial stress, HPA axis reactivity is higher in patients with LTLE than in HCs.•fMRI activation to stressful negative feedback in patients with LTLE is attenuated compared with HCs.•HPA axis reactivity but not stress perception is associated with seizure frequency.•There is a correlation between seizure frequency and medial prefrontal activation.

Stress is commonly reported as a seizure precipitant in individuals with poorly controlled seizures including temporal lobe epilepsy. The aim of the study was to assess the neural and physiologic correlates of psychosocial stress response during functional magnetic resonance imaging (fMRI) and their relationship with seizure occurrence in patients with left temporal lobe epilepsy (LTLE). We enrolled 23 patients with LTLE and 23 age- and sex-matched healthy controls (HCs); all underwent fMRI with control math task (CMT) and stress math task (SMT) and pre-/post-fMRI salivary cortisol analysis (acute stress reactivity calculated as % reduction from post-stress to recovery baseline; dCORT). The Beck Depression Inventory-II (BDI-II) and Perceived Stress Scale (PSS-10) were administered. T-tests of performance and cortisol variables were performed. Processing and single-subject modeling of fMRI response to CMT positive feedback and SMT negative feedback, group comparisons, and whole-brain correlation of seizure occurrence and fMRI response in patients with poorly controlled LTLE were performed. Patients with LTLE and healthy controls were similar in demographics, math performance, heart rate, and PSS-10 scores (all p > 0.05). Patients with LTLE exhibited greater dCORT (p = 0.048) and lower BDI-II scores (p = 0.016) compared with HCs. Patients with poorly controlled LTLE showed a positive association between seizure frequency and dCORT (r = 0.73, p = 0.016). Functional MRI activation to feedback was similar between groups, including midfrontal, temporal, parietal, and occipital regions. Regression analyses revealed no group differences to positive feedback, but, compared with HCs, patients with LTLE showed decreased activation to negative feedback in the left cerebellum/middle occipital/fusiform gyri, left hippocampus/parahippocampus, bilateral medial frontal/cingulate/superior frontal gyri, right postcentral gyrus/inferior parietal lobule, and right insula/postcentral gyrus (p < 0.05, corrected). Patients with poorly controlled LTLE showed negative association between seizure frequency and activation in the bilateral subgenual anterior cingulate (p < 0.05, corrected). This study is the first to characterize the cortical and physiologic responses to acute psychosocial stress and to show a significant relationship between seizure control in LTLE and both the hypothalamic-pituitary-adrenal axis and fMRI signal reactivity to acute psychosocial stress. These findings extend our understanding of the complex interplay between stress, physiologic stress markers, and seizures/epilepsy.