Deep brain stimulation (DBS) is a neurosurgical technique involving stereotactic implantation of electrodes in specific nodes of brain circuits and their connection to a subcutaneous pulse generator. It is an established therapy for some neurological disorders, but in treatment resistant major depression it remains an investigational strategy: results from open label cohorts have often been favorable, whereas randomized studies with a sham phase have produced less consistent results.
DBS should therefore not be presented as a “more powerful” version of rTMS or tDCS. It is an invasive procedure, with surgical and hardware risks, intended for research and highly specialized centers treating patients with exceptionally refractory depression. Unlike electroconvulsive therapy, it does not induce generalized seizures, and stimulation is chronic and programmable.
Scientific interest stems from the idea that treatment resistant depression does not depend on a single area but on dysfunctional networks. The main targets studied include the subcallosal cingulate, ventral capsule and ventral striatum, nucleus accumbens, medial forebrain bundle and other limbic frontal nodes. More recent research focuses less on isolated anatomical coordinates and more on the circuits and white matter tracts actually engaged by stimulation.
The subcallosal cingulate, often referred to as area 25, is connected with the prefrontal cortex, limbic structures, hypothalamus and brainstem. Early studies showed that stimulation of this region could be associated with sustained improvements in some patients with treatment resistant depression. However, response depends critically on contact location and the fibers traversed by the electric field; a difference of only a few millimeters can mean engagement of different circuits.
The ventral capsule and ventral striatum have been studied for their role in motivational and reward circuits and frontostriatal control. The nucleus accumbens and medial forebrain bundle more directly target circuits related to anhedonia and motivation. None of these targets has demonstrated definitive, replicated superiority sufficient to establish a universal standard.
Modern models interpret DBS as modulation of distributed networks: high frequency stimulation does not simply “switch off” a nucleus, but can alter impulse propagation along axons, network oscillations and communication between regions. The effect may emerge gradually over weeks or months and is not explained by a single neurochemical response.
DBS studies in depression generally include patients with extreme treatment resistance, long lasting episodes, severe functional impairment and failure of multiple pharmacological, psychotherapeutic and somatic strategies, often including ECT. The exact definition of resistance varies across protocols, contributing to heterogeneity of results.
Before implantation is considered, the diagnosis must be reviewed in depth. Bipolar disorder, psychotic disorders, substance use, personality disorders, neurological diseases, sleep disorders and medical conditions can alter both the diagnosis of treatment resistant depression and surgical risk. The adequacy, duration and adherence of previous treatments must also be documented because pseudo resistance does not justify an experimental neurosurgical procedure.
Selection includes multidisciplinary psychiatric, neurosurgical, neuropsychological and anesthesiological assessment. The ability to understand uncertainty of benefit, willingness to undergo very long follow up, social support and capacity to manage programming and device revisions are essential. Informed consent must make clear that DBS for depression is not a standard treatment and may provide no benefit despite technically correct implantation.
Implantation uses structural imaging and stereotactic planning to define the target and trajectory. Electrodes are placed bilaterally and connected through subcutaneous extensions to a pulse generator, often placed in the chest. Surgical techniques and the type of anesthesia vary by center, target and protocol.
After implantation, a programming phase begins in which active contacts, amplitude, frequency and pulse width are adjusted. The aim is to achieve effective circuit engagement while limiting adverse events. In depression this phase can be lengthy because the therapeutic effect is not necessarily immediate and overly frequent changes can make it difficult to attribute the clinical course to a specific setting.
Pulse generators require checks, possible recharging or battery replacement, and surveillance of electrode and extension integrity. MRI examinations and electromedical procedures must follow the specific conditions of the implanted system. Effective DBS is therefore a chronic treatment pathway, not a single surgical procedure.
Planning uses structural neuroimaging and stereotactic coordinates and, in more advanced programs, connectivity information to try to engage white matter tracts associated with response. After implantation, clinical work continues with wound checks, impedance measurements, programming of contacts, amplitude, frequency and pulse width, and systematic observation of mood and behavior. Response may require weeks or months and often requires multiple adjustments. Rechargeable or nonrechargeable batteries, extensions and the generator also require maintenance and possible revisions. For this reason, psychiatric DBS is a prolonged multidisciplinary pathway, not a single neurosurgical procedure.
Uncontrolled cohorts have reported substantial rates of response and remission sustained over time across several targets. However, in a fluctuating disorder and in patients receiving intensive follow up, open label studies are vulnerable to regression to the mean, expectations, concomitant treatment changes and participant selection. This is why sham controlled studies are crucial.
The large multisite study of subcallosal cingulate stimulation published in 2017 did not demonstrate a significant advantage of active stimulation during the planned controlled phase, and similar difficulties emerged in studies of the ventral capsule and ventral striatum. In 2026, a new randomized study of the subgenual cingulate showed overall improvement but no significant difference between conditions during the controlled phase, whereas some patients continued to improve during open follow up. These findings are compatible with the possibility that some patients respond, but they do not yet demonstrate a predictable and standardizable effect.
Meta analyses combine small studies, different targets and heterogeneous designs. Signals of efficacy are stronger in open phases than in sham comparisons. This gap is the main reason DBS for depression must be described as experimental despite circuit plausibility and durable results observed in subgroups.
Risks include intracranial hemorrhage, infection, seizures, anesthetic complications and, rarely, neurological deficits. Long term complications may include delayed infections, lead migration or breakage, skin erosion, malfunctions and the need for surgical revisions. These events are particularly important because treatment is elective and benefit is not guaranteed.
Stimulation can induce psychiatric effects such as anxiety, irritability, insomnia, activation, hypomania or behavioral changes, depending on the target and parameters. In patients with highly treatment resistant depression, baseline suicide risk is high and must be monitored independently of the procedure; implantation does not eliminate the need for a safety plan and comprehensive psychiatric treatment.
Reviews of cognitive function do not show a consistent pattern of global deterioration, but samples are small and selected. Serial neuropsychological assessment remains necessary both to detect possible effects and to distinguish changes related to depression from those associated with surgery or stimulation.
Beyond perioperative risks, hardware related problems include infection, erosion, pain at the generator site, component fracture or migration, and the need for replacement. Stimulation can produce changes in mood or activation that require parameter adjustment; reports of hypomania, anxiety, worsening depression or suicidality require specialist monitoring without automatically attributing every event to stimulation. Available reviews do not show consistent global cognitive deterioration, but samples are small and selected. Safety must therefore be interpreted by considering surgical risk, stimulation effects, disease progression and prolonged exposure to the device.
The main challenge is to identify who will respond and where to stimulate. Connectomic targeting uses tractography and connectivity to select contacts that engage specific frontolimbic tracts; closed loop approaches seek neurophysiological signals that enable adaptive stimulation. These methods aim to reduce the heterogeneity that limited early trials based on relatively uniform anatomical coordinates.
Another priority is to separate the effects of surgery, expectations and intensive monitoring from the specific effect of stimulation. Trials with sufficiently long controlled phases, shared selection criteria and functional outcomes in addition to symptom scales are essential before routine clinical use.
Today, DBS therefore has value primarily for scientific and translational research. In patients with treatment resistant depression that can be treated with ECT, rTMS, VNS, ketamine or esketamine, or pharmacological strategies, these options should be considered according to their level of evidence before an experimental intracranial procedure.
Recent research seeks to move beyond the model of a single anatomical target for everyone. Connectomic approaches attempt to identify the tracts actually involved in the depressive circuit, while chronic neural recordings and physiological biomarkers may in the future enable more individualized programming or adaptive systems. These strategies are promising but have not yet produced a validated clinical standard. Before routine use, larger controlled trials, uniform definitions of treatment resistance, credible sham procedures and long term data on efficacy, safety and cost are needed. At present, DBS for depression should therefore remain restricted to highly specialized research protocols and exceptionally rigorous patient selection.
Informational notice: the information contained on this page is provided solely for informational and educational purposes and does not replace the advice, diagnosis or treatment provided by a physician. If needed, always consult a qualified healthcare professional.
Artificial intelligence transparency: this page was created with the support of artificial intelligence tools, used to assist in the production and processing of its content.