Key points

Introduction: why we need neuromodulation for depression

Major depressive disorder (MDD; codes 293.2 and 293.3 in Diagnostic and Statistical Manual of Mental Disorders , 5th edition and the International Classification of Diseases, 9th revision) is a common, disabling, and undertreated condition. With a lifetime prevalence of 1 in 6 and an annual prevalence of 6.6%, at least 20 million Americans will have an episode in 2013. In middle-income and high-income countries, MDD already ranks above ischemic heart disease as causing annually the greatest disability for both men and women. It has been estimated that more than 40% of North Americans with MDD do not receive any treatment.

For those who do receive treatment, the landmark STAR*D study (Sequenced Treatment Alternatives to Relieve Depression) found that less than one-third of adults with MDD remitted with their first medication trial and, thus, a majority need to try something else to aid recovery. The likelihood of achieving remission decreases with each successive pharmacologic treatment while the 12-month rate of relapse increases (71% after 3 failures). There clearly is a need for treatments that have both greater efficacy and durability of benefit.

Treatment-resistant depression generally refers to those patients who remain ill despite repeated vigorous attempts using adequate doses of medication for trials of an adequate duration ; this might better be termed pharmacoresistance than treatment resistance, as it is generally applied in patient care to signify failures of medication treatments. Neuromodulation represents a family of interventions that may have complementary mechanisms of action, and therefore may offer new hope for recovery for patients with pharmacoresistant depression.

The use of neuromodulation for mood disorders is not itself a new idea, but there has been a recent proliferation of approaches. Electroconvulsive therapy (ECT, convulsive therapy) has been in use since the late 1930s, and is believed to achieve a therapeutic alteration of brain activity by the use of electrical currents that pass through the brain and produce a seizure. This seizure activity has been hypothesized to be transduced into clinical benefits through such mechanisms (cf. Bolwig ) as affecting neurogenesis, neuroendocrine regulation, or cytokine levels ; normalizing patterns of cerebral metabolism or glutamatergic neurotransmission ; or altering gene expression. Recent efforts to address the cognitive side effects by focusing the stimulation have led to innovations such as focal electrically administered seizure therapy (FEAST ) and magnetic seizure therapy (MST ). At the other end of the spectrum, methods for nonconvulsive, low-energy stimulation of the brain have been examined, with approaches such as transcranial direct current stimulation (tDCS) and transcranial alternating current stimulation (tACS), as forms of cranial electrotherapy stimulation (CES), whereby current flux lines pass through brain tissue but no seizure is effected.

From a long and expanding list of neuromodulation approaches with potential use in depression, the editors of this issue have asked the authors to focus on just 3: deep brain stimulation (DBS), transcranial magnetic stimulation (TMS), and trigeminal nerve stimulation (TNS). All 3 of these approaches have histories in which the developments of neurologic and psychiatric applications have gone hand in hand. One of these interventions (TMS) has regulatory approval by the US Food and Drug Administration (FDA) for use in depression, and is reimbursed by some insurance carriers in the United States, whereas the other approaches do not yet have US regulatory approval; these experimental devices are limited to investigational use in the United States, although they may have regulatory approval in other countries. These treatments differ notably in terms of their mode of delivery and risk profile: DBS involves a neurosurgical implantation procedure, whereas the TMS is noninvasive, and TNS can be administered either noninvasively or with a minimally invasive approach. This article describes salient features of each of these interventions. For each intervention, the authors primarily have surveyed controlled studies that have been completed or other peer-reviewed data that have been published and are widely accessible, and have noted where salient trials are ongoing or completed but for which published data are not available at the time of writing. Across all modalities, Table 1 summarizes critical aspects of stimulation to consider in evaluating a clinical trial, namely, the anatomic target(s) being stimulated, and the nature of signal (electrical or magnetic) being used for stimulation.

Introduction: why we need neuromodulation for depression

Major depressive disorder (MDD; codes 293.2 and 293.3 in Diagnostic and Statistical Manual of Mental Disorders , 5th edition and the International Classification of Diseases, 9th revision) is a common, disabling, and undertreated condition. With a lifetime prevalence of 1 in 6 and an annual prevalence of 6.6%, at least 20 million Americans will have an episode in 2013. In middle-income and high-income countries, MDD already ranks above ischemic heart disease as causing annually the greatest disability for both men and women. It has been estimated that more than 40% of North Americans with MDD do not receive any treatment.

For those who do receive treatment, the landmark STAR*D study (Sequenced Treatment Alternatives to Relieve Depression) found that less than one-third of adults with MDD remitted with their first medication trial and, thus, a majority need to try something else to aid recovery. The likelihood of achieving remission decreases with each successive pharmacologic treatment while the 12-month rate of relapse increases (71% after 3 failures). There clearly is a need for treatments that have both greater efficacy and durability of benefit.

Treatment-resistant depression generally refers to those patients who remain ill despite repeated vigorous attempts using adequate doses of medication for trials of an adequate duration ; this might better be termed pharmacoresistance than treatment resistance, as it is generally applied in patient care to signify failures of medication treatments. Neuromodulation represents a family of interventions that may have complementary mechanisms of action, and therefore may offer new hope for recovery for patients with pharmacoresistant depression.

The use of neuromodulation for mood disorders is not itself a new idea, but there has been a recent proliferation of approaches. Electroconvulsive therapy (ECT, convulsive therapy) has been in use since the late 1930s, and is believed to achieve a therapeutic alteration of brain activity by the use of electrical currents that pass through the brain and produce a seizure. This seizure activity has been hypothesized to be transduced into clinical benefits through such mechanisms (cf. Bolwig ) as affecting neurogenesis, neuroendocrine regulation, or cytokine levels ; normalizing patterns of cerebral metabolism or glutamatergic neurotransmission ; or altering gene expression. Recent efforts to address the cognitive side effects by focusing the stimulation have led to innovations such as focal electrically administered seizure therapy (FEAST ) and magnetic seizure therapy (MST ). At the other end of the spectrum, methods for nonconvulsive, low-energy stimulation of the brain have been examined, with approaches such as transcranial direct current stimulation (tDCS) and transcranial alternating current stimulation (tACS), as forms of cranial electrotherapy stimulation (CES), whereby current flux lines pass through brain tissue but no seizure is effected.

From a long and expanding list of neuromodulation approaches with potential use in depression, the editors of this issue have asked the authors to focus on just 3: deep brain stimulation (DBS), transcranial magnetic stimulation (TMS), and trigeminal nerve stimulation (TNS). All 3 of these approaches have histories in which the developments of neurologic and psychiatric applications have gone hand in hand. One of these interventions (TMS) has regulatory approval by the US Food and Drug Administration (FDA) for use in depression, and is reimbursed by some insurance carriers in the United States, whereas the other approaches do not yet have US regulatory approval; these experimental devices are limited to investigational use in the United States, although they may have regulatory approval in other countries. These treatments differ notably in terms of their mode of delivery and risk profile: DBS involves a neurosurgical implantation procedure, whereas the TMS is noninvasive, and TNS can be administered either noninvasively or with a minimally invasive approach. This article describes salient features of each of these interventions. For each intervention, the authors primarily have surveyed controlled studies that have been completed or other peer-reviewed data that have been published and are widely accessible, and have noted where salient trials are ongoing or completed but for which published data are not available at the time of writing. Across all modalities, Table 1 summarizes critical aspects of stimulation to consider in evaluating a clinical trial, namely, the anatomic target(s) being stimulated, and the nature of signal (electrical or magnetic) being used for stimulation.

Deep brain stimulation for depression

DBS uses the stereotactic neurosurgical implantation of stimulating electrodes at 1 or more specific anatomic target locations in the brain, then applies electrical currents at controlled signal parameters to achieve the intended therapeutic effects. Direct stimulation of the brain was observed to affect complex behaviors in the mid-twentieth century (by, eg, Delgado ). However, societal fears regarding the adequacy of informed consent for neurosurgical procedures to treat mental illness prompted establishment of legal constraints in some jurisdictions (eg, California Welfare and Institutions Code § 5325 from 1967). The use of DBS for treating neurologic conditions, such as movement disorders, falls outside these legal considerations. Building on a pathophysiologic model of an electrically induced reversible or simulated lesion that could be adjusted postoperatively, unlike once-only ablative surgical interventions (eg, pallidotomy, thalamotomy), DBS systems were approved by the FDA for essential tremor in 1997, for Parkinson disease in 2002, and for dystonia in 2003. A humanitarian device exemption (HDE) approval was granted by the FDA for use in obsessive-compulsive disorder (OCD) in 2009, as an initial psychiatric indication for a DBS system. (HDE is an FDA regulatory approval pathway for devices that will be used in a very limited number of patients each year and does not have the same level of clinical trial evidence to support safety and efficacy as a regular pathway [i.e., investigational device exemption pivotal study followed by a premarket approval to permit sales]. Some have drawn an analogy to orphan drugs.) CE Mark regulatory approval for OCD was issued the same year. (CE Mark [Conformité Européenne] is a requirement for certain medical [and other] devices to be made commercially available in the European Economic Area, namely, the 27 member states of the European Union plus Iceland, Norway, and Liechtenstein.)

Effects on mood were observed in many of the patients implanted for the treatment of neurologic conditions, prompting interest in developing DBS specifically for use in depression. Several different targets have been explored for potential in the management of depression (cf. Hauptman and colleagues, Anderson and colleagues ), guided by different insights into structures implicated in the pathophysiology of mood disorders and the behavioral, cognitive, and affective symptoms of depression. These structures include: (1) subgenual (subcallosal) cingulate cortex, Brodmann area 25 (SCG); (2) ventral anterior internal capsule/ventral striatum (VC/VS); (3) nucleus accumbens (NAcc) and ventral striatum; (4) inferior thalamic peduncle (ITP); (5) lateral habenula (LH); (6) median forebrain bundle (MFB); and (7) internal globus pallidus (GP) ( Fig. 1 ). Other targets have been proposed (cf. Hauptman and colleagues ), but no publications about clinical effects for these other locations could be found via online searches.

Common targets of deep brain stimulation. Several anatomic targets for stimulation have been examined: GP, internal globus pallidus (not shown); ITP, inferior thalamic peduncle; LH, lateral habenula; MFB, median forebrain bundle; NAcc, nucleus accumbens; SCG, subgenual cingulated gyrus; VC/VS, ventral capsule/ventral striatum.

Table 2 summarizes key publications, findings, and considerations for each of these targets. Of note, none of these have used a parallel-group, double-blind, randomized controlled trial (RCT) model. A review of preclinical studies and an accompanying commentary amplified some of the rationale for the targets, and also highlight nuanced aspects of stimulation beyond the reversible lesion concept (eg, hyperpolarization of 1 part of a neuron while other portions are depolarized) that must be considered in trying to understand the potential mechanisms of action at work in DBS. A recent report by Ramasubbu and colleagues examined the relationship between stimulation parameters and acute and chronic symptom responses using a systematic approach to explore variations in the type of stimulation (see Table 1 ). The investigators reported that clinical response was related more to pulse width than to frequency of stimulation, a finding that emphasizes that additional studies to explore different stimulation parameters may be important in improving clinical outcomes in patients using DBS to address depression.

Regardless of target or the particular device being used, treatment with DBS is generally undertaken by a multidisciplinary team, including psychiatrists and stereotactic neurosurgeons, working together with neuropsychologists and other support staff. Because implantation is conventionally performed bilaterally, it is common to place an electrode into the target of each hemisphere, and each is then connected to a programmable neurostimulator (also called an implantable pulse generator or IPG); the IPGs are often placed pectorally, near the clavicle, and are connected to the implanted leads via subcutaneously tunneled extension wires. Experimental protocols in depression have followed the example of those used for neurologic indications, and subjects often wait for 2 or more weeks postoperatively before the device is activated and a period of adjustment is begun to find the best settings. In this process, physicians select the specific electrode contact(s) to be stimulated (ie, to fine-tune the exact anatomic region being stimulated), and adjust key signal parameters such as signal amplitude (current or voltage, depending on the device), pulse repetition frequency, and pulse width, to achieve the desired symptom control. Individuals are seen frequently (commonly weekly or biweekly for 1 to 3 months, then monthly) in the year following implantation to allow close monitoring, and usually less often thereafter. This monitoring includes assessments of clinical response as well as of tolerability and side effects. It is worth noting that DBS is currently used adjunctively as an addition to ongoing pharmacotherapy and psychotherapy.

Complications of DBS affect a minority of individuals, and can best be conceptualized as those related to (1) the surgical procedure acutely (eg, intracranial hemorrhage, postoperative infection, stroke), (2) the chronic presence of the device in the body (eg, lead migration or erosion), or (3) the stimulation signal itself (eg, undesired behavioral or affective symptom changes). Adverse events in this last category have included the development of manic/hypomanic states, anxiety symptoms, and worsening of depression, but these generally have been found to respond to adjustments in the stimulation parameters. Suicides have been reported after DBS implantation (eg, Kennedy and colleagues ), but it is challenging to differentiate reliably the risk ascribed to device stimulation as opposed to the risk associated simply with chronic unremitting depression.

At present, no DBS system has received FDA approval for use with an indication for depression, either unipolar MDD or depression in bipolar-spectrum disorders, although trials for these indications are currently registered in the ClinicalTrials.gov database. Because clinical benefits only emerge over months to years of use, DBS for treatment-resistant depression is not an acute intervention, and other approaches continue concomitantly at least until a response is obtained. Should regulatory approval occur, health economic factors will be considered as the questions of cost-benefit and cost-effectiveness that affect reimbursement and insurance coverage policies are addressed. Initial DBS device and implantation costs may exceed US$200,000 per patient, raising questions about how this intervention might best fit into treatment algorithms in this era of attention to health care costs, some of which may be addressed through comparative effectiveness studies of at least 12 months’ duration. As points of comparison, vagus nerve stimulation (VNS) received FDA regulatory approval in 2005, but in 2007 the Centers for Medicare and Medicaid Services issued a decision that VNS would not be a covered benefit, a decision which it reiterated in 2013. The situation with TMS is discussed next, but it is clear that FDA approval is a necessary but not sufficient step for a treatment to become accessible to patients.