Therapeutic deep brain stimulation targeting BNST-NAc circuit driven large-scale brain networks in treatment-resistant depression

Therapeutic deep brain stimulation (DBS) targeting the striatum shows promise for treatment-resistant depression (TRD), but its effects on large-scale brain network dynamics remain unclear. This study aimed to elucidate how DBS targeting the bed nucleus of the stria terminalis-nucleus accumbens (BNST-NAc) circuit modulates network dynamics, assessed via electroencephalography (EEG) microstates, and how these changes relate to local circuit activity. In a randomized, double-blind, crossover trial with 10 TRD patients, synchronous resting-state scalp EEG and NAc local field potential (LFP) data were collected during active and sham stimulation. Microstate analysis identified five consistent microstates (A-E). Compared to sham, therapeutic DBS increased the coverage and occurrence of microstates A and B. Transition probabilities involving A ⇌ C, E → B, and B → A were increased during active stimulation, whereas C ⇌ D transitions were reduced. Several of these changes, notably the reduced transitions between C and D, were associated with symptom improvements. Critically, time-locked analysis revealed that a significant increase in NAc gamma-band aperiodic activity specifically preceded transitions from microstate C to D, but not from D to C. These findings provide multi-scale mechanistic evidence that BNST-NAc DBS drives clinically relevant EEG microstate alterations in TRD patients, which are driven by specific gamma aperiodic activity patterns in the NAc. Combined EEG-LFP microstate signatures may therefore serve as valuable biomarkers for DBS treatment response.