Neurostimulation Therapies

Abstract

Depression is one of the most disabling conditions in the world. In many cases patients continue to suffer with depressive disorders despite a series of adequate trials of medication and psychotherapy. Neuromodulation treatments offer a qualitatively different modality of treatment that can frequently prove efficacious in these treatment-refractory patients. The field of neuromodulation focuses on the use of electrical/electromagnetic energy, both invasively and noninvasively, to interface with and ultimately alter activity within the human brain for therapeutic purposes. These treatments provide another set of options to offer patients when clinically indicated, and knowledge of their safety, risks and benefits, and appropriate clinical application is essential for modern psychiatrists and other mental health professionals. Although neuromodulation techniques hold tremendous promise, only three such treatments are currently approved by the United States Food and Drug Administration (FDA) for the treatment of major depressive disorder: electroconvulsive therapy (ECT), vagus nerve stimulation (VNS), and repetitive transcranial magnetic stimulation (rTMS). Additionally, numerous other neurostimulation modalities (deep brain stimulation [DBS], magnetic seizure therapy [MST], transcranial electric stimulation [tES], and trigeminal nerve stimulation [TNS]), though currently experimental, show considerable therapeutic promise. Researchers are actively looking for ways to optimize outcomes and clinical benefits by making neuromodulation treatments safer, more efficacious, and more durable.

Depression is one of the most disabling conditions in the world today, both on an individual patient level and a global scale. Since the advent of antidepressant medications in the late 1950s, progress toward newer and more effective treatments has been challenging. With the possible exception of electroconvulsive therapy (ECT), pharmacotherapies targeting neurotransmitter systems have been the staple depression treatment for the past 60 years. For the most part, attempts to improve upon earlier generations of antidepressant drugs have struggled to demonstrate significantly greater benefits or novel mechanisms of action. Further, the challenge of managing treatment-refractory depression reaches an additional level of complexity, as many of these patients will not respond to a series of antidepressant trials and continue to suffer for years, even decades, with debilitating depression.

Neurostimulation therapies offer a unique approach for effecting change within the brain and nervous system. Focusing primarily on the use of electrical and magnetic forces, this category of therapeutics has ushered in a new and different way of not only thinking about studying mental illness and neuropsychiatric pathology in the brain but also an alternate strategy for modulating the brain and its neural elements to benefit patients struggling with depression and various other disorders of the brain. Encompassing older methods such as electroconvulsive therapy and newer technologies such as transcranial magnetic stimulation and vagus nerve stimulation, neurostimulation therapies have proven efficacious and exciting options for clinicians and researchers to utilize and explore. This chapter focuses on neurostimulation therapies and the role they play in the treatment of depression and other illnesses, with a focus on clinical methodology and use.

1 Electroconvulsive Therapy

1.1 Introduction

Electroconvulsive therapy (ECT) is one of the oldest and most effective psychiatric treatments still in regular use today. Despite repeated demonstrations of ECT’s efficacy in depression and other psychiatric conditions, it remains controversial due to apprehensions over adverse effects and misunderstandings about the treatment itself. Modern techniques have minimized prior safety concerns, and continued research and methodological advances have led to a renewed interest in ECT in recent decades.

1.2 Historical Background

The use of chemically induced seizures dates to the 1930s, when the Hungarian neuropsychiatrist Meduna hypothesized a “biological antagonism” between schizophrenia and epilepsy in patients, such that having seizures was “protective” against psychosis (Coffey and Weiner 1990). Despite Meduna’s success in treating patients with schizophrenia with camphor-induced seizures, the use of this modality was painful and unreliable. In 1938, the Italian physicians Cerletti and Bini successfully administered the first electrically induced therapeutic seizure on a patient (Coffey and Weiner 1990). Subsequently, the practice of convulsive therapy rapidly spread throughout Europe and the United States and became a dominant form of somatic therapy for psychiatric disorders.

The use of ECT waned in the 1950s with the development of antipsychotic and antidepressant drugs. Misconceptions promulgated by sensationalized depictions of ECT, such as portrayed in the film One Flew Over the Cuckoo’s Nest, and negative media portrayals led to further public disinterest (Jenkusky 1991). Although early attempts at treatment were riddled with injuries and complications, techniques have dramatically evolved with the routine use of anesthetics, muscle relaxants, oxygenation, and seizure monitoring. In recent decades, ECT has experienced a resurgence, partly in light of a growing recognition of its benefits and safety profile and partly due to the limitations of existing pharmacologic modalities (Lisanby 2007).

1.3 Mechanisms of Action

Despite decades of research and experience, the exact mechanism underlying ECT’s benefits has yet to be elucidated. The possibility of a placebo or psychological effect was eliminated by multiple trials demonstrating significant efficacy of ECT over sham ECT (Janicak et al. 1985; UK ECT Review Group 2003; Tharyan and Adams 2005). Sackeim (1994) has described over 100 other theories, including alterations in neurotransmitters, the hypothalamic-pituitary-adrenal axis, neuroendocrine pathways, neurophysiological changes, and synaptic plasticity. The role of the electrical seizure itself has been a research subject of much interest. Cronholm and Ottosson (1996) classic research demonstrated that ECT is ineffective when the seizure is pharmacologically blocked by lidocaine. However, although the presence of a seizure is necessary, it is not sufficient for clinical efficacy: studies have established a lack of efficacy from marginally suprathreshold yet seizure-inducing treatments. Further, the dose of electrical charge relative to seizure threshold used to trigger generalized seizures appears to play a critical role in determining efficacy depending on electrode placement (i.e., unilateral vs. bilateral vs. bifrontal lead placement) (Sackeim et al. 1993, 2000).

1.4 Indications (Table 1)

1.4.1 Unipolar Depression

ECT is well-established as the most effective treatment for major depressive episodes, with response rates of 70–90% and remission rates well over 40–50%, even in treatment-resistant patients (Prudic et al. 1996; UK ECT Review Group 2003; Kellner et al. 2006). However, ECT is largely utilized as a secondary treatment in patients who have failed to respond to a series of adequate antidepressant trials. Given that an “adequate” trial generally refers to 4–6 weeks of a medication at a therapeutic dose, this entire process of titration, waiting, and switching may span months and even years, leaving the patient with prolonged and unnecessary suffering (Beale and Kellner 2000).

According to practice guidelines from the American Psychiatric Association, ECT should be considered as first-line treatment for major depression with high symptom severity, such as concurrent psychosis or catatonia, and urgent cases necessitating a rapid symptomatic response (American Psychiatric Association 2010). This includes patients who are at high suicide risk, nutritional compromise (e.g., food refusal, dehydration or failure to thrive from loss of appetite), and severely ill inpatients. There is greater consensus on initiating ECT for such emergently ill cases than for medication-resistant depression. ECT is also indicated as first-line treatment for patients who have previously shown a positive ECT response or patients who self-request (American Psychiatric Association 2010).

1.4.2 Bipolar Disorder

ECT has been less extensively studied for bipolar depression. Several older, controlled studies have found ECT to be as or more effective than MAOIs, tricyclics, and placebo (Zornberg and Pope 1993). As with unipolar depression, ECT is a reasonable treatment in bipolar depression associated with life-threatening conditions, psychotic or catatonic features, affective symptoms occurring during pregnancy, as well as treatment-resistant cases (American Psychiatric Association 2002). ECT is also efficacious for acute mania, with marked improvements in approximately 80% of patients (Mukherjee et al. 1994). In small prospective comparison studies, ECT was superior in efficacy to lithium and the combination of lithium and haloperidol (Small et al. 1988; Mukherjee et al. 1994). ECT may play a role in the treatment of delirious mania, rapid-cycling mania, treatment-refractory cases (Perugi et al. 2017), as well as emergent situations (e.g., mania leading to physical exhaustion) and pregnant patients with severe mania (American Psychiatric Association 2002).

Mixed episodes in bipolar disorder are notoriously difficult to treat with conventional pharmacologic modalities (Kruger et al. 2005). In a prior trial of ECT for patients with mania, the strongest predictor of clinical response was baseline depressive symptoms, suggesting ECT’s possible benefits in mixed episodes (Small et al. 1988). Case reports have also proposed that ECT can successfully treat mixed states, but no prospective, randomized controlled studies have been completed (Devanand et al. 2000; Gruber et al. 2000; Ciapparelli et al. 2001).

1.4.3 Catatonia

Catatonia may develop in up to one third of patients during mania and is associated with increased episode severity and poor short-term outcomes (Braunig et al. 1998). Antipsychotic medications have had relatively poor utility in treating catatonia. Typically, benzodiazepines are the drugs of choice in catatonia; however, their use is based largely on anecdotal cases or small studies with deficiencies in methodology or reporting of findings, as was recently summarized in a Cochrane review (Hawkins et al. 1995; Gibson and Walcott 2008). ECT should be considered if benzodiazepines do not improve symptoms or if immediate resolution is necessary (e.g., malignant catatonia).

1.4.4 Schizophrenia

Several trials have found ECT to be less effective than antipsychotic medications, particularly clozapine, as first-line treatment in patients with schizophrenia (Small 1985; Tharyan and Adams 2005). However, ECT has shown therapeutic benefit when given together with antipsychotics, particularly if rapid symptomatic improvement is desired (Tharyan and Adams 2005). Patients with catatonia or prominent affective symptoms may be more likely to respond to treatment (Konig and Glatter-Gotz 1990). ECT is recommended by the APA for psychosis resistant to antipsychotics, particularly failure of clozapine, as well as catatonia and emergent cases (American Psychiatric Association 2004). Although only 5–10% of ECT courses in the United States are given for patients with schizophrenia, many developing countries continue to use ECT for this indication, as the treatment is relatively available, effective, and inexpensive (Leiknes et al. 2012).

1.4.5 Other Conditions

ECT may be the treatment of choice in cases that preclude the use of medications, particularly due to safety concerns; this includes geriatric patients, the medically ill, and even pregnant patients who wish to avoid the potential teratogenic side effects of psychotropics. ECT also has potential benefits in the on-off phenomenon of Parkinson’s disease, intractable epilepsy, and neuroleptic malignant syndrome (Faber and Trimble 1991; Trollor and Sachdev 1999; Lisanby et al. 2001). ECT is not effective for treating personality disorders, and, in fact, comorbid personality disorders have been associated with decreased efficacy (Prudic et al. 2004).

Table 1 Indications for electroconvulsive therapy

1.5 Contraindications

While there are no absolute contraindications to ECT, certain patients are at increased risk for complications. These include individuals with recent myocardial infarction, unstable symptomatic cardiac disease (e.g., arrhythmias, severe hypertension, unstable angina), decompensated congestive heart failure, and increased intracranial pressure (e.g., space-occupying brain lesions) and those at increased risk for cerebral bleeding (e.g., recent hemorrhage, unstable aneurysms (Lisanby 2007)). Some patient populations may also be deemed to be of higher anesthesia risk due to underlying medical or surgical conditions and will require closer anesthesia monitoring. ECT has been shown to be safe in pregnancy, the elderly, and persons with cardiac pacemakers or implantable cardioverter-defibrillators (Dolenc et al. 2004a).

1.6 Administration of ECT

1.6.1 Pre-treatment Evaluation

A thorough pre-ECT work-up should be conducted to detect the aforementioned conditions, which may increase a patient’s risk of ECT-related adverse events. This entails a detailed medical and psychiatric history, physical and neurological exam, pre-anesthesia exam conducted by the anesthesia team, basic lab work (i.e., complete blood count, serum electrolytes), and an electrocardiogram (EKG). Other imaging is typically indicated only on an as-needed basis, for example, a chest x-ray for suspected acute pulmonary disease, a spinal x-ray in cases of severe osteoarthritis/osteoporosis, and brain imaging if specific concerns for neurological condition arise (e.g., concerns about stroke, previous history of head trauma or skull fracture). Patients with histories of skull fracture should have careful identification of the location of the healed sutures, so lead placement is distant from these sites to prevent electricity from directly entering the brain.

1.6.2 Preparation for Treatment

All of the patient’s medications must be carefully reviewed for potential adverse effects or interactions. Benzodiazepines are typically reduced or withdrawn due to their anticonvulsant activity and potential for increasing postictal confusion. Antiepileptics also raise seizure threshold and are often decreased or discontinued, though patients with comorbid depression and epilepsy may require these medication to avoid spontaneous or prolonged seizures, in which case minimizing the antiepileptic dose becomes the goal. Concurrent administration of lithium may increase the risk for cognitive disturbances, delirium, and spontaneous seizures, although recent case reviews and prospective studies have challenged this (Dolenc and Rasmussen 2005; Thirthalli et al. 2011); clinicians should weight the risks of neurotoxicity against relapse of symptoms on a case-by-case basis. MAOIs were historically recommended to be discontinued prior to ECT, but the current literature suggests that they are safe to continue (Dolenc et al. 2004b; Horn et al. 2010). Non-psychotropic medications that should be avoided include theophylline, which increases seizure duration, lidocaine, which increases seizure threshold, and reserpine, which compromises the cardiorespiratory system.

As with any procedure requiring general anesthesia, patients must not have anything to eat or drink for 6–8 h prior to treatment. Routine, medications needed for significant underlying medical illnesses can be administered by mouth, however, with minimal amounts of water. The patient’s mouth should be checked for foreign bodies or loose teeth, and dentures should be removed. Fake nails or nail polish that could interfere with pulse oximetry should also be avoided.

1.6.3 Anesthesia

ECT is performed under general anesthesia in order to prevent injuries and control the pronounced sympathetic response and subsequent hemodynamic changes associated with seizure activity. The procedure is typically administered in a specialized suite in some centers and in a standard operating or recovery room in others. A peripheral intravenous catheter is inserted to administer medications, and a bite block inserted in the mouth just prior to electrical stimulation to protect the teeth and tongue. The patient’s blood pressure, pulse rate, EKG, and oxygen saturation are carefully monitored throughout treatment.

The patient is pre-oxygenated with 100% oxygen by mask and is continued on mask ventilation until the procedure ends and the patient resumes normal respiration. Neuromuscular blocking agents are administered to prevent skeletal muscle contractions and injuries during tonic-clonic convulsions. Succinylcholine is the paralytic agent of choice for nearly all individuals but may lead to prolonged paralysis in those with pseudocholinesterase deficiencies; non-depolarizing muscle relaxants are used in these cases instead. Muscarinic anticholinergics (e.g., glycopyrrolate, atropine) are given on a case-by-case basis to minimize salivation and bradycardia. A short-acting anesthetic, typically low-dose etomidate, methohexital, or propofol, is used for anesthesia. Despite propofol producing shorter duration seizures than methohexital, seizure quality and therapeutic outcomes of the two agents have been comparable (Mårtensson et al. 1994; Geretsegger et al. 2007).

1.6.4 Electrode Placement

Three electrode arrangements are commonly used in practice today: bilateral, right unilateral (RUL), and bifrontal. Bilateral placement offers more predictable efficacy and speed of response than RUL, but this placement is associated with considerably greater cognitive side effects. Sackeim et al. (2000) found that RUL ECT treatment, with an electrical stimulus 500% above the seizure threshold, produces therapeutic effects comparable to that seen with bilateral ECT with less cognitive dysfunction. A common treatment strategy is to start with a course of RUL and then switch to bilateral placement if the patient does not respond after adjustments in stimulus dosing (Mankad et al. 2010) (Fig. 1).

Fig. 1

Electroconvulsive therapy electrode placements

Bifrontal lead placement is not as well studied as bilateral and unilateral placements. It was initially believed to be as efficacious as bilateral placement, but with less cognitive side effects (Letemendia et al. 1993; Bailine et al. 2000). However, recent studies have failed to confirm these advantages, with the majority of research indicating that bifrontral lead placement has efficacy and cognitive side effects comparable to RUL placement (Eschweiler et al. 2007; Bjølseth et al. 2015; Dybedal et al. 2016). Further studies are still needed to clarify its place among the other treatment techniques.

1.6.5 Electrical Stimulus

Stimulus dosing should be adjusted on an individual basis to induce an adequate generalized seizure. A dosage titration technique is commonly employed as follows: (1) the approximate seizure threshold is determined during the initial ECT session by a method of limits approach; (2) once the seizure threshold is determined, subsequent sessions are performed at a factor of the threshold (6× seizure threshold for right unilateral treatment; 1.5–2× seizure threshold for bilateral (Coffey et al. 1995; Mankad et al. 2010)). Stimulus dosing can also be established via an age-based dosing algorithm, but this method is limited by the fact that age alone does not adequately account for the variance in individual seizure thresholds (Coffey et al. 1995). Of note, existing dosing research has been performed only for bilateral and RUL placement, while little is known regarding optimal dosing for bifrontal placement.

Originally, ECT was delivered via a sinusoidal electrical waveform, which was physiologically inefficient and resulted in significant cognitive side effects (Peterchev et al. 2010). Present ECT machines use constant-current brief (0.5–2.0 ms) or ultrabrief (≤0.5; typically 0.5 or 0.3 ms) rectangular pulses. Compared to sine wave generators, seizures can now be induced at significantly lower charges, with less adverse effects and equivalent efficacy (Peterchev et al. 2010).

1.6.6 Seizure Monitoring

The optimal ECT seizure lasts at least 20 s in motor duration and 30 s in electroencephalogram (EEG) duration (American Psychiatric Association 2001). Because treatments are performed under muscle relaxation, motor duration should be assessed by occluding flow of the muscle relaxant into the right ankle with an inflated blood pressure cuff. One can then observe the unmodified seizure in the foot and count how many seconds the seizure episode lasts. The ankle ipsilateral to the side of stimulation in unilateral treatments is used to ensure generalization of the seizure.

1.6.7 Treatment Course

A typical ECT course consists of 6–12 treatments administered 2–3 times/week, with the actual number of treatments dependent on the patient’s clinical response and side effects. Given a significant risk of symptom relapse, all patients must receive maintenance treatment, either pharmacologic, continued ECT, or both (American Psychiatric Association 2001). If offered as maintenance, ECT treatments may be gradually decreased in frequency to monthly visits eventually. This process remains highly empirical due to a paucity of data on the optimal maintenance interval. Even with vigorous maintenance therapy, the relapse rate remains substantial at an estimated 40–50% at 6 months (Sackeim et al. 2001; Kellner et al. 2006); without effective maintenance treatment (e.g., placebo), relapse rates exceed 80%. Furthermore, Prudic et al. (2004) noted high relapse rates in community settings compared to clinical trials, possibly secondary to less aggressive maintenance treatment or premature discontinuation of ECT.

1.7 Side Effects

ECT-associated cognitive effects have been the subject of intense investigation. Patients commonly experience a brief period of disorientation and even delirium following the seizure and emergence from anesthesia (Lisanby 2007). ECT can cause anterograde amnesia, or inability to recall newly learned information, which is short-lived and resolves after ECT is terminated (Lisanby 2007). Patients may also experience retrograde amnesia, or the forgetting of information learned before treatment, extending back months or years. This retrograde memory loss improves in the first few months after completing ECT in some individuals, but may be prolonged in others (Lisanby 2007). In addition, ECT may affect memory of prior personal events – that is, autobiographical memory – predominantly those occurring within 6 months of treatment, although some subjective accounts have reported this amnesia to be persistent as well (Fraser et al. 2008).

The cognitive deficits from ECT are related to specific factors, including bilateral electrode placement, higher stimulus dosages, and longer pulse waveforms (Sackeim et al. 2008). Older age, baseline cognitive status, and co-administration of medications such as anticholinergics may also be contributing factors. Because psychiatric disorders, particularly depression, are associated with deficits in concentration and executive functioning, some patients report an improvement in memory, as well as quality of life, following ECT (Prudic et al. 2000; McCall et al. 2006). A meta-analysis also showed that multiple cognitive measures, including processing speed, anterograde memory, and executive function, improved beyond baseline levels after 15 days (Semkovska and McLoughlin 2010).

Other side effects experienced by patients receiving ECT include muscle soreness, jaw pain, headaches, and nausea. In most instances, these lesser side effects can be resolved with analgesics (e.g., intravenous NSAIDs) and antiemetics given during the treatment. Rarely, episodes of hypotension, hypertension, tachycardia, and transient arrhythmias may occur with seizure activity; such cases require optimizing blood pressure pre-treatment and administering antihypertensives or antiarrhythmics as needed (American Psychiatric Association 2001). Prolonged seizures of greater than 120 s during ECT are also extremely rare and can be terminated via intravenous administration of either a repeat anesthetic dose or a short-acting benzodiazepine (e.g., midazolam 1–2 mg); this should be repeated after 2 min if the seizure is still not aborted (Greenberg 1985; Mankad et al. 2010). The overall mortality rate of ECT is estimated to be 1 per 10,000 patients or 1 per 80,000 treatments, which is no greater than the risk associated with general anesthesia alone and notably less than the mortality rate of inadequately treated depressed patients (Avery and Winokur 1976; American Psychiatric Association 2001).

2 Summary

ECT is backed by an extensive evidence base supporting its use and remains a standard psychiatric treatment around the world. Despite the extensive knowledge base regarding various aspects of treatment, much is left to be discovered, particularly in relation to how ECT works mechanistically and how to further reduce cognitive effects. It is crucial for clinicians to be aware of recent advances in technique that have allowed ECT to be even more effective and better tolerated than ever before, as well as to educate patients and counter the stigma surrounding this well-established treatment.

3 Repetitive Transcranial Magnetic Stimulation (rTMS)

3.1 Introduction

Transcranial magnetic stimulation is a unique tool that can be used to noninvasively induce an electrical current in the brain. This electrical current is induced via the creation of a temporary magnetic field that can pass unimpeded into the brain tissue and depolarize neurons, primarily in the cortex. As a magnetic field is not shunted or deflected by the human skull in the same way that an electrical current is (e.g., as in ECT), this allows TMS to be utilized in a manner that is safe and well-tolerated by patients while they are awake and alert. Transcranial magnetic stimulation administers brief “pulses” of electromagnetic energy that stimulate a small area (~2 cm) of the cortex of brain (Thielscher and Kammer 2002; Deng et al. 2013).

In addition to the potential clinical uses of TMS, researchers have used this modality to probe different areas of the brain and look for perceptual, behavioral, or motor changes to assess for cortical integrity. For example, a stimulation applied over the primary motor cortex of the brain will induce a motor-evoked potential (i.e., a small muscle twitch) in the region of the body controlled by the population of neurons in the cortex being stimulated. Hence, assuming the motor pathway is intact, stimulating the hand representation region in the motor cortex results in a muscle twitch in the hand.

Repetitive TMS (abbreviated rTMS) is a process of repeatedly pulsing electromagnetic stimuli into certain regions of the brain with specific parameters. Studies have shown that rTMS can induce long-lasting functional changes, which can then be used for therapeutic purposes. For this reason, rTMS is being explored for a host of different neuropsychiatric conditions. Although rTMS is a promising new tool for clinicians and researchers in neurology, neurorehabilitation, psychiatry, and neuroscience, there remains a great deal to be learned about the mechanism of action and how to optimize the treatment to interface with the brain in a therapeutic manner.

3.2 Brief History and Physics

A TMS device is a large capacitor that simply produces and stores an electrical charge. This device is attached to a “wand” that rests upon the patient’s scalp. The wand contains a copper coil through which the electrical charge from the capacitor flows. Electrical current passing through a coil induces a magnetic field. This magnetic field can pass right through the scalp and skull. Since the brain uses electrical energy as a method for transmitting signals between neurons and down nerves to other areas of the body, it can “translate” this magnetic field back into electrical energy, thus inducing an action potential in the neurons under the coil in the wand (Fig. 2) (Frye et al. 2013).

Fig. 2

Creation of a magnetic field (Adapted from Frye et al. (2013))

The shape of the magnetic field determines how large an area of cortex is affected; hence, different magnetic coils can create different magnetic fields in the brain (Deng et al. 2013). The shape of the field, in addition to several other parameters, determines how deep the field will penetrate as well as how diffusely across the cortex the field will spread. There will be considerably different effects created by stimulating a small focal area versus a large swath of brain tissue, and researchers are actively investigating the best way to shape magnetic fields to address different research questions/disease conditions. Additionally, different TMS device manufacturers produce different coil shapes with different magnetic fields. It is important for TMS clinical practitioners to understand the properties of different magnetic fields so as to best understand how the rTMS treatment they are administering is interacting with cortical neurons.

Although achieving more popularity and attention in the modern era, the notion of using electromagnetic energy to induce changes in the brain is not a new one. The concept of electromagnetic induction was first described by Michael Faraday in 1831, and since then researchers have been attempting various forms of transcranial magnetic stimulation. D’Arsonval (Vidal-Dourado et al. 2014) first produced phosphenes by placing a volunteer’s head into a magnetic coil in the late 1800s; this was later confirmed by the work of Silvanus Thompson in 1910 (Thompson 1910). The modern TMS device and coil was developed by Anthony Barker and colleagues in 1974, and the first stimulation of the human motor cortex took place in 1985 (Vidal-Dourado et al. 2014). Since then, there has been a very large expansion of TMS uses, both investigative and clinical.

3.3 How Does TMS Work?

It has been clearly demonstrated that a TMS pulse will induce an electrical current in the effected neuronal elements. The most dramatic forms of this occur with TMS of the primary motor cortex, which can produce a visible muscle twitch, and TMS of the primary visual cortex, which can lead to a patient experiencing phosphenes in their visual field. Critically, TMS pulses can travel “transynaptically,” meaning that a pulse can cross neuronal synapses and effect neurons further downstream from the stimulated neuron. This ability provides the rationale for how TMS can act both locally and across different brain networks to induce effects. This is important to recognize because the astute TMS practitioner can take advantage of “connectivity pathways” in the brain by applying current to specific cortical “nodes” in an effort to induce downstream (perhaps subcortical) effects on networks thought to be pathological; hence, TMS may be able to noninvasively influence deeper brain regions (Fox et al. 2014). Although single TMS pulses do not seem to have any durable effects, repetitive TMS has been shown to induce longer-lasting changes in brain regions and corresponding networks.

The exact effects of repetitive TMS are less well understood. Some evidence suggests that rTMS is capable of inducing long-term potentiation and long-term depression-like effects in targeted brain regions or networks, although this is admittedly a simplified and inexact description of the underlying processes, which are more complex (Cirillo et al. 2017). The general nomenclature suggests that “low frequency rTMS” (defined as pulses delivered at a frequency of 1 Hz or less) are inhibitory in the underlying neurons, whereas “high frequency rTMS” (defined as pulses delivered at a frequency greater than 1 Hz and usually greater than 5 Hz) is excitatory. Most of the evidence for these hypotheses comes from work done with repetitive stimulation of the primary motor cortex (motor-evoked potentials). If a high-frequency rTMS paradigm reduced the amount of energy required to produce a motor-evoked potential, that paradigm was thought to be excitatory and vice versa for low-frequency rTMS (Chen et al. 1997; Wu et al. 2000).

Research now demonstrates this conceptualization is an oversimplification; many factors will determine whether a stimulation protocol is excitatory or inhibitory, or has any effect at all, including the following:

• 1. Duration since the previous pulse sequence (Julkunen et al. 2012)

• 2. Ongoing tasks or cognitive processes occurring during stimulation time (Suzuki et al. 2014)

• 3. Orientation of the magnetic coil in reference to the neuronal elements being stimulated (Thomson et al. 2013)

• 4. Orientation of the neuronal elements themselves (e.g., operating in parallel or perpendicular to the induced current, often determined by the location of the gyri and sulci)

• 5. Any underlying pathologic process in the brain

• 6. Medication changes, sleep changes, caffeine or alcohol intake changes

• 7. Likely numerous other factors yet to be identified (Fitzgerald and Daskalakis 2013)

Several of these factors can also play a role in the size and quality of the motor-evoked potential obtained with stimulation of the primary cortex (see “motor threshold” below).

In addition, rTMS has been associated with brain changes identified using various research modalities, including changes in motor cortical excitability and plasticity, EEG signaling, PET imaging, functional connectivity MRI, MR spectroscopy, SPECT imaging, brain-derived neurotrophic factor (BDNF), genetic transcription factor activity, dopamine levels or dopamine-binding activity, and functional near-infrared spectroscopy, among others (Fitzgerald and Daskalakis 2013). Details of the myriad identified changes are beyond the scope of this chapter but strongly suggest that rTMS is having a significant impact on brain neurochemistry and neurophysiology that lasts well beyond the duration of the stimulation.

3.4 Clinical Indications and Clinical Utility

Although it is being studied for almost every neuropsychiatric illness, there are only a few conditions for which TMS and rTMS are currently FDA-approved for clinical use in the United States. These are as follows (For a summary of clinical and experimental indications for rTMS, see Table 2):

• 1. The use of high-frequency rTMS of the left prefrontal cortex for major depressive disorder that has failed to respond to at least one medication

• 2. The use of TMS for motor and speech mapping for presurgical planning in neurosurgery, often for epilepsy or brain tumor resections

• 3. The use of single-pulse TMS to the occiput for the abortion of migraine headache with aura

The use of rTMS for major depressive disorder (MDD) is the focus of this chapter, after which we will briefly discuss its experimental use for other neuropsychiatric indications.

The use of repetitive transcranial magnetic stimulation for unipolar major depressive disorder has been an FDA-approved indication for rTMS since 2007. The first device to be approved for rTMS in depression was the Neuronetics device. Neuronetics sponsored the pivotal trial that demonstrated efficacy of the rTMS treatment (compared to sham treatment) in major depressive disorder. Since then, several other rTMS devices have been approved. Notably, although the FDA’s medical device approval was for a specific indication, the device can be used for off-label purposes, as deemed appropriate by the physician. Thus, although a specific protocol was utilized in the pivotal Neuronetics trial, the FDA approval indicated use of high-frequency rTMS to the left prefrontal cortex (no specific parameters required).

In general, “high-frequency” TMS refers to any frequency >1 Hz, although most studied high-frequency protocols utilize 5 Hz or greater. The frequency used in the Neuronetics trial was 10 Hz (10 pulses/s). In addition to frequency, there are several other parameters of the TMS stimulus which can be adjusted to alter the treatment and its effects on the brain. These include changing the percentage of machine output (total stimulus charge), the length of each “train” of pulses, the duration of time between pulse trains (the “inter-train interval”), and the total number of trains and pulses delivered. Although some research suggests that longer courses of treatment with more total pulses may lead to additional benefit for patients with depression (Perera et al. 2016), the optimal parameters are still being studied.

3.5 Dosing of rTMS

The current method used to “dose” rTMS is based on the patient’s “motor threshold.” By delivering a single TMS pulse over the primary cortex, the stimulation intensity required to induce a muscle twitch in the hand can be determined. The exact “dose” of stimulation is variable and is determined by several factors. The “motor threshold” is defined as the minimum stimulus intensity at which a muscle twitch is elicited in the contralateral hand muscles in at least five of ten successive single-pulse stimulation trials. The treatment stimulation dose is then set based on a percentage above or below the motor threshold. For example, in the Neuronetics trial, treatments were delivered over the prefrontal cortex at a stimulus intensity 120% of the patient’s individual motor threshold.

3.5.1 Treatment Parameters

For the Neuronetics trial, which led to the FDA approval of rTMS for depression, treatments were delivered at 120% of the motor threshold, over the left dorsolateral prefrontal cortex, which was defined as a target 5 cm anterior to the motor threshold target of the primary motor cortex. A frequency of 10 Hz was used, and pulses were clustered into 4 s trains (10 pulses/s × 4 s = 40 pulses/train), for a total of 3,000 pulses per session. There was a 26 s break in between each “train” of pulses. Sessions were performed daily, 5 days/week, for 4–6 weeks (Perera et al. 2016).

Significant changes in a standardized depression scale (Hamilton Depression Rating Scale − 24 items) were identified at the 4-week and 6-week time points for a subcategory of patients who had failed one antidepressant medication, leading to the initial FDA approval. As several additional subsequent confirmatory studies have emerged, the FDA approval has been expanded to allow for treatment of patients with major depressive disorder who have failed at least one antidepressant medication of adequate dose and duration in the current depressive episode. Although some physicians may adjust the parameters of the treatments, the vast majority of rTMS practitioners in MDD use the parameters employed in the Neuronetics trial.

3.6 Clinical Indications in MDD

When is it clinically appropriate/therapeutic to use rTMS for MDD? The answer may vary from practitioner to practitioner depending on several factors, including access to an rTMS device/provider, as well as access to other treatment options (e.g., ECT). Generally speaking, rTMS is often considered for patients who have failed several medication trials or who cannot tolerate side effects of medications. It is often considered prior to a trial of ECT for patients with treatment-resistant MDD. Many times patients who suffer from treatment-resistant depression (TRD) will come seeking rTMS due to fears of undergoing a course of ECT or having previously experienced intolerable side effects to ECT. Finally, rTMS is often considered in severely medically ill patients (if there are concerns over medical tolerability of ECT). For a summary of clinical and experimental indications for rTMS, please see Table 2.

Table 2 Current and experimental indications for transcranial magnetic stimulation therapy

3.7 ECT vs. rTMS

Although studies in the literature are mixed about whether rTMS is as efficacious as ECT, the general consensus among experts is that ECT is a more effective treatment for treatment-refractory depression as of the time of this writing and should still be considered as an option even if a patient fails a course of rTMS (Fitzgerald and Daskalakis 2013). If a patient is suffering from bipolar depression, bipolar mania, or any type of psychotic illness, ECT or pharmacotherapy would be the treatment of choice, as there is insufficient evidence to suggest that rTMS is an effective therapy in these populations.

3.8 rTMS in Other Clinical Populations

TMS is being explored as a therapy for specific populations of depressed patients, as well as for other neuropsychiatric indications. Ongoing research looking at subpopulations of depressed patients, such as the use of rTMS in child and adolescent depression, geriatric depression, depression secondary to traumatic brain injury, bipolar depression, and postpartum depression, is ongoing and promising (Fitzgerald and Daskalakis 2013).

In terms of other neuropsychiatric indications, rTMS of various brain regions, using various treatment parameters, is being studied for obsessive compulsive disorder, Tourette’s and tic disorders, post-traumatic stress disorder, cravings associated with substance use disorders, auditory hallucinations in schizophrenia, negative symptoms of schizophrenia, cognitive enhancement in Alzheimer’s dementia, poststroke recovery, epilepsy, Parkinson’s disease, autism, tinnitus, ADHD, chronic pain, headache, and other disorders (Wassermann and Lisanby 2001; Kobayashi and Pascual-Leone 2003).

3.9 Practical Considerations and Patient Selection

rTMS is considered a safe and well-tolerated procedure, which is a significant factor leading to its popularity as a novel therapy for depression. Unlike ECT, rTMS requires no general anesthesia or induction of a seizure. It can be performed on an outpatient basis, and patients have no restrictions on diet or activities before or after treatment. Nonetheless, there are some practical considerations the rTMS specialist must address when consulting on a patient considered for rTMS therapy.

3.9.1 Patient Evaluation

Usually a patient will be referred to an rTMS practitioner for a psychiatric diagnostic evaluation to consider rTMS treatment. This evaluation serves two primary purposes – to confirm the MDD diagnosis and to ensure the patient is appropriately medically stable. In addition to the confirmation of MDD, it is important to note any comorbid psychiatric illnesses, which can significantly contribute to likelihood of response. For example, comorbid anxiety has been identified as a poor predictor of rTMS response (Lisanby et al. 2009). In contrast, more recent work has suggested rTMS that targets the dorsolateral prefrontal cortex may address psychiatric symptoms that span several diagnostic categories, perhaps by addressing underlying transdiagnostic problems influenced by cognitive control networks in the brain (Taylor et al. 2014).

In terms of medical safety, a practitioner’s main concern should be evaluating for risk of inducing a generalized seizure, the most dangerous risk of rTMS application. Hence, medical conditions/circumstances that increase the patient’s risk for a seizure need to be determined. Along these lines, the practitioner should inquire about any history of head injuries, concussions, neurological surgery, epilepsy, or seizures. As rTMS induces a brief magnetic field in the area around the coil, the practitioner also needs to query about implanted devices near the coil, including devices such as deep brain stimulators, vagal nerve stimulators, or cochlear implants. Further, patients should be asked about other implanted devices within the body, specifically those which are ferromagnetic. Sometimes treatment parameters can be adjusted to accommodate such patients, but these adjustments are patient-specific and will not be discussed here.

As a practical matter, rTMS is a time- and resource-intensive process. Patients need to be able to come in for treatments 5 days/week for 4–6 weeks. This requires reliable transportation to and from the hospital and a flexible daily schedule that allows the patient access to treatment. This logistical matter is very important to discuss in advance with new patients considering rTMS therapy, so that commitment expectations are clear.

3.9.2 Pre-treatment Work-Up

There is little recommended or required in the way of blood tests, imaging, or diagnostic tests prior to pursuing rTMS therapy for a healthy individual. If a patient has a history of head trauma or has had a surgical procedure on the head or neck, obtaining previously performed imaging or ordering new imaging might be worthwhile. The goal of reviewing this imaging would be to ensure there is no ferromagnetic material in the scalp or head and to ensure that there is no evidence of visible neurologic damage that could place the patient at higher risk for a seizure. Other work-up should be obtained only as clinically indicated for special circumstances – for example, a urine drug screen in a patient with a history of substance abuse if there is concern for active use, as use of certain substances, may alter seizure threshold. Obtaining metabolic blood samples in a patient with chronic medical problems such as hypernatremia may also help to evaluate for seizure risk.

If a patient has a history of seizures, inquiring further about the nature of these seizures and ensuring a proper work-up has been conducted by a neurologist is important. If the seizure was provoked by fever, medications, or illicit drugs, and there is no concern for continued seizure risk, it may not result in an absolute contraindication for treatment; again, this is something that must be evaluated on a case-by-case basis, and discussion with the patient and other physicians involved in the patient’s care is important for adequate determination of risks and benefits of rTMS therapy.

3.9.3 Medication Adjustments

Some practitioners may consider making medication adjustments based on the premise that certain medications may inhibit neuroplasticity (thus decreasing potential response to rTMS therapy), while others may lower seizure threshold (placing the patient at higher risk with treatment). The general practice is to continue psychiatric medication management with rTMS therapy augmentation. Although some medications (e.g., lamotrigine (Manganotti et al. 1999)) have clearly been shown to alter cortical excitability, and others (e.g., bupropion) are known to alter/lower seizure threshold, there are no clinical guidelines at this time to suggest that removing or adjusting doses of certain medications is necessary to achieve an rTMS response. Altering medication doses is often left to the best judgment of the rTMS practitioner in collaboration with the patient’s primary psychiatrist. Indeed, ensuring a patient is not in active withdrawal (e.g., from benzodiazepines or antiepileptic mood stabilizers) during the administration of rTMS, therapy is important to minimize seizure risk. In general, avoidance of drastic medication changes prior to adding rTMS may be the safest course.

3.9.4 The Treatment Course

rTMS is usually administered by an rTMS-certified technician. This can be the physician with special rTMS training, but oftentimes it will be a nurse practitioner, physician’s assistant, nurse, behavioral health expert, or a trained technician. It is usually recommended that the technician have basic life support training and rTMS training with a focus on how to be a first responder in the case of seizure or behavioral emergency (Perera et al. 2016).

The rTMS course usually lasts 4–6 weeks. Some research suggests that patients will continue to respond with longer courses even if they do not respond in the first 4–6 weeks (Perera et al. 2016); however, practically speaking, this is difficult unless the patient is motivated and has an insurance environment that will continue to reimburse for the continued provision of treatments (typically infrequent in the United States beyond 6 weeks). If a patient fails to respond to rTMS therapy, referral to ECT is commonly a consideration. Other patients will return to their primary provider to consider other pharmacotherapeutic options.

As with most therapy for major depression, a higher level of treatment resistance is a poor prognostic factor for rTMS response (Lisanby et al. 2009). However, several studies have now demonstrated that even in patients who have failed several medications, TMS can be a viable and effective alternative (Fitzgerald and Daskalakis 2013; Perera et al. 2016). Although some rTMS practitioners may try to augment or alter the treatment protocol to address issues of treatment resistance or nonresponse, the evidence for doing such is mixed (Blumberger et al. 2016).

Durability of rTMS is a subject of active research. There are no well-accepted practice guidelines for rTMS discontinuation/taper in MDD treatment. There is also no clear indication regarding whether or how to offer maintenance rTMS therapies after a successful rTMS course. Some data suggests rTMS may be more durable than ECT; however, a large proportion of patients will still relapse within 6–12 months of discontinuing a course (Perera et al. 2016). The current common practice is to taper the patient off rTMS (a common taper used in studies is the 3-2-1 taper – three times per week for a week, two times per week for a week, one time per week for a week, and then stop) and then discuss pharmacologic maintenance therapies. Although one study suggested that monthly rTMS maintenance could offer a mild benefit over no treatment at all (Philip et al. 2016), the current consensus is that a more aggressive frequency maintenance strategy would likely be required to keep a higher percentage of patients from relapsing. At this time, rTMS continuation or maintenance is not common practice.

3.10 Adverse Effects of rTMS

As mentioned previously, the most serious adverse effect associated with rTMS treatment is seizure induction. All seizures induced with rTMS therapy have been self-limited. There are no cases of patients developing epilepsy, or any type of seizure disorder, after a course of rTMS. The risk of seizure is minimal, estimated to be approximately 1 in 30,000 treatments (Perera et al. 2016). Usually with appropriate precautions, this risk can be minimized and is negligible. There are several ways to evaluate and minimize seizure risk:

• 1. Ensure the patient is taking their medication reliably and regularly – A patient who misses a dose of an antiepileptic or benzodiazepine that was not appropriately tapered could be at higher risk for a seizure. Be aware of any medication changes that are made during the course of the treatment.

• 2. Evaluate for any medications or illicit drug use that may lower seizure threshold, either in the intoxicated state (stimulants) or in the withdrawal state (barbiturates, alcohol, benzodiazepines).

• 3. Ensure your patient is attempting to get an adequate amount of sleep and is maintaining good sleep hygiene, as sleep deprivation will lower seizure threshold.

• 4. Be sure to stay within the safety guidelines for rTMS stimulation parameters. Seizures are shown to occur more commonly with primary motor cortex stimulation and high-frequency stimulation protocols.

Other potential-related adverse effects are syncope, dizziness, scalp irritation, headaches, muscle twitching, and scalp burns (if the coil malfunctions and overheats).

There are strategies for minimizing or mitigating these side effects:

• 1. Slowly increase a patient’s dose over the course of a single treatment or several treatments to the 120% motor threshold dose. Sometimes side effects can be minimized or eliminated by increasing the dose rather than subjecting patients to the full dose at the start of session one.

• 2. Keep the treatment coil in proper working order, and maintain a good relationship with the device manufacturer so any problems can be addressed expeditiously.

• 3. Be aware of warning signs of syncope or seizure and terminate or pause a session early if needed to ensure safety.

  • (a) Warning signs for syncope – Dizziness, diaphoresis, lightheadedness, increased heart rate, tunneling vision.

  • (b) Warning signs for seizure – Purposeless or stereotyped movements, unresponsiveness to commands or questions, contralateral motor movements during the prefrontal stimulation trains (especially if the movements continue beyond the end of the train; if motor movement is occurring during a prefrontal stimulation, strongly consider repositioning the coil or troubleshooting to ensure your target location is correct, as this could be a harbinger of impending seizure).

• 4. Recommend over-the-counter analgesic medications such as acetaminophen or ibuprofen to manage headaches or scalp irritation. Patients can be reassured that the irritation rarely lasts longer than a week of treatment – Studies replicate that scalp pain rapidly dissipates with repeated stimulation.

3.11 Future Directions

The use of transcranial magnetic stimulation in its various forms for clinical and experimental neuropsychiatric applications is a rapidly evolving and exciting field. Briefly we will discuss some of the areas of active and promising research:

3.11.1 Targeting Treatments

Considerable recent research has focused on finding better ways to target rTMS treatments to specific regions in the brain. Research groups are now looking at using personalized head measurements or neuroimaging to target treatments more precisely and reliably.

3.11.2 State-Dependent Treatment

The notion that brain state plays an important role in the efficacy of treatment is gaining ground. Research is ongoing in linking EEG readings with TMS stimulation to allow for parameter adjustments based on a patient’s personal EEG output. Other ongoing work is combining rTMS with psychotherapy (Donse et al. 2018), mindfulness, and other activities that might place the brain into a “state” that is more receptive to beneficial neuroplasticity changes.

3.11.3 Maximizing Neuroplasticity

The concept of neuroplasticity, or the brain’s ability to adapt and change its function and structure over time, is an exciting discovery. Researchers are trying to find ways to enhance neuroplasticity using novel rTMS protocols and other augmenting agents. This includes the use of specialized rTMS protocols, such as “priming” rTMS or “theta burst” rTMS, to more effectively and efficiently enhance the neuroplasticity of the brain.

3.11.4 Coil Variation

The electrical field induced by an rTMS pulse is dependent on the shape and physical properties of the magnetic coil being used. Coils are being developed that allow for varying levels of brain coverage and depth of penetration. In 2013 a “deep TMS” device was approved for use in major depressive disorder, which provides much deeper and broader brain stimulation, allowing modulation of deeper brain structures (Perera et al. 2016). This device has shown similar efficacy in achieving response and remission for major depressive disorder when compared to more superficially stimulating coils. Newer coils are being developed with the intent to target much smaller cortical areas or to simultaneously stimulate multiple cortical targets.

3.11.5 Parameter Space

As one can imagine, a dizzying amount of parameter changes can be made, such as altering frequency, number of pulses, number of treatment sessions per day, time between treatment sessions or pulse trains. Studies looking at “high-dose” rTMS protocols or unique treatment targets are underway for almost every neuropsychiatric known psychiatric condition.

3.12 Conclusion

Repetitive transcranial magnetic stimulation is an exciting new technology that affords physicians and researchers a new method for modulating and probing the human brain. It offers several practical advantages over the more cumbersome and invasive ECT treatment in the management of MDD, although currently rTMS does not rival ECT in antidepressant effectiveness, especially in highly refractory MDD. As more research is conducted, and we learn better ways to optimize this versatile treatment, it is likely that treatment success rates will improve, methods to sustain the effects will emerge, and new neuropsychiatric indications will arise.

4 Vagal Nerve Stimulation

4.1 Background

The concept that stimulation of the left cervical vagus could lead to an antidepressant response emerged from the epilepsy literature. Vagus nerve stimulation (VNS) has demonstrated efficacy in the treatment of certain forms of refractory epilepsy (Morris and Mueller 1999; Food and Drug Administration 2005); during these trials, it was anecdotally reported that epileptic patients suffering from comorbid depression experienced depressive symptom improvement. Subsequently, two studies (Elger et al. 2000; Harden et al. 2000) assessed for the reduction in depressive symptomology in refractory epileptic patients receiving VNS, and both studies found a trend toward depression reduction in those patients receiving VNS, independent of reduction in seizure frequency. This exciting finding prompted the study of VNS in treatment-refractory major depression (TRD) without concomitant epilepsy.

In 2005, in response to two large clinical trials that demonstrated antidepressant efficacy (Rush et al. 2000, 2005a, b), the US Food and Drug Administration (US-FDA) approved the use of adjunctive VNS for the treatment of major depressive disorder in patients not responding to four or more antidepressant courses. Despite this FDA approval, in 2007 the United States Committee on Medicare and Medicaid Services (CMS) elected not to reimburse VNS for TRD, contending that the treatment was experimental/unproven. Since that time, most private insurers have fallen in line with this decision, leading to very limited availability of VNS to TRD patients. Currently, there are ongoing efforts to further study/understand how VNS works in TRD with the eventual goal of making it available to individuals with severe, refractory depression.

4.2 Surgery and Stimulus Delivery

The primary means of VNS delivery in the United States and Europe is via the Neurocybernetic Prosthesis™ system (NCP; Cyberonics, Houston, TX; Fig. 3). The VNS device is an implantable, multi-programmable, battery-operated current generator, which is typically implanted under the skin below the clavicle (typically the surgeon will access this area via the axilla). The lead from the device is then run under the skin into the neck region, where a second incision is made to expose the left cervical vagus. The bipolar lead is typically attached above the cardiac branch of the vagus.

Fig. 3

The implanted VNS generator device (Neurocybernetic Prosthesis™ system (NCP; Cyberonics, Houston, TX). The VNS generator is subcutaneously implanted beneath the left clavicle, with the emerging lead subcutaneously tunneled in the skin over the clavicle. A second incision is made in the mid-neck region to expose the vagus nerve. The leads (see enlargement) are then attached at three points, a cathode, anode, and ground. Source: Reproduced with permission of LivaNova

Once turned on, the device delivers around-the-clock vagal stimulation. Table 3 describes the modifiable electrical stimulation parameters used in VNS and the standard ranges allowable for these parameters.

Table 3 Modifiable electrical parameters of VNS

Animal studies, and experience in human studies, demonstrate that at current clinical levels of stimulus delivery (0.25–3.5 mA), the vast majority of the electrical stimulus is directed afferently (toward the brain). Hence, the thoracic (lung and heart) as well as the abdominal organs (gastrointestinal tract) supplied by the vagus are minimally affected by VNS. The more proximal/afferently located recurrent laryngeal nerve (supplies the larynx) does frequently receive afferent stimulus; hence, approximately two thirds of patients experience hoarseness/stridor during VNS stimulation for TRMD.

4.3 Clinical Studies of TRMD

To date, there have been six clinical trials to test the antidepressant efficacy of VNS in TRMD summarized in Table 4 (Rush et al. 2000, 2005a; Marangell et al. 2002; Nahas et al. 2005; Schlaepfer et al. 2008a; Bajbouj et al. 2010; Aaronson et al. 2013, 2017). Only one of these (Rush et al. 2005b) had a double-blind, placebo arm (all subjects implanted, only 50% had devices turned on). The most recently published trial, described below, was an FDA-mandated, open-label registry of patients with TRMD (Aaronson et al. 2017) that spanned a 5-year period. With exception of the 2013 dose-finding study (Aaronson et al. 2013), all of the studies had similar initial dosing patterns.

Table 4 Summary of clinical trials of VNS in adults with treatment-resistant depression

The single double-blind, placebo-controlled study came close to but did not achieve statistical significance after 10 weeks of stimulation; however, importantly, it has since been determined that the majority of VNS patients require sustained VNS, typically in the range of 6–12 months, before achieving maximal antidepressant response.

Notably, all VNS TRMD studies to date demonstrate antidepressant efficacy with response rates ranging from 30 to 53% (see Table 4). A vexing problem of TRMD treatment is maintaining efficacy. Studies assessing existing treatments (including electroconvulsive therapy [ECT]) demonstrate that the typical 1-year response rate in TRMD is abysmal (~10% (Dunner et al. 2006)). In contrast, studies of VNS in TRMD demonstrate that the majority of patients achieving antidepressant response maintain this response at 1 (Marangell et al. 2002) and 2 years (Nahas et al. 2005); hence, there is emerging evidence that VNS has sustained efficacy in TRMD. In fact, a recent published case series described six TRMD patients with 15–33 years of preimplantation depression, who have maintained antidepressant remission post-VNS treatment for a mean of 9.2 years (Salloum et al. 2017).

As part of an FDA-mandated registry, Aaronson et al. (2017) followed a collection of 795 TRMD patients in the largest and longest study comparing VNS against treatment as usual (TAU). The TAU cohort was able to receive any treatment for the study duration (n = 301) and were compared to TAU plus adjunctive VNS (VNS + TAU, n = 494). After 5 years of follow-up, VNS + TAU had higher cumulative response rates (67.6% vs. 40.9%, p < 0.001) and remission rates (cumulative first-time remitters, 43.3% vs. 25.7%, p < 0.001). Further analyses demonstrated VNS + TAU had a more rapid response to treatment and had a slower relapse rate. Additionally, when comparing antidepressant response rates between patients who had not responded to ECT across both cohorts, VNS + TAU demonstrated greater antidepressant efficacy (59.6% vs. 34.1%, p < 0.001), demonstrating that failing a course of ECT does not necessarily predict failing VNS.

Aaronson et al. (2013) attempted to compare efficacy of VNS in TRMD using a “dosing study” of three different electrical parameters: a “low,” “medium,” and “high.” The doses differed in pulse width and current; however, the groups held constant the duty cycles (30 s “on,” 5 min “off”) and pulse frequencies (20 Hz). This trial had an acute phase (first 22 weeks) and a long-term phase (subsequent 28 weeks). The acute phase had “fixed” (unchangeable) parameters; however, the long-term phase allowed for upward dose titration. During the acute phase of the trial, the higher current and pulse width groups (“medium” and “high”) demonstrated numerically higher response rates (than the low-dose group) at 22 weeks, although these groups did not achieve statistical significance. However, at the end of the long-term phase, the “medium-” and “high”-dose cohorts were less likely to have a depressive relapse, suggesting that a higher VNS treatment initiation dose may help maintain antidepressant response.

4.4 Surgical Procedure

In general, the implantation procedure is done on an outpatient basis. The entire procedure (including VNS therapy generator implantation and attachment of the lead to left vagus nerve) takes about 1.5–2 h. The procedure is typically very well-tolerated and has low rates of complication (~1% infection rate).

4.5 Delivery of VNS in TRMD

Typically, due to postoperative swelling and pain, it is advisable to allow a period of 2 weeks of postoperative surgical recovery before initiation of stimulation. During each of the “titration visits,” the electrical current is gradually increased (typically over two to three visits separated by 5–7 days).

In light of the research that supports that higher electrical current likely provides better sustained antidepressant effects (Aaronson et al. 2013), the current standard practice (subject to change with further research findings) is to attempt to push the treatment to the patient’s “highest tolerable” current level. In general, our experience is that TRMD patients may not tolerate aggressive ramping up of the electrical current (less so than refractory epileptic VNS patients). Additionally, there is considerable between-patients tolerability of VNS: some patients develop discomfort at very low levels of current (e.g., 0.5 mA), while others tolerate very high currents (2.5 mA). In general, we recommend attempting to achieve an initial treatment current in the 1.5–2.0 mA range. Further, experience suggests that TRMD patients slowly acclimate to the upward titration of current, so the “start low, go slow” process is advisable to insure that patients achieve higher current. In general, this is best achieved over two to three visits during which the current increases are 0.25 mA/current increase, allowing the patient to experience several “firing cycles” at this dose before titrating further upward.

Additionally, experience has also demonstrated that certain stimulation parameters are more frequently associated with pain or discomfort. In particular, frequencies above 20 Hz and pulse widths greater than 250 μs are avoided during initial titrations, as we have found these to be more frequently associated with patient discomfort.

In summary, our experience, and that of many users of VNS in TRMD, is to start with low frequency (20 Hz), low pulse width (250 μs), and a “standard” duty cycle of 30 s “on” and 5 min “off.” We typically use the first two to three office visits to titrate up to a tolerable dose with a period of observation of 15–20 min between upward output current titrations.

4.6 Programming the VNS Device

Similar to the method employed with programming a cardiac pacemaker, a handheld programming computer is attached to a “wand” that allows the programmer to check device integrity and modify electrical parameters (Fig. 4). The modifiable electrical parameters involved in VNS include output current (milliamps, mA), current frequency (Hertz, Hz), pulse width (microseconds, μs), and duty cycle (time “on” vs. time “off”). For titrating VNS in TRD, the clinician must first take into account patient comfort.

Fig. 4

The VNS device programming wand. This instrument is held against the skin directly over the implanted VNS generator. The wand is attached to a handheld computer, similar to a smartphone, which allows for modification of electrical parameters changes and is used for assessments of circuit integrity (successful transmission of current to the vagus nerve), as well as programming the electrical parameters being delivered during VNS. Source: Reproduced with permission of LivaNova

4.6.1 Example Titration

Based on our experience with VNS in TRMD, we present the following example titration:

Office Visit #1: Stimulation is initiated with the following parameters:

Frequency::

20 Hz

Pulse width::

250 μs

Duty cycle::

30 s “on” and 5 min “off”

Output current::

0.25 mA

Starting with an output current of 0.25 mA, have the patient sit in the waiting room for 20–25 min to allow the device to cycle three to four times to assess tolerability. If the patient tolerates these settings without pain/discomfort/side effects, increase the output current by another 0.25 mA (to 0.50 mA), followed by another 20–25 min observation. This is repeated a third time on the first office visit with a final first visit output current (assuming patient tolerability) of 0.75 mA. If at any time during the upward titration the patient experiences pain/discomfort/side effects, we decrease the output current by 0.25 mA to the previously tolerated level. We then have the patient return in 1 week and reattempt to increase the output current by at least 0.25 mA.

Office Visit #2:

Frequency::

20 Hz

Pulse width::

250 μs

Duty cycle::

30 s “on” and 5 min “off”

Output current::

0.75 mA

Similar to Visit #1, we increase the output current by 0.25–1.0 mA and observe the patient while the device cycles, three to four times; during this process, we ask the patient if they are experiencing any pain/discomfort/side effects. This is repeated one to two more times during this visit. Some clinicians will be more aggressive with their upward titration (increasing by current output increments >0.25 mA); however, we have observed that a more gradual titration allows for both greater final output currents and greater patient comfort.

Once you have achieved the patient’s maximal tolerable current, we recommend holding this dose for a sustained period, typically 9–12 months.

4.6.2 When Should You Make Further Parameter Adjustments?

Evidence from clinical and neuroimaging studies (Rush et al. 2005b; Nahas et al. 2007; Conway et al. 2013) strongly suggest that response to VNS in TRD typically develops longer term. Response rates (i.e., a 50% drop in standard MDD measures) appear to increase most precipitously at 6–12 months. For this reason, we believe that once the maximally tolerated output current (during original titration) is achieved, it is wise to maintain these parameters for at least 12 months. If the patient is having partial or no response to treatment, our experience suggests that increasing the amount of charge delivered over time has the greatest influence on antidepressant outcome. This can be achieved by either increasing the amount of “on” time or decreasing the amount of “off” time between charge deliveries. The VNS therapy user guide, which accompanies the VNS programming system, details the allowable percentage on time/duty cycle. A duty cycle in excess of 50% “on” time equal to or greater than “off” time is not recommended. It should be noted that increasing the amount of charge delivered in a given time span will also more rapidly decrease battery life; therefore, this step should be reserved for situations in which standard parameter settings have not proven successful.

4.6.3 What Constitutes a VNS Antidepressant Response in TRMD?

Several key points about VNS response in TRD:

• 1. Though there are some TRD patients who respond quickly to VNS, clinical studies and brain imaging studies suggest that the majority of TRD patients take 6–12 months of VNS before exhibiting a response.

• 2. From experience, the antidepressant response tends to be very subtle at first, and then increases with time, typically over many weeks.

• 3. If a TRMD patient does not have a response after 8–10 months of stimulation at his or her “maximal tolerated dose,” we recommend increasing the charge/time delivered using the method described above.

• 4. Unpublished data (as of this writing) suggests that many patients who fail to fall below the 50% drop in standard antidepressant rating scales (classic antidepressant response and remission) still report significant improvements in quality of life which is not directly captured on these scales. In general, with careful patient selection, our experience suggests that approximately two out of three patients experience a clinically meaningful improvement in quality of life, such that they do not desire to have the device removed/turned off.

5 Other Evolving Neurostimulation Treatments

5.1 Deep Brain Stimulation (DBS) for Treatment-Resistant Major Depression

Deep brain stimulation involves the placement of a lead into specific, targeted brain regions. The use of DBS for TRMD emerged from the highly successful use of DBS in movement disorders. More recently, DBS has been successfully employed in obsessive compulsive disorder (OCD).

Currently, there are studies/reports showing the successful use of DBS on six different brain regions: the nucleus accumbens (NAcc), ventral capsule/ventral striatum (VC/VS), Brodmann area 25 or subgenual cingulate cortex (SCC), lateral habenula, inferior thalamic peduncle, and medial forebrain bundle.

Most of the DBS studies published to date do not employ placebo/sham groups; to date only four (Mayberg et al. 2005; Schlaepfer et al. 2008b; Holtzheimer et al. 2012) are controlled trials with sham stimulation periods.

Though research is proceeding in all the abovementioned regions, the three regions currently with the most active research include the NAcc, VC/VS, and SCC. Early open-label studies of the NAcc showed great promise, with early studies demonstrating antidepressant benefits early [6 months (Schlaepfer et al. 2008b)]. Larger sample studies from the same group over a longer period (Bewernick et al. 2010, 2012) demonstrated approximately 50% response rates.

Based on several studies demonstrating that targeting the VC/VS in OCD patients led to improvement in concomitant depressive symptoms (Greenberg et al. 2006, 2010; Goodman et al. 2010), researchers now study this region as a target for TRMD. The term VC/VS includes the NAcc and the ventral aspect of the anterior limb of the internal capsule (ALIC). In contrast to the NAcc DBS studies, the VC/VS trials used larger electrodes. Malone et al. (2009, 2010) reported on a series of 17 TRMD patients receiving VC/VS DBS from several sites. This group reported a 53% response rate at 12 months and a 71% response rate at the last follow-up (14–67 months, average stimulation duration 37.4 months). Of note, two patients receiving VC/VS DBS (one with bipolar disorder) reported symptoms of mania (Malone et al. 2009); whether this is a relevant consideration regarding future use of VC/VS DBS in bipolar disorder is yet to be determined. Dougherty et al. (2015) also conducted a study of VC/VS DBS in 30 TRD patients. Half of the TRD patients were randomized to active treatment for 16 weeks, and half to sham. There was no difference in response rate observed between active treatment and sham at study conclusion, though an open-label extension of the trial suggested some longer term response (20–27% response rate over 2 years extension). Finally, a recent multicenter, prospective trial of VC/VS DBS (sponsored by Medtronic) failed to show significant antidepressant improvement after 16 weeks of stimulation and was discontinued due to the design which predicted futility (Underwood 2013). Researchers in this subtype of VC/VS DBS remain hopeful that modifications of technique and patient selection may enhance outcomes in larger multicenter trials.

Another well-studied, DBS-targeted region is Brodmann’s area 25 or the SCC. Mayberg et al. (2005) reported that four of six patients responded to SCC stimulation at 6 months. Subsequently, larger studies with longer stimulation durations found similar significant response rates (Kennedy et al. 2011): mean response rates for years 1, 2, and 3 of 63%, 46%, and 75%, respectively (Kennedy et al. 2011). Similarly, another large sample (N = 21) open-label, longitudinal study demonstrated antidepressant efficacy, though it did not achieve statistical significance (29% response rate at 12 months; mean drop of 41% on depression scale, with 62% of patients having a greater than 40% improvement (Lozano et al. 2012)). In a single-blinded study of SCC DBS, Holtzheimer et al. (2012) reported a very high response rate in a population of TRMD patients (unipolar and bipolar). Remission and response rates increased with time: 18% remission and 41% response at 24 weeks (n = 17) up to 58% remission and 92% response at 2 years (n = 12). Notably, this study used higher current DBS (highest of the current SCC DBS studies), ranging from 6.0 to 10 mA. The bipolar TRMD patients responded with equal frequency as unipolar TRMD patients, and there were no reported cases of manic emergence. Despite these very impressive early studies, a large multicenter, prospective trial of SCC DBS for TRMD (the BROADEN study, sponsored by St. Jude Medical) had to be discontinued due to futility analysis at 6 months of stimulation (statistical probability of response determined to be less than 17.2%; letter from St. Jude Medical Clinical Study Management). Very active work on SCC DBS in TRMD remains, with researchers remaining hopeful that perhaps more precise localization of lead placement, as well as more precise patient selection, can lead to improved outcomes.

There are other regions targeted for DBS in TRMD, including the lateral habenula, inferior thalamic peduncle, and medial forebrain bundle, but the results of these are beyond the scope of this section. For a nice summary of DBS findings in TRMD, please see Morishita et al. (2014).

5.2 Magnetic Seizure Therapy (MST)

Magnetic seizure therapy involves the use of an electromagnetic transcranial magnetic stimulation device to purposefully induce a generalized seizure under anesthesia. This is utilized in a manner similar to electroconvulsive therapy (ECT), although is capable of inducing a seizure with a less intense and more focal electric field than ECT (Lee et al. 2016). This increased focality allows for MST to initiate a seizure with primarily superficial cortical stimulation of the brain. The premise for MST is based on the theory that by sparing deeper brain structures from passage of electrical current, as occurs with ECT, some of the negative sequelae, especially cognitive, of ECT can be avoided. Early studies seem to suggest that magnetic seizure therapy may have fewer cognitive side effects when compared to ECT, as demonstrated by a faster post-procedure reorientation time and better acute cognitive performance following treatments (Lisanby et al. 2003; Cretaz et al. 2015).

Despite theoretical promise, thus far studies have not reliably demonstrated that MST can achieve the same antidepressant efficacy in major depressive disorder as ECT. MST response and remission rates in MDD have been highly variable, ranging anywhere from 38 to 69% and 15 to 46%, respectively. One systematic review of MST therapy suggested that these levels of therapeutic efficacy were still significantly lower than those achieved in most studies of right unilateral ECT at six times seizure threshold, the standard treatment delivered in many ECT practice settings (Cretaz et al. 2015). Despite this, MST is in its infancy as a tool for treating mood disorders and continues to be refined and improved in the hope of developing a new technology with similar efficacy to ECT and potentially fewer side effects (Radman and Lisanby 2017).

5.3 Transcranial Electrical Stimulation (tES)

Transcranial electrical stimulation is an umbrella term that encompasses various forms of noninvasive, subconvulsive electrical stimulation applied to the human head, with the intent of altering cortical excitability to achieve therapeutic goals. As compared to ECT, which also involves application of an electrical stimulus, tES therapies do not require anesthesia due to low-intensity stimulation that is attributable to few side effects and minimal discomfort. The current is administered via conducting pads applied to the scalp. Unlike rTMS therapy, which delivers a relatively focal stimulation in a repetitive fashion, tES therapies have a broader area of stimulation and constant current delivery. Additionally, unlike rTMS, tES stimulation intensity is usually not strong enough to induce depolarization of neurons. The type of current delivered defines the specific tES nomenclature; for example, under the tES umbrella, there exist transcranial direct current stimulation (tDCS), transcranial alternating current stimulation (tACS), and transcranial random noise stimulation (tRNS), among others.

Due to the relative affordability, simplicity, tolerability, safety, and durability of these devices, they have been studied for numerous neuropsychiatric applications, with varying degrees of success. Indeed, outside the realm of medicine, many of these devices are currently being marketed and sold as nonclinical agents for improving video game skills, cognitive ability, sleep, and energy, frequently resting on the basis of questionable and unreplicated data.

Within the realm of medicine, to date the modality with the biggest body of research underneath the tES umbrella is tDCS, with studies primarily looking at its uses for cognitive enhancement, depression, poststroke rehabilitation, and pain (Fregni et al. 2006; Zimerman et al. 2012; Marlow et al. 2013; Tremblay et al. 2014; de Aguiar et al. 2015). Mechanistic studies have supported evidence that tDCS has some effect on cortical excitability, as measured by induced changes in the amplitude of evoked potentials in the underlying cortical regions [usually motor cortex (Ammann et al. 2017)]. Additionally, some studies suggest that the effects of tDCS may be neurotransmitter-mediated. For example, studies of neuropharmacologic agents that act on various neurotransmitter systems including NMDA, serotonin, and dopamine demonstrate the ability to enhance or suppress tDCS-mediated effects on cortical excitability (Liebetanz et al. 2002; Nitsche et al. 2004, 2009; Kuo et al. 2008).

In terms of tDCS’s practical application, the general notion is that a battery generated electrical current and conducting pads of positive (anode) and negative (cathode) charge are applied to the scalp to incorporate brain tissue into the electrical circuit. Thus, the standard practice is to apply a pad delivering anodal stimulation at one scalp location and a second pad delivering cathodal stimulation at a second scalp location. Early evidence suggested that the anodal stimulation site would increase cortical excitability in the underlying neural elements, and cathodal stimulation would decrease cortical excitability (Nitsche et al. 2003). This phenomenon occurred via alterations in the membrane potentials of underlying neurons, thus increasing or decreasing the likelihood that they would fire an action potential.

The primary brain regions tDCS targets in the treatment of MDD are typically the left dorsolateral prefrontal cortex (anodal stimulation site) and the right dorsolateral prefrontal cortex (often cathodal stimulation site), based upon data demonstrating mood effects during rTMS stimulation of these regions and their presumptive involvement in mood regulation networks in the brain. The most common side effects of treatment include scalp tenderness, headache, fatigue, and, in rare instances, scalp burns. There have been no serious adverse events causally related to tDCS stimulation, with the biggest risk being the induction of mania in some bipolar patients (Antal et al. 2017).

Although tDCS is remarkably safe and well-tolerated, many researchers have questioned whether such a low-intensity stimulus could be inducing meaningful changes in the brain (Horvath et al. 2015). To date, the studies of its use in depression have been mixed and difficult to interpret. Although some early studies indicated a potential benefit for treating depressive symptoms when anodal tDCS was applied daily to the left DLPFC for several weeks at a time (Boggio et al. 2008; Loo et al. 2012; Brunoni et al. 2013), this has more recently been contradicted by a large, multisite, randomized controlled trial which showed no benefit (Loo et al. 2018). Ultimately, tDCS faces many of the same challenges identified to optimize rTMS treatments for MDD, that is, an almost-infinite parameter space in which to work, complicated by underpowered, poorly designed, or unblinded trials with conflicting results.

5.4 Trigeminal Nerve Stimulation (TNS)

Trigeminal nerve stimulation is a novel neuromodulation treatment that uses mild electrical signals to stimulate branches of the trigeminal nerve, also known as cranial nerve V (CN V). Similar to the vagus nerve, CN V is thought to have effects on mood regulation (DeGiorgio et al. 2011). However, unlike CN X, CN V is superficially located and has three branches traversing beneath the skin of the face, allowing for transcutaneous stimulation and obviating the need for an implanted device (DeGiorgio et al. 2011). Additionally, unlike CN X, CN V contains no autonomic outflow fibers to pose any cardiac risks (DeGiorgio et al. 2011). After demonstrating some success in decreasing seizure frequency in adults with drug-resistant epilepsy, TNS was then applied to depression research (DeGiorgio et al. 2003, 2006, 2009).

In an 8-week, open-label, pilot trial, Cook et al. (2013) demonstrated that TNS resulted in significant improvements in depression and quality-of-life measures in 11 adults with treatment-resistant major depressive disorder. Another 8-week study of 12 adults with comorbid MDD and post-traumatic stress disorder (PTSD) demonstrated significant improvements in both depression and PTSD severity from TNS (Cook et al. 2016). Finally, a recent study of the antiepileptic effects of TNS in 50 epilepsy patients also found significant improvements in mood independent of antiepileptic response as a secondary outcome (DeGiorgio et al. 2013).

Currently, the only commercially available TNS device is NeuroSigma’s Monarch™ external trigeminal nerve stimulation (eTNS™) delivery system in the European Union and Canada, where it has been approved for both depression and epilepsy (NeuroSigma 2012). The eTNS™ system is also approved in the European Union for attention-deficit/hyperactivity disorder (ADHD) after one promising pilot study (McGough et al. 2015). Notably, patients are able to use the physician-prescribed device in the comfort of their own homes. The device is approximately the size of a smartphone, with a wire connecting to a patch that adheres to the patient’s forehead. A minimally invasive subcutaneous trigeminal nerve stimulation (sTNS™) system is also under development by the same company. Further research efforts are required for FDA approval of TNS modalities in the United States. Replication of aforementioned studies with double-blinded conditions, as well as investigation of the durability of antidepressant efficacy following an acute treatment course, is all needed at this time.