Parameter-based analysis of clinical efficacy of combined bed nucleus of the stria terminalis–nucleus accumbens deep brain stimulation for treatment-resistant depression

OBJECTIVE Treatment-resistant depression (TRD) is a severely disabling psychiatric condition that responds poorly to conventional treatments. Deep brain stimulation (DBS) has been proposed for the treatment of patients with TRD in numerous studies. Several deep brain nuclei are considered as potential targets for TRD-DBS, but their clinical efficacy needs further validation. This study carried out dual-target combined stimulation of the bed nucleus of the stria terminalis (BNST) and nucleus accumbens (NAc) to investigate the effectiveness of the treatment for TRD patients. METHODS An 8-contact DBS electrode was used in the study with a surgical path that crossed the BNST and NAc targets. Stimulation parameters and the corresponding severity of symptoms evaluated by the 17-item Hamilton Depression Rating Scale (HAMD-17) and other scales were obtained at each follow-up. The accuracy of electrode positions, the effect of combined stimulation, and the corresponding stimulation parameters were evaluated. Sweet spot prediction models were used to assess the effective stimulation sites in the treatment. RESULTS The study included 23 TRD patients undergoing DBS at a single center from March 2021 to May 2023. At the last follow-up (range 4–24 months), 14 patients had responded to the treatment (HAMD-17 score improved ≥ 50%), 7 of whom had achieved clinical remission (HAMD-17 score ≤ 7). Electrode position analysis suggested that the BNST may be more important for the improvement of depressive symptoms than the NAc. Overlapped volumes of volume of tissue activated (VTA) and BNST were significantly correlated with absolute (ρ left = −0.377, p < 0.001; ρ right = −0.251, p < 0.001) and percent (ρ left = −0.249, p < 0.001; ρ right = −0.098, p = 0.102) changes in HAMD-17 score. The sweet spot model of HAMD-17 improvement also suggested that the VTA overlap with the dorsal side of BNST was associated with the impact on depressive symptoms (t = −4.10, p < 0.05). CONCLUSIONS Combined BNST-NAc stimulation of TRD can effectively improve depressive symptoms, in which the BNST seems to have a dominant therapeutic effect. The results of this study not only help to optimize the DBS programming parameters, but also offer an opportunity to further understand the differences between the two targets. In the future, larger prospective cohorts are needed to verify the results of combined BNST-NAc DBS.