Review and Comparison of Somatic Therapies

Electroconvulsive therapy (ECT) was the first device-based treatment to become available for psychiatric illnesses in the 1940s and consists of the application of an electric stimulus to the surface of the head, with the aim of inducing a seizure. ECT is used in clinical practice and is considered the gold standard for treatment-resistant depression (depression that has not responded to two or more adequate pharmacologic trials). Current guidelines from the American Psychiatric Association (APA) (1) recommend ECT when there is need for a rapid response because of the severity of a psychiatric condition, when there is a history of poor treatment response or a good response to ECT, or if there are intolerable adverse side effects from medications.

According to APA guidelines, ECT should be considered only for patients with major depression with a high degree of symptom severity, for cases in which psychotic symptoms or catatonia are present, or for cases in which there is an urgent need for response, such as patients who are suicidal or refusing food. The biological mechanisms of action of ECT are not currently known and numerous hypotheses have been proposed (2–4).

Before ECT, a general anesthetic is administered to the patient together with a short-acting muscle paralyzing agent in order to prevent injuries. The parameters of the electric stimulus can vary widely (pulse width from 0.25 to 1 msec, frequency from 10 to 140 Hz, duration of the stimulus: 0.5–8 sec), and are titrated according to seizure threshold (the minimum amount of energy that induces a seizure in an individual patient), clinical efficacy, and the side effects for each individual. The sites of application of the stimulus can be bilateral, right unilateral, or bifrontal. The EEG recorded during the seizure usually presents patterned sequences consisting of high-voltage sharp waves and spikes, followed by rhythmic slow waves that in most cases end abruptly. For decades it was common belief that the induction of a generalized seizure provided the necessary and sufficient conditions for ECT’s antidepressant effects (5). However, a randomized, double-masked trial contrasting three dosages of right unilateral ECT (at 1.5, 2.5, or 6 × seizure threshold) and a high-dosage form of bilateral ECT (2.5 × seizure threshold) has demonstrated how the response rate differs significantly between “low-dose” versus “high-dose” ECT, in particular for unilateral therapy, where high-dosage right unilateral (6 × seizure threshold) and bilateral ECT did not differ in any efficacy measure, while the two lower dosages of right unilateral ECT had inferior outcome (6). Moreover, the antidepressant efficacy and cognitive side effects of right unilateral ECT are thought to be dependent on the magnitude of the stimulus dose relative to the seizure threshold for a specific individual, with a dose-response relationship (7). As a result, right unilateral ECT has become standard practice in most ECT centers, as it is associated with lower incidence of cognitive side effects, while bilateral ECT is preferred in those cases where a rapid response is necessary, or in the case of no response to right unilateral ECT. Further developments of ECT technique with the goal of minimizing cognitive adverse effects are the administration of an ultrabrief pulse to induce the seizure (8, 9) and bifrontal application of the stimulus (for a review, see Dunne and McLoughlin [10]). The clinical efficacy of ECT is well-established and substantially exceeds that of any other form of antidepressant treatment, with a 60%–90% rate of response in treatment-resistant depression (11). In a large, multicenter clinical effectiveness study in the community setting that included 398 patients with various diagnoses, the overall response rate was 64% (12).

ECT is available in academic and community settings and an acute ECT treatment series is typically administered two to three times per week for an average of 3 weeks. Common adverse effects of ECT are usually temporary and include arrhythmias, headaches, muscle aches, minor dental and tongue injuries, and nausea (13). Serious medical complications, such as heart attack, stroke, or death, are exceedingly rare even in patients with severe cardiovascular illness. A recent paper estimates the mortality rate associated with ECT to be less than 1 death per 73,440 treatments (14).

The cognitive side effects produced by ECT have been the subject of intense investigation (6, 15). Immediately after the seizure induction and upon emergence from anesthesia, patients experience a usually brief period of disorientation, with impairment in orientation, attention, and memory (16). The technique utilized in delivering ECT has a profound impact on the nature and magnitude of acute cognitive side effects, with bilateral electrode placement, higher electrical dosage relative to seizure threshold, shorter treatment intervals, and higher dosage of anesthetic for induction being independently associated with more severe side effects (17, 18). Moreover, patients vary considerably in their predisposition to develop adverse effects, with patients manifesting global cognitive impairment before treatment and those experiencing prolonged disorientation in the acute postictal period being the most vulnerable to persistent retrograde amnesia for autobiographical information (19). In those patients who experience deficits in attention and concentration due to the underlying condition, ECT usually produces a significant improvement (20).

ECT can also cause some form of amnesia, either anterograde (inability to recall newly learned information, usually short-lived) or retrograde (the forgetting of information learned before treatment). Following termination of ECT, the anterograde amnesia usually resolves, while persistent deficits are usually greater for events that occurred at the time of the ECT series (15). A recent meta-analysis showed that cognitive abnormalities associated with ECT are mainly limited to the first 3 days posttreatment, and after 15 days, processing speed, working memory, anterograde memory, and some aspects of executive function overall were improved beyond baseline levels (21). Overall, ECT is considered safe, and it has a well-established efficacy as an antidepressant treatment in patients with treatment-resistant depression, bipolar depression, and psychotic depression.

Magnetic seizure therapy (MST) is a form of convulsive therapy in which seizures are elicited through a magnetic rather than electrical impulse, and it was developed as a possible alternative to ECT. In fact, in all neurostimulation therapies, because of the high impedance of the skull, an unknown proportion of the electrical stimulus is involved in neuronal depolarization. The stated aim of MST is to avoid inducing significant seizure spread in medial temporal lobe structures, therefore sparing as much as possible the hippocampus and other regions critical for memory. From initial animal data, MST appeared to cause induced current, less robust seizure spread, and less marked anatomical changes in the dentate gyrus than ECT (22, 23).

MST requires general anesthesia and the administration of a muscle relaxant. The tonic-clonic seizures are similar to those seen with conventional ECT, and both treatments produce significant increases in heart rate as well as systolic and diastolic blood pressure, suggesting that the magnitude of the sympathetic discharge due to catecholamine release is similar (24). Early studies suggested that MST is safe and well-tolerated in nonhuman primates (23, 25) and in humans (26, 27). A recent review that included 11 studies suggested that MST has overall few cognitive adverse effects (28).The small sample size of the original studies, however, must be taken into consideration. Regarding MST efficacy, only a small number of clinical trials have been conducted since the first application of MST in 2001 (24). A recent open-label study of 20 patients (29) suggested a more benign side effect profile of MST compared with ECT; another randomized, double-masked study on 20 patients has been concluded but the results are not yet available; and other double-blind studies are currently being conducted in Germany, the United States (comparing MST and ECT), and Canada (source: Clinicaltrials.gov).

Transcranial direct-current stimulation (tDCS) is a noninvasive method of brain stimulation utilized mostly for research purposes and involves the use of direct current applied to the head or scalp. The device used in tDCS is a small current generator capable of delivering a constant electrical current flow, usually of up to 2 mA, that is attached to two electrodes soaked in saline or water and placed inside sponges. One of the electrodes is placed over the region of interest and the other electrode, the reference electrode, is placed in another location in order to complete the circuit. Although parameters of stimulation may vary, the current density, duration, polarity, and location of stimulation have been shown to have important implications in the neuromodulatory outcome of stimulation (30). The mechanism of action is not known, and tDCS is thought to exert a neuromodulatory effect on neuronal membrane potentials, increasing or decreasing cortical excitability and therefore modulating functional connectivity (31, 32).

Overall, tDCS is considered a safe technique with mild and transient adverse effects. However, data on safety and tolerability are largely provided from single-session studies in healthy volunteers. A recent systematic review of tDCS clinical trials identified 209 studies, and the most commonly reported adverse events were itching, tingling, headache, burning sensation, and discomfort (33). There have been initial reports of therapeutic effects of tDCS in numerous conditions, including Parkinson’s disease (34), tinnitus (35), fibromyalgia (36), and poststroke motor deficits (37).

The outcomes reported by these studies were quite variable, probably reflecting the considerable diversity of tDCS methodology used in terms of electrode placement (e.g., various parts of the head, trunk, limbs), electrode size, current amplitude (20–500 mA), and stimulation duration (2 minutes to 8 hours).

Two small double-blind, sham-controlled studies from the same group in Brazil reported positive results in reducing depressive symptoms using left prefrontal anodal tDCS at higher stimulation intensities (1–2 mA) (38, 39). However, two other controlled studies from an Australian group also using left prefrontal anodal tDCS did not detect any difference between groups and sham, and reported one participant with bipolar disorder becoming hypomanic after active tDCS (40, 41). A recent meta-analysis including six randomized controlled trials, with a cumulative sample of 96 patients treated with active and 80 with sham tDCS, reported active tDCS to be more effective than sham for the reduction of depression severity (42), although the studies were heterogeneous. Finally, a clinical trial comparing tDCS, sertraline, and placebo for the treatment of depression is currently underway in Brazil, as well as other trials in patients with fibromyalgia, pain, and stroke (source: Clinicaltrials.gov).

Cranial electrical stimulation or CES is produced by the application of low-level, pulsed electrical currents (usually a few milliamperes) to the head. Its primary indications are anxiety, depression, and insomnia; it has also been utilized in drug addiction and pain (43). CES devices deliver a current between 1-15 mA at a frequency variable between 0.5 and 100 Hz. The electrodes can be placed bitemporally, forehead to posterior, or through electrodes clipped to the earlobes. A generic clinical protocol of application is 20–45 minutes daily or twice a day for 1 to 3 weeks. The patient can use the device at home, although the manufacturers recommend supervision by a physician at least every 3 months.

There are numerous CES devices commercially available on the market that are FDA-approved and classified as “Class III Devices,” which can be purchased by patients “upon written authorization from a licensed healthcare practitioner” as indicated on the website of all manufacturers. The FDA funded a large safety study of CES, conducted by the National Research Council, and the results indicated that the device poses no significant health risk; rare side effects include headache, dizziness, mild tingling sensation at the electrode sites, and skin irritation (44). The mechanism of action remains unclear, although it is hypothesized that direct current modulates spontaneous neuronal activity with effects that are transmitted throughout the brain via specific pathways (45). A recent study testing 0.5- and 100-Hz stimulation on default mode network via fMRI suggested that CES causes cortical brain deactivation, with a similar pattern for high- and low-frequency stimulation, and alters connectivity in the default mode network (46). A recent Cochrane analysis of CES efficacy in patients with chronic pain (47) reported insufficient evidence to draw firm conclusions, given that no double-blind trials have demonstrated a clear benefit of active CES over sham stimulation.

The efficacy of CES for depression remains uncertain. While a substantial body of published research does exist, the research design and quality of the studies vary widely. A meta-analysis published in 1995 (48) identified 18 studies of CES versus sham treatment, and showed CES to be significantly more effective than sham treatment; however, most studies failed to report sufficient data and in all but two trials, the therapists were not blinded to treatment arm. A recent pilot study investigated the efficacy of CES in 12 patients with generalized anxiety disorder and reported positive results (49).

In summary, CES appears to be a safe device-based treatment that may help relieve a number of symptoms, particularly in nonclinical populations, but it has not been studied in double-blind sham-controlled studies to determine its efficacy in different psychiatric conditions. Two randomized trials are currently underway, one in patients with major depression and one in patients with anxiety disorders (source: Clinicaltrials.gov).

Transcutaneous vagus nerve stimulation (tVNS) stimulates superficial branches of the vagus nerve without surgical procedures by delivering electrical impulses to the surface of the left ear stimulating the superficial branches of the vagus nerve. One type of device is commercially available on the European market upon prescription of a provider, with the indication for pain relief therapy and treatment-resistant epilepsy (50). Currently this device is not available in the United States. Potential mechanisms of action for tVNS are not currently known, and very few studies have been conducted to investigate the effects of the stimulation on different brain structures (51). It has been reported that tVNS reduced symptoms of pain sensation in healthy volunteers (52), but no controlled studies have been published. According to the manufacturer, the recommended daily stimulation dose is 4 hours in several sessions, and potential side effects reported in a retrospective pilot study include itching, discomfort in the area of the outer ear, and local pain at the stimulation side (53). Three randomized sham-controlled trials have been completed in Germany in patients with epilepsy, schizophrenia, and pain, but the results have not yet been published and one trial for the treatment of tinnitus is currently underway (source: Clinicaltrials.gov).

Transcranial magnetic stimulation (TMS) was introduced as a neurophysiological technique in 1985 with the aim of delivering noninvasive stimulation of the cerebral cortex (54). In TMS devices, a brief electric current passes through a magnetic coil positioned above the surface of the cranium. The electric current induces a transient, high-intensity magnetic pulse that penetrates through the scalp, skull, and meninges to the underlying cortex. This pulse, in turn, generates an electric field within the targeted cortical regions that can induce depolarization of superficial cortical neurons, penetrating at a depth of approximately 2–3 cm below the surface (55). Variations in stimulation parameters include orientation of the magnetic field, single or repeated stimulation, frequency of stimulation, number of pulses, and intensity and site of stimulation. The minimum amount of energy needed to produce contraction of the thumb (abductor pollicis brevis) is called the motor threshold and is the base for the choice of the amount of energy delivered per pulse for each individual (between 80% and 120% of motor threshold).

TMS’ mechanism of action is unknown and is thought to be mediated by gabaergic and glutamatergic systems. Low-frequency (<1 Hz) repeated TMS (rTMS) is thought to inhibit cortical firing in certain regions by inducing decreased regional cerebral blood flow in the stimulated area (56).

Alternatively, high-frequency (>1 Hz) rTMS is thought to be excitatory by inducing increased regional cerebral blood flow in the stimulated area (57). There is also evidence that the application of TMS reaches deeper structures in the brain as shown by a magnetic resonance spectroscopy study reporting increased glutamate levels in the left anterior cingulate cortex after stimulation of the left dorsolateral prefrontal cortex (58).

The initial clinical application of rTMS to patients with depression was driven by functional imaging data demonstrating that patients with depression have reduced activity in the left dorsolateral prefrontal cortex (59, 60) and the underlying hypothesis was that repeated stimulation of this hypofunctional area with TMS could lead to clinical improvement in symptoms of depression. More recently, however, investigators have focused on the idea of an imbalance in the activity of the frontal lobes, with hypofunction in the left frontal lobe and excessive inhibitory activity in the right frontal lobe. This led to the proposal of an alternative pattern of treatment with low-frequency rTMS of the right dorsolateral prefrontal cortex (56).

TMS devices are available to clinicians and researchers. TMS is a noninvasive technique that does not require anesthesia and can be safely administered as an outpatient procedure. A typical TMS session lasts between 30 and 60 minutes; the sessions are five times a week, usually for a period of 3 to 4 weeks for a total of 20 to 30 sessions.

TMS is usually well tolerated, with no cognitive adverse effects and exceedingly rare medical complications, while the most common adverse effects are headaches and facial pain (61). Cautionary use is indicated in patients with history of seizures, with implanted VNS leads, or any implanted medical device located less than 30 cm from the rTMS magnetic coil. The presence of magnetic-sensitive metal implanted in the head such as cochlear implants, aneurysm clips, coils, or stents are considered absolute contraindications for TMS. Because of its availability and safety, TMS use has been investigated in pilot studies for the treatment of diverse neuropsychiatric disorders, including anxiety (62, 63), psychosis (64), and addiction (65) with mixed results.

There is now evidence to support the efficacy of TMS for MDD. Early randomized, double-blinded, sham-controlled, parallel trials of TMS did not show clear mood benefits; however, these trials were limited by small sample sizes and were likely underpowered for signal detection. In a comprehensive Cochrane review of TMS for depression, published in 2002 (66), the authors concluded that high-frequency left prefrontal rTMS and low-frequency right prefrontal rTMS were statistically superior to a sham comparison. The most recent meta-analysis (67), which included 34 studies comparing rTMS to sham treatment, demonstrated a statistically and clinically significant benefit for rTMS, with an effect size of 0.55. The largest single study of TMS, sponsored by a TMS equipment manufacturer, provided the basis for FDA approval and also found an effect size for TMS of 0.55. In this double-blind, multisite study, 301 medication-free patients with MDD, who had not responded to prior treatments, were randomly assigned to either active (N=155) or sham TMS (N=146) conditions (68). TMS separated from sham by weeks 4 and 6: response rates were 18% versus 11% and 24% versus 12%, respectively; and remission rates were approximately twofold higher at week 6, with active TMS versus sham. In 2009, a large multicenter study was designed by NIMH (69) to address key methodological limitations (e.g., adequacy of masking, validity of sham treatment, training of raters, and reliability of outcome evaluation) and to maximize the likelihood of robust antidepressant effects by utilizing high-intensity stimulation (120% motor threshold), a high number of pulses (3000 stimuli per session), and magnetic resonance imaging (MRI) adjustment for proper scalp placement. The results of the primary efficacy analysis revealed a difference in the proportion of remitters receiving rTMS (14.1%) versus sham (5.1%). The major point of controversy in the field resides in the clinical meaning of the observed improvement and the total number of patients who were in remission in those clinical trials. In 2011, an Agency for Healthcare Research and Quality (AHRQ) report pooled the results from 14 trials for a total of 497 patients, and found that rTMS was beneficial for severity of depressive symptoms, response rate, and remission rate, with strong evidence for severity of depressive symptoms and response rate, and moderate strength evidence for remission rate (70). Other more recent naturalistic and open-label studies in patients with TRD have been published (71, 72). In summary, daily TMS stimulation of the left dorsolateral prefrontal cortex for several weeks can have an antidepressant effect; it is safe and well tolerated. The majority of data indicate that TMS is an effective treatment for those unable to benefit from initial antidepressant medication (71).

Vagus nerve stimulation (VNS) was originally developed for the treatment of epilepsy. In 1997, VNS was approved by the FDA as “an adjunctive therapy for reducing the frequency of seizures in adults and adolescents who were refractory to antiepileptic medications.” Early anecdotal clinical observations of improvement of mood symptoms in epilepsy patients that appeared to be independent from improvement in seizure control (73) suggested that VNS could also have an effect on depressive symptoms.

Although the vagus nerve can be stimulated using a noninvasive transcutaneous method (see tVNS above), VNS involves surgical implantation of a pacemaker-like pulse generator device. Because approximately 80% of vagus nerve fibers are afferent, most of the electrical pulses applied to the nerve are propagated from the point of attachment toward the brain. The VNS device currently available on the market is manufactured by Cyberonics and consists of a lithium battery and a lead wire system with electrodes. The generator is implanted in the left chest wall and connected to a lead with the left vagus nerve by a neurosurgeon, either under local or general anesthesia. Following the surgery, the device is activated telemetrically by a wand connected to a hand-held computer. The device delivers a cyclical train of pulses of a few milliamperes (0.25 to 3.5 mA), with a pulse width between 130 and 1000 microseconds and a frequency between 1 to 50 Hz. The settings are adjusted to optimize efficacy and tolerability for each patient.

The safety of the VNS device is well established from its use in the treatment of epilepsy. In total, more than 60,000 patients worldwide have had a VNS device implanted since its approval in 1997 (source: http://www.cyberonics.com/). Possible adverse events associated with VNS surgery include wound infections, pain at the surgical site, left vocal cord paresis (rare), and arrhythmias during the initial testing of the device. The most common side effects associated with stimulation are hoarseness, dyspnea, cough, and reversible bradyarrhythmias; these are thought to be dose dependent and correlate with stimulation intensity. VNS has also been shown to worsen preexisting obstructive sleep apnea/hypopnea (74). The rate of stimulation-induced switch to mania or hypomania in the VNS trials was overall low (75, 76), and those symptoms generally subside with adjustment of stimulation parameters.

The mechanism of action of VNS is currently not well understood. Stimulation of the vagus nerve affects the function of multiple brain regions known to regulate mood, appetite, sleep, energy, reward, and motivation. The afferent fibers traveling in the vagus terminate largely in the nucleus tractus solitarius in the medulla. The nucleus tractus solitarius, in turn, innervates the noradrenergic nucleus locus coeruleus, which projects to the orbitofrontal cortex and the insula. Human studies using fMRI and PET techniques show that VNS induces neuronal activity changes within amygdala, hippocampus, and thalamus (for a review see Nemeroff et al. [77]). A recent fMRI study has also linked the effects of VNS therapy to deactivation of the ventromedial prefrontal cortex and activation of right insula in patients with depression, a mechanism similar to antidepressant drugs (78). Positron emission tomography with oxygen-15 labeled water identified changes in regional cerebral blood flow in response to acute VNS stimulations in 13 subjects with treatment-resistant depression—decreases in the left and right lateral orbitofrontal cortex and left inferior temporal lobe, and significant increases in the right dorsal anterior cingulate, left posterior limb of the internal capsule/medial putamen, the right superior temporal gyrus, and the left cerebellar body (79).

The primary indication for VNS in psychiatry has been as an add-on therapy for patients with treatment-resistant depression. In the first open-label trial of VNS (80), 30 patients with a diagnosis of major depressive episode (MDD or bipolar depression) of at least a 2-year duration (i.e., chronic) who had failed at least two adequate antidepressant trials (average number of failed trials was 4.8 [SD=2.7]) received VNS augmentation over 10 weeks. The results at endpoint were promising, with a response rate of 40% and a remission rate of 17% over 10 weeks (80). This led to a second pilot study during which 30 new patients were added to the initial cohort and sequenced through an identical trial design to generate a total sample size of 60 (81). Although response and remission rates for the total sample were lower (30% and 15%, respectively), the difference in efficacy was thought to be moderated by higher levels of treatment resistance (particularly nonresponse to ECT) in the second sample (82). An open-label, European study (N=74) reported similar results, with a response rate of 37% and remission rate of 17% after 3 months (83). Perhaps the most compelling finding in the long-term follow up of those studies was the apparent growing benefit over time in response and remission rates: for the U.S. cohort, a response rate of 44% and a remission rate of 27% were seen at 1 year, and rates of 42% and 22%, respectively, were seen at 2 years (82); in the European cohort, a response rate of 53% and remission rate of 33% were seen after 1 year, and rates of 53% and 39%, respectively, were seen at 2 years (84). Other naturalistic follow-up studies and open-label case series have been reviewed in detail in a recent paper (85).

The largest randomized, sham-controlled, multicenter study of adjunctive VNS enrolled 235 patients and compared active VNS to a control group (86). The control group had the surgical procedure to implant the VNS device, but they did not have the device turned on. More than 40 percent of the sample had four or more prior antidepressant treatment failures and 50% of patients received ECT lifetime, indicating a high degree of treatment resistance. In a modified intent-to-treat analysis that excluded those noncompliant with the medication protocol, the results did not demonstrate a statistically significant difference between the two groups for the primary outcome (24-item Hamilton depression scale score). No differences were found in the percentage change in depressive severity (−16.3% for VNS versus −15.3% for control, p=0.639) or the response rates (15.2% versus 10.0%, p=0.25) (86). However, follow-up observations of this cohort replicated earlier studies, showing increasing treatment benefit over time, with an overall response rate of 33% after 2 years (87). There are also additional data suggesting that VNS is associated with decreased suicidality (ideation and attempts), rates of hospitalization, medical costs, and mortality at 2 year follow-up (88).

In summary, VNS is FDA-approved for patients with chronic or recurrent depression, either unipolar or bipolar, with a history of failing to respond to at least four antidepressant trials. VNS is usually considered as an adjunct to pharmacologic treatment and it can safely be combined with ECT in case of an acute relapse. VNS cannot be considered an acute treatment for treatment-resistant depression because achieving benefits may take up to 6–12 months. Since VNS is FDA-approved for treatment-resistant depression in the absence of Class I evidence of efficacy, insurance companies have resisted reimbursement for the implant and, so far, this lack of reimbursement has limited access to this device in the United States.

Deep brain stimulation is a reversible neurosurgical procedure consisting of implanting electrodes at specific anatomical locations, and through those electrodes delivering an electrical impulse of variable intensity and frequency. The mechanism of action is unknown, and DBS is thought to alter complex firing patterns of the neurons in the region and thus modify activity in the neuronal circuits (89).

This technology was initially developed for refractory neurologic disorders such as tremor, Parkinson’s disease, and dystonia (90).

In patients implanted with DBS for movement disorders, it has also been observed that both acute and chronic stimulation of different targets could induce mood changes, including hypomania, dysphoria, and anhedonia; these observations led to the development of clinical trials to test the efficacy of DBS in refractory mood disorders. A number of research groups are currently investigating different sites for implantation of electrodes (for a complete review see Hauptman et al. [91]) including the subcingulate –Brodmann’s area 25 (SCG 25), the ventral anterior internal capsule/ventral striatum (VC/VS), the nucleus accumbens, and the inferior thalamic peduncle as stimulation targets for refractory depression.

The implantation of DBS electrodes in patients is a complex neurosurgical procedure. Under stereotactic guidance, two electrodes are placed relative to a set of anatomical landmarks in deep structures of the brain. During the whole procedure the patient remains awake. Two programmable neurostimulators are implanted in the chest area under each clavicle and are connected to the corresponding electrode by extension wires tunneled subcutaneously under general anesthesia. After several weeks, systematic outpatient adjustment of stimulation parameters is performed and the parameters can be adjusted by varying number of active contact(s) on each electrode, pulse amplitude and duration, and stimulation frequency. Frequent follow-ups are necessary, especially during the first 6—12 months after implantation, to enable optimization of stimulation parameters, monitoring of the patient, and coordination of other pharmacological and behavioral therapies.

The rates of surgical complications are quite variable in the DBS literature (mostly derived from large clinical trials in DBS for movement disorders) and include intracranial hemorrhage, stroke, infection, and lead fracture. Overall, infection of the hardware (leads, connectors, or batteries) is the most commonly reported serious surgical complication, and it may require explant of the device (92).

Neuropsychiatric side effects of DBS include manic or hypomanic symptoms, anxiety, restlessness, worsening depression, apathy, and impulsivity (93); however, these symptoms are thought to be transient and to respond to modification in parameters of stimulation. DBS in treatment-resistant patients requires a dedicated multidisciplinary team of neurosurgeons, psychiatrists, neuropsychologists, and support staff. Replacements of the batteries add to the total burden for the patient, being necessary on average every 12–24 months

Regarding efficacy, DBS was approved by the FDA in 2009 for treatment of otherwise intractable obsessive-compulsive disorder under the FDA’s Humanitarian Device Exemption program. Four pilot studies have been conducted, two at the VC/VS target (94, 95), one at the nucleus accumbens target (96), and one at the subthalamic nucleus. One multicenter NIH-sponsored randomized, sham-controlled study with the VC/VS target is being conducted in the United States and one at different targets in France (source: Clinicaltrials.gov)

In patients with treatment-resistant depression, open-label studies have been conducted for three targets, SCG25, VC/VS, and the nucleus accumbens. The largest cohort of patients was of 20 patients with DBS at the SCG25 followed for 3 to 6 years after the implant (97), and those data support the long-term efficacy and safety of DBS.

There appeared to be no significant loss of effect requiring dose adjustments over time, a phenomenon that parallels the relative stability of DBS stimulation parameters in patients with other neuropsychiatric disorders.

A more recent open-label trial of DBS implanted at SCG25, conducted independently by a Spanish group, also supports the potential efficacy of DBS at this target (98) for patients with severe treatment-resistant depression.

There has been one published open-label trial of DBS targeting the VC/VS that was conducted on a sample of 15 patients (99) and one open-label trial at the nucleus accumbens (100) including 10 patients. Finally there is one case report of DBS implanted at the inferior thalamic peduncle site (101).

The data on efficacy in treatment-resistant depression are currently limited to a series of open-label studies involving the most refractory patients with a diagnosis of major depression. Additional double-blind, randomized, sham-controlled trials will be necessary to establish the efficacy of DBS. Of note, DBS is not a treatment indicated for an acute depressive decompensation, since the benefits may take weeks to months to manifest.

At present, DBS for treatment-resistant depression continues to be an experimental field, with active investigation into optimal neuroanatomical locations for electrode placement and parameters for stimulation.