Neuromodulation Surgery for Psychiatric Disorders Treatment & Management

Updated: Oct 05, 2016
  • Author: Jonathan P Miller, MD; Chief Editor: Brian H Kopell, MD  more...
  • Print

Medical Therapy

Any patient being considered for neuromodulation surgery will be refractory to medical therapy or have side effects such that they are unable to tolerate medical therapy. Typically, however, whatever medical therapy the patient is on at the time of evaluation for surgery will be continued until neurostimulation parameters can be optimized, at which point medications can start to be tapered as indicated.


Surgical Therapy

Deep brain stimulation

Deep brain stimulation (DBS) implantation for refractory obsessive-compulsive disorder (OCD) or major depressive disorder (MDD) is nearly identical to that for movement disorders. [27] The procedure is typically staged by first implanting the leads (see the first image below) and then implanting the neurostimulator or pulse generator (see the second image below).

Examples of deep brain stimulation (DBS) leads. Examples of deep brain stimulation (DBS) leads.
Example of implantable neurostimulator for deep br Example of implantable neurostimulator for deep brain stimulation (DBS) leads (implantable pulse generator, IPG).

The lead implantation process starts with the acquisition of a high-resolution MRI, on which the surgeon identifies the target region or structure and plans a trajectory that avoids blood vessels and minimizes risk to eloquent areas. To precisely attain the planned trajectory, the surgeon will use a stereotactic frame or similar computer-assisted targeting system.

Typically, the patient will be sedated while incision(s) are made and bilateral holes are drilled in the skull. The patient will then be awakened to allow monitoring of changes in patient behavior in response to intraoperative stimulation. Adverse behavioral changes may prompt the operative team to adjust the position of the electrode. Having the patient awake also facilitates microelectrode recording, a process used at many centers whereby microelectrodes are passed through the brain along the planned trajectory and used to record the electrical activity of the areas that they traverse. This process allows for a physiologic correlate to the anatomy seen on MRI; in the context of psychiatric disease, however, the role of microelectrode recording is much less established than for movement disorders.

Once the optimal position of the electrode is determined, the lead is locked into place with the use of commercially available burr-hole mounting systems. The distal aspect of the lead is often tunneled to the region just posterior to the pinna in preparation for the second phase.

The next phase involves the placement of the implantable pulse generator (IPG), which can be performed under general anesthesia. Typically, the IPG is implanted in the upper chest/subclavicular region while an extension lead is tunneled from the chest to the posterior auricular region. There, the distal aspect of the implanted DBS lead is connected to the IPG via the tunneled extension lead. If bilateral DBS leads are implanted, this phase is repeated for the second side.

Currently, programming DBS for refractory OCD or MDD has no standard approach. In general, high-frequency (>100 Hz) stimulation has been explored at various pulse widths (duration of each electrical pulse in milliseconds) and current settings. Although acute effects have been reported, settings are made and behavioral effects are observed over a period of weeks before another adjustment is made. DBS programming for refractory psychiatric disease is still in the early stages of investigation.

Vagus nerve stimulation

The implantation of the vagus nerve stimulation (VNS) system is done in one stage under general anesthesia. The surgical target is the left vagus nerve, and the procedure is identical to the one used for refractory epilepsy. The patient is placed in a supine position on the operating room (OR) table with the head slightly turned to the right side. Either a longitudinal incision based along the anterior border of the sternocleidomastoid muscle or a transverse incision based on the level of the laryngeal prominence can be used. A standard anterior, trans-cervical approach is performed to expose the left vagus nerve in the carotid sheath.

Once exposed, a 2-contact, 1-reference electrode is placed around the main trunk of the vagus nerve. The distal aspect of this lead is tunneled down to the subclavicular region while the lead itself is anchored to the deep and superficial cervical fascia to prevent dislodgment with neck movement. In the subclavicular region, the lead is attached to a neurostimulator unit in the same fashion as DBS.

Two weeks after implantation of the VNS IPG, the implant is programmed to be active. The IPG works automatically to deliver the programmed parameters. The most common settings used to commence therapy are an output current of 0.25 mA (1.0 mA median value at 12 mo in a pivotal study completed by Rush in 2005 [28] ), a pulse width of 250-500 microseconds (500 microseconds median value at 12 mo in a pivotal study completed by Rush in 2005 [28] ), a signal frequency of 20-30 Hz (20 Hz median value at 12 mo in a pivotal study completed by Rush in 2005 [28] ), and a duty cycle of 10% that activates for 30 seconds every 5 minutes.

The output current is then increased over the following visits by 0.25 mA increments up to a tolerable level or, in this author’s experience, the derived target level of 1.75 mA. Very little data is available to guide individualized target doses. Subsequent visits include interrogating the IPG and running diagnostic tests and adjusting the parameters as needed. By adjusting output currents by 0.25 mA at a time, once or twice during each early office visit, treating physicians can identify the settings that produce the desired balance of benefits to side effects. Once patients respond, increasing the output current any further is unnecessary. Lower output currents extend the battery life.



Deep brain stimulation

The complications of deep brain stimulation (DBS) for refractory psychiatric disease can be broadly classified into 2 categories: general complications of the surgical implantation and side effects of stimulation.

  • General: The complications of DBS implantation for refractory psychiatric disease are likely to be the same in character and frequency as the complications of DBS implantation for movement disorders, due to their similarities. These include intracranial hemorrhage, infection, and hardware complications (including lead breakage and erosions of hardware through the skin). [29]
  • Side effects: DBS for these conditions is still in the early phases of investigation. Side effects from the small numbered case series include the following: fear/panic response, [30] hypomania, [19] and disinhibition and subjective memory disturbance. [31] Note that these effects are immediately reversible with cessation of stimulation or adjustment of stimulation settings. There have been no reported decrements in scores of neuropsychological tests following DBS implantation; indeed, improvements in scores have been reported. [19]

Vagus nerve stimulation

See the list below:

  • General: Possible surgical complications include a low risk of infection at the incision site and possible left vocal cord paresis, which is often transient. Asystole in the operating room has been rarely reported.
  • Side effects
    • Common side effects
      • Side effects related to the use of VNS are only experienced during stimulation.
      • The most common adverse events noted are hoarseness, dyspnea, and cough, which are often related to the intensity of the output current.
      • Hoarseness is the most common adverse event and is generally mild in severity.
      • The adverse events tend to ameliorate with time, and only the hoarseness tends to persist.
      • VNS does not result in sexual dysfunction, dry mouth, urinary retention, weight gain, or other common side effects of psychotropics. [17]
    • Uncommon side effects
      • Hypomanic symptoms may occur, as is true with any antidepressant treatment. In one pilot study, 3 patients (1.2%) developed mild hypomanic symptoms. [32] Two of the three patients had a history of bipolar disorder, and symptoms resolved without intervention.
      • In the pivotal trial, 3 patients developed mania. [28] Two cases were mild and developed within 3 months of patients having starting VNS, and they subsided spontaneously within 1-2 weeks. The third patient developed a manic episode that lasted for 2 months and required hospitalization and cessation of VNS. The VNS was ultimately safely restarted.
      • In the trials, no evidence has emerged that suggests a potential for VNS to worsen depression or induce suicidality.
      • VNS does not appear to have any negative cognitive effects and may improve cognition by improving depression.
      • Sackheim et al looked at 27 patients with depression as they were implanted with VNS and gave them a neurocognitive battery at baseline and at 12 weeks. [33] They found no decrements in cognitive functioning, and 40% had improved depression.

Outcome and Prognosis

Deep brain stimulation

Multiple targets for deep brain stimulation in psychiatric disease have been proposed, some of them stemming from findings in stereotactic lesioning studies while others stem from models of psychiatric circuitry derived from anatomic and functional imaging studies. One target, the subthalamic nucleus, has been used extensively in the treatment of movement disorders and was discovered by serendipity to have possible effects in OCD. Outcomes data are discussed below by target.

Anterior internal capsule

In 1999, a joint investigative group from Belgium and Sweden released an initial communication in The Lancet. [5] In this report, the group implanted a model 3887 Pisces Quad Compact electrode (Medtronic, Inc), generally used for spinal cord stimulation, into the internal capsules of 4 patients with obsessive-compulsive disorder (OCD). In June 2003, this same group published the long-term follow-up of their initial data presented in 1999. [34] The data were presented from 6 patients who were implanted with bilateral DBS electrodes in the internal capsule with follow-up that ranged from 3-31 months. Patient selection criteria were narrowly and specifically defined using Yale-Brown obsessive compulsive scale (Y-BOCS) and global assessment of functioning (GAF) scores. Under double-blinded conditions, the 6 patients were assessed in a crossover design that used Y-BOCS, clinical global severity (CGS) scale, clinical global improvement (CGI) scale, and the Beck depression inventory (BDI).Onepatientdid not receive any benefit. Another patient required such high voltages for stimulation that the batteries were replaced every 5 months with only limited beneficial effects. The electrodes were electively removed. Both this patient and the first patient underwent standard RF capsulotomy. With chronic stimulation, 3 of the 6 patients were considered responders with improvements on Y-BOCS of at least 35%. Another patient who received chronic DBS had some improvement (less than 35%) and was considered a nonresponder. In the "off" state, the 3 responding patients had worsening mood and OCD symptoms that returned to baseline.

Another published report of anterior internal capsule (AIC) DBS is from a group at the Loyola University Medical Center. In 2003, they published a case report of DBS on the anterior internal capsule for OCD. [35] Although full details of diagnosis were not given in this report, the patient, a 35-year-old female, had severe illness and functional impairment, reflected by a Y-BOCS score of 34 and a GAF score of 40. Bilateral DBS electrodes were placed in a target considerably more lateral and anterior to the Belgium/Sweden group. Stimulation parameters were kept constant: 2 V, 210 ms pulse width (PW), 100 Hz. Unipolar stimulation was used, but no mention was made of which contact served as the cathode. At 10 months of follow-up, the patient returned to work with all compulsions "abated." At 3 months, her Y-BOCS score fell to 7. No adverse effects were reported.

In 2005, Abelson et al performed capsular DBS in 4 OCD patients, using a more anterior target than that of Nuttin et al and a different stimulating electrode. [36] Three of the 4 patients had significant benefit (>35% reduction in Y-BOCS scores). Disturbingly, one of the responders committed suicide despite a significant change in OCD symptoms.

Most recently, Greenberg and colleagues reported on a series of OCD patients who met stringent criteria for severity and treatment resistance and underwent DBS at a ventral internal capsule/ventral striatum target based on that described by Nuttin et al. [16] Eight patients were followed for at least 36 months. Group Y-BOCS scores decreased from 34.6 (mean) at baseline (severe range) to 22.3 (moderate range) at 36 months (P < 0.001). Four of 8 patients had at least a 35% decrease in Y-BOCS severity at 36 months. In 2 other patients, scores declined 25-35%. Depression and anxiety also improved. GAF scores improved from 36.6 at baseline (indicating major functional impairment) to 53.8 (indicating moderate impairment) at 36 months (P < 0.001). This corresponded to improvements in self-care, independent living, and work, school, and social functioning.

This same group presented their findings of DBS in the AIC for treatment of refractory MDD in 2006. Six patients who were highly refractory to medication, psychotherapy, and bilateral electroconvulsive therapy (ECT) were enrolled in the study from 2003-2005. Stereotactic implantation of bilateral DBS leads in the ventral anterior internal capsule was performed. Four of the 6 patients met the response criterion of more than 50% reduction in depression severity on the Montgomery-Asberg depression rating scale at a minimum of 6 months follow-up. Quality-of-life measures also improved, and patients had progressive improvement in mood and functioning over time.

Nucleus accumbens

In 2003, a group from Cologne, Germany, published their results using DBS for OCD symptoms. [37] Their target, the nucleus accumbens (NA) shell, is in a similar anatomic area as that in the groups that use DBS in the ventral anterior internal capsule. Noting the relatively higher voltages used in capsular DBS for OCD, the group chose the NA because of its proximity to the ventral internal capsule and its relationship to the amygdala, the prefrontal/orbitofrontal cortex, and the dorsomedial thalamus, all areas implicated in the pathogenesis of OCD. NA shell DBS was performed on the right side in 4 patients. Three of the 4 patients had a "nearly total recovery" of their OCD, although explicit Y-BOCS scores were not given. The fourth patient was noted to have had the DBS electrode not in the target area, although no information was given regarding reimplantation.

Ventral striatum

In 2004, a group from France presented a case report in which they implanted DBS electrodes into the ventromedial caudate in order to address refractory OCD and depression. They addressed this target specifically because of its anatomic relationships and because they observed that the high currents needed for capsular DBS "raised the essential issue of the exact area to be targeted." [38] They placed the DBS electrode through the head of the caudate nucleus in such a way that the 2 lower contacts were located within the nucleus accumbens and the 2 upper contacts were within the ventromedial portion of the caudate nucleus. Depression symptoms resolved first, at 6 months, with OCD symptoms resolving at 15 months.

Cingulate gyrus

In 2005, Mayberg et al published their results that used a method of DBS for major depression. [39] DBS electrodes were bilaterally implanted into the subcortical white matter in the region of area 25 in 6 patients. Patients were selected for notable but not extreme levels of treatment resistance and for a relative lack of psychiatric comorbidity. Four of the 6 implanted patients, at 6-month follow-up, were determined to have had their depression go into remission, which is defined by a 50% reduction in their Hamilton rating scale for depression scores. In 2008, they published results from an expanded cohort of 20 (including the original 6) that showed 60% response and 35% remission at 6 months after DBS activation [40] . A multicenter trial in 2012 treated an additional 21 patients, with 62% of patients showing at least a 40% reduction in depression severity at 12 months and 29% showing greater than 50% reduction [41] .

Inferior thalamic peduncle

In 2005, Jimenez presented a case report at the 2005 annual meeting of the World Stereotactic and Functional Neurosurgery Society Meeting in Rome, Italy. [42] DBS electrodes were placed bilaterally in the inferior thalamic peduncle (ITP) in a woman with refractory depression. This stimulation, via effects propagated by way of ITP fibers that continue rostrally in the ventral portion of the anterior limb of the internal capsule, is expected to modulate the projections of the dorsolateral prefrontal cortex (DLPFC), the orbitofrontal cortex (OFC), and the ventromedial striatum to the dorsomedial and intralaminar thalamus. After a substantial "microlesion" period (a benefit gained by the mass effect of the peri-electrode edema by the mere implantation of the electrodes), the patient was reported to have had continuing benefit from stimulation in this region. Further exploration and follow-up is necessary to establish whether this approach is both safe and beneficial. [42]

Subthalamic nucleus

The use of the subthalamic nucleus (STN) as a treatment target for OCD was sparked by the serendipitous observation that in two patients with concommitant Parkinson's Disease and OCD, stimulation of the STN resulted in marked improvement of the patients' psychiatric symptoms. [43] This led to a prospective, multi-center, crossover, double blind trial which was published in The New England Journal of Medicine in 2008. [44] The cohort of 16 patients demonstrated significant decrease in symptom severity and significant increase in global function with stimulation on as compared with stimulation off. The adverse event profile was inkeeping with that seen for STN stimulation in Parkinson's Disease.

Vagus nerve stimulation for major depressive disorder

Vagus nerve stimulation (VNS) was FDA approved for use in treatment-resistant depression in May 2005, based on both short-term and long-term data. The short-term data come from a 10-week pilot study in which 60 patients with severe, refractory depression were implanted with VNS for depression. [32] The long-term data come from a larger, double-masked, sham-controlled, 10-week trial with 235 participants, also with severe, refractory depression. [28] In all of these studies, VNS was used adjunctively (in addition to their regular medication regiment, which has, by definition, not brought much relief from the depression).

In the pilot study, 18 (30.5%) of the patients had a response (≥50% reduction from baseline HAMD28 score) and 9 (15.3%) remitted (HAMD28 score of 10 or less). The mean time to response was not acute at 48.1 days. The treatment was well tolerated.

In the larger study, VNS was compared with sham treatment (ie, the implant was not activated) for 12 weeks in patients with unipolar or bipolar depression that was refractory with a mean duration of current depressive episode of over 4 years. This 12-week study did not have 12 weeks of stimulation. VNS was activated after a 22-week postimplant recovery period, and the dose was titrated for another 2 weeks and was then held constant for the remaining 8 weeks. In this study, VNS did not separate from sham. The response rate (HDRS-24) for the study group was 15.2% and the sham group was 10.0% (P =0.251, last occurrence carried forward [LOCF]).

The long-term data come from following the responders in the previously discussed studies. The patients that responded in the pilot study [17] at 3 months (early responders) or 1 year (late responders) were followed for 2 years. [45] The 3-month response rate of 30.5% increased to 44.1% at 1 year and remained at that level (42.4%) at the 2-year time point. At 2 years, 55.6% of early responders and 78.6% of late responders continued to be responders.

All of the eligible patients (N=205) from the pivotal trial had their implants activated and were followed naturalistically after the trial. [28] Patients in the active group in the 3-month trial continued on VNS for an additional 9 months, totaling 12 months. Patients from the sham group in the 3-month trial were activated and followed for 12 months of active VNS. [28] In addition to receiving VNS therapy, all patients continued to receive treatment as usual (TAU). At LOCF end point, the response rate was 27.2%; 15.8% of patients remitted. Rates of response and remission doubled between 3 and 12 months of treatment (P < 0.005), indicating progressive clinical improvement after the initial 3 months of VNS therapy.

Finally, a comparison of VNS with a matched group of TAU patients was undertaken. Data from the 205 patients who completed the 12-month naturalistic study were compared with a matched control group of 124 patients with refractory depression who received only TAU. At the end point, VNS therapy plus TAU was associated with a mean improvement on the IDS-SR30 score of 9.3 points, which represented a significantly greater improvement than TAU (4.2-point improvement; P < 0.001).

The VNS for MDD efficacy database that was used to make VNS approvable has been controversial. Some reviewers are unimpressed with the data set and point out that in the only sham-controlled trial that exists for VNS, the treatment’s primary measure did not separate from placebo. [28] Others point out that this short trial was of insufficient length to show the superiority of VNS and that the secondary measure did actually separate from placebo.


Future and Controversies

Current scope of practice

With the exception of vagus nerve stimulation (VNS), there is no FDA-approved treatment, surgical or otherwise, for refractory OCD or major depression. The postapproval road for VNS has been a rocky one, fraught with controversy and refusal of private payers to reimburse VNS as a "covered benefit." Furthermore, the Centers for Medicare and Medicaid Services (CMS) issued a denial of coverage policy in 2007, citing that the data submitted to the FDA do not meet criteria to be deemed as "necessary and reasonable" for use in depression. VNS trials currently underway hope to address treatment selection and optimum treatment parameters.

The absence of FDA approval, however, is not deterring some practitioners from employing these techniques. A survey of functional neurosurgeons in North America conducted in 2011 found that 40 of the 82 respondants practiced some form of psychosurgery, with the majority of procedures being stimulation rather than lesioning based. [46] A worldwide survey by the same group found that 56 of 104 respondents (53.6%) were practicing psychosurgery in some form. [47]

New targets

A potentially promising target for the treatment of MDD which has not yet been the subject of a formal trial is the lateral habenular complex. Two patients with severe treatment-resistant depression have been implanted at that site to date, and both are reportedly in remission. [48]

New disorders

Anorexia nervosa is a devastating and potentially fatal psychiatric disorder with a high rate of treatment resistance. Two small pilot studies have looked at DBS for anorexia, targeting either the shell of the nucleus accumbens or the subcallosal cingulate. [49, 50] Both studies showed an improvement in the BMI of treated patients, but neither has yet published long term follow-up. There has also been no sham-controlled trial to date.

Substance abuse and addiction disorders exact an extremely high personal and societal toll, with high rates of treatment resistance and relapse. Studies using either stereotactic lesioning or DBS of the nucleus accumbens have shown promise, [51, 52] and other groups have also suggested the lateral habenular complex as a potential target. [53]

New techniques

As discussed earlier, neuroimaging has identified the DLPFC as an important node in the mood circuitry. Repetitive transcranial magnetic stimulation (rTMS), a noninvasive means to focally stimulate brain areas, has been shown to have antidepressive effects in multiple small, uncontrolled trials. [54] rTMS is an intermittent therapy, however, and requires the patient to be present at the therapy delivery site. Two groups have conducted trials of implanted epidural electrodes for direct stimulation of the DLPFC for TRD. [55, 56] While more invasive than rTMS, this technique allows substantially more precise spatial resolution, better temporal continuity, and the ability to test multiple stimulation parameters. Regarding results of the studies, 4/12 patients treated with left DLPFC achieved remission [56] and 3/5 treated with bilateral stimulation achieved remission [55] .

Several promising options are on the horizon. The researchers must now be sure that these new options are safe and effective and past mistakes are not repeated.