*3.6. Thalamic Peduncles*

The inferior thalamic peduncle (ITP) is a bundle of fibers connecting the OFC to the thalamus. The OFC is thought to play a role in the non-reward tractor theory of depression, where the orbitofrontal non-reward system is more easily triggered in depression, causing negative emotional states [77] (Table 2). Stimulating the ITP could disrupt this enhanced triggering and lead to less depressive symptoms. ITP stimulation for OCD has already shown improvements of the score on the Yale-Brown Obsessive-Compulsive scale in five OCD patients [45]. A case study in one TRD patient reported that DBS of the ITP decreased depressive symptoms [78]. However, within this study, two brain regions were investigated, the second being the BNST.

### *3.7. Bed Nucleus of the Stria Terminalis*

The BNST is involved in a range of behaviors, such as stress response, social behavior, and extended duration of fear states. This nucleus assesses sensory information from the environment, coupled together with the subjects current mood and arousal, integrating a proper response to environmental and social setting changes [22] (Table 2). Raymaekers et al. (2017) indicated that both BNST and ITP stimulation could alleviate depressive symptoms; however, due to a small sample size, no statistical analyses were conducted [78].

### *3.8. Medial Forebrain Bundle*

The medial forebrain bundle (MFB) is a fiber tract connected to various parts of the limbic system thought to play a role in reward-seeking systems [21] (Table 2). In one trial, DBS of the superolateral branch of the MFB resulted in more than a 50% decrease in depressive symptoms in six out of seven TRD patients within seven days [47]. An additional interim analysis of MFB-DBS in TRD confirmed

these findings, showing more than a 50% decrease in depressive symptoms in three out of four patients within seven days of stimulation. At 26 weeks follow-up, two patients showed more than an 80% decrease in depression rating scales [49] (Table 1).

Taken together, the results of the aforementioned studies of DBS for TRD imply that stimulation at a number of di fferent brain areas can alleviate depressive symptoms, which is in line with the view that MDD is a circuitopathy involving various brain regions and networks mainly within the limbic CSTC mood circuits [12,79]. However, how DBS of those targets improves the depressive symptoms is not completely clear. Moreover, stimulation parameters vary between studies due to a need to adjust and balance therapeutic e ffects to side e ffects.

MDD is a circuitopathy that involves a wide range of brain structures and exhibits diverse clinical manifestations. Therefore, a one-size-fits-all approach to the DBS targeting may not be beneficial in all patients, whereas a patient-centric selection based on individually disrupted neurocircuits could improve therapeutic outcomes. In evaluating the e ffects of DBS, one needs to focus on overall improvement on depression rating scales as well as individual scores and symptom-specific improvements. This will enhance the understanding of the e ffects of DBS and eventually contribute to the development of more personalized treatment approaches. Seemingly, this also applies in other psychiatric disorders such as OCD, where personalized approaches with content-specific DBS targets have already proven to be beneficial [80].

### **4. Towards a More Personalized DBS Treatment Approach for Treatment-Resistant Depression**

Since open-label trials and RCT data on DBS in TRD show inconsistent results, this gives rise to discussion about the chosen study designs, the correct interpretation of results, and the best target(s) for neuromodulation. Depression entails di fferent clinical subtypes and looking at homogenous subgroups of depressed patients may lead to a personalized DBS approach. This would be superior to looking at primary outcomes across all participants. Importantly, a prerequisite to this approach is the ability to determine pathoanatomical substrates of specific subtypes. How to implement such a more personalized approach to DBS treatment for TRD is discussed below.

### *4.1. Clinical and Neurophysiological Subtypes of Depression*

Most response rates in depression treatments to date have been measured with changes in average levels among all patients treated. However, depressive symptomatology varies highly among individuals, making the standardization of positive outcomes challenging. Mood, sleep rhythm, concentration, psychomotor, and cognitive domains can all be disturbed in depression, while treating one selected brain structure within the mood circuit may not have an e ffect on all aforementioned symptoms nor have an e ffect on the main symptomatology of all depressed patients.

Subdividing TRD into di fferent subtypes, involving distinct clinical symptoms as well as distinct patterns of dysfunctional connectivity in limbic and frontal striatal networks, may reveal di fferent subtype-related outcomes for each investigated brain region, and if so, patient selection for a given brain target could enhance treatment e ffectiveness [81]. Analysis of resting-state connectivity biomarkers previously revealed four connectivity-based biotypes of depression characterized by either anxiety, increased anhedonia, psychomotor retardation, and/or increased anergia and fatigue. Moreover, patients could not be di fferentiated into a particular subtype based on clinical features alone and clustering them based on functional connectivity was needed [82]. Therefore, imaging procedures as well as featured symptoms should be taken into account when treating TRD with DBS. It is conceivable that subdividing TRD patients according to connectivity-based biotypes will shed new light on the interpretation of previous DBS study results, and that the integration of functional connectivity in future DBS studies will reveal clinically relevant subgroups that might respond to DBS of a specific target within the mood circuit. Altogether, it can be suggested that better assessment of therapeutic outcomes at symptom level might be accomplished when TRD patients with dominant anergia/fatigue symptoms (biotype 2) are stimulated within the CG 25; and patients characterized by more anxiety (biotype 4) are stimulated within the thalamic region, as suggested by Drysdale and coworkers [82]. Likewise, SCG stimulation could alleviate sleep disturbances and NAcc stimulation could improve anhedonia (Table 2).
