*3.2. Nucleus Accumbens*

Another brain region involved in MDD is the nucleus accumbens (NAcc), part of the mesolimbic dopaminergic circuit involved in di fferent cognitive functions such as motivation and reward [33] (Table 2). DBS of the NAcc exerts immediate and long-term positive clinical e ffects in TRD and has been shown to significantly improve depression scores within one week [33]. Visualized with PET–computed tomography (PET-CT or PET/CT), NAcc-DBS increased metabolic activity in the ventral striatum, dlPFC, dorsomedial PFC (dmPFC), cingulate cortex, and the amygdala. Furthermore, metabolic activity in the vmPFC, ventrolateral prefrontal cortex (vlPFC), dorsal caudate nucleus, and part of the thalamus were decreased. Targeting the NAcc was essential for the e ffect of DBS on anhedonia (i.e., the inability to feel pleasure) in patients su ffering from TRD. However, when Schlaepfer et al. (2008) looked at single items of depression rating scales, capturing aspects of anhedonia such as 'work and activities', 'apparent sadness', and the 'inability to feel', no significant improvements were found following NAcc-DBS. A follow-up study showed a 50% response rate in 10 patients su ffering from TRD undergoing NAcc-DBS after 10 months [53]. In a more recent study reporting the long term e ffects of NAcc-DBS, 45% of TRD patients (*n* = 11) were classified as responders with a 50% reduction in HDRS scores after 12 months of stimulation, which remained until the last follow-up of 4 years [54] (Table 1). Several side e ffects were reported, such as seizure, agitation, and a transient increase in anxiety. In addition, one attempted suicide and one completed suicide were reported, for which the relation with the DBS treatment is uncertain.

### *3.3. Ventral Capsule*/*Ventral Striatum*

The ventral capsule/ventral striatum (VC/VS) is thought to be hyperactive in MDD [36] (Table 2). Capsulotomy (i.e., lesioning) of the VC/VS improved not only OCD symptoms but also depressive symptoms, inspiring stimulation of the VC/VS for TRD [15]. In an open-label trial that stimulated the VC/VS in 15 TRD patients, responder rates at three months, six months, and 12 months were 53.3%, 46.7%, and 53.3%, respectively, using the MADRS as an outcome measure, and were 46.7%, 40%, and 53.3%, respectively, using the HDRS as an outcome measure [55]. Adverse events ranged from pain or discomfort at the incision site, to hypomania, mixed bipolar state, and increased depression due to battery depletion.

The first RCT of DBS of the VC/VS for TRD was performed by Dougherty et al. (2015) who stimulated 30 patients for 16 weeks. There were no significant di fferences in response rates between the intervention and sham group in the double-blind phase [68,69]. Another RCT of VC/VS DBS in eight TRD patients was discontinued after an interim futility analysis of active vs. sham stimulation showed no di fference in e ffects between the two groups after 16 weeks. These results were never published but were discussed by Rezai et al. [70].

### *3.4. The Ventral Part of the Anterior Limb of the Internal Capsule*

The anterior limb of the internal capsule (ALIC) is another brain region that was initially studied for DBS in OCD. One study aimed at stimulating the NAcc discovered that most treated OCD patients (9 out of 16) actually received DBS in the ventral part of the ALIC (vALIC), which improved obsessive compulsive scale scores, showed anti-depressive e ffects, and led to the clinical implementation of vALIC-DBS in TRD [29]. DBS of the vALIC has also been associated with a decreased metabolism in the OFC, subgenual anterior cingulate cortex, and right dlPFC [71–73] (Table 2).

The first RCT of DBS of the vALIC for TRD was conducted by Bergfeld et al. (2016), investigating 25 TRD patients during a 52 week open-label trial, which resulted in a significant decrease in HDRS scores in the whole group during the optimization phase, although overall HDRS scores were still in the depression range (22.2 at baseline vs. 15.9 after optimization phase). Ten of the 25 patients could be classified as responders, with a more than 50% decrease on the HDRS. After the optimization phase, a RCT with a cross-over design including nine responders and seven non-responders ensued and showed a significantly lower score in the active DBS phase compared with the sham DBS phase (mean HDRS score of 13.6 (95% CI; 9.8–17.4 vs. 23.1 (95% CI; 20.6–25.6)) (HDRS < 0.001). However, the scores on the HDRS in the active treatment group were still within the mild to moderate depression range [74]. Both crossover phases lasted approximately 21 and 18 days, respectively.
