History
Interest in the effects of vagus nerve stimulation (VNS) on the central nervous system began in the early 20th century, with Bailey and Bremer demonstrating in 1938 that electrical stimulation of the vagus nerve of cats resulted in increased electrographical activity of cortical structures ( ). Dell and Olson discovered a suggestive link between the vagus nerve and affective disorders in 1951 in their own study of cats, demonstrating an evoked response in the amygdala in response to vagal stimulation ( ). As a clinical intervention, VNS was first investigated for its potential role in treating epilepsy. Beginning in the 1980s, began animal experiments that demonstrated VNS’s ability to both terminate acute seizure activity and prevent seizures chronically with continuous stimulation. The first human implanted VNS device occurred in 1988 by Penry and Dean in four patients with intractable partial seizures, of whom two obtained complete seizure control ( ). VNS for medication refractory epilepsy received approval in Europe in 1994 and in the United States in 1997 ( ).
These epilepsy trials suggested a potential role for VNS in the treatment of depression. Reports of improved mood and cognition in patients receiving VNS for epilepsy came out of the trials used to gain VNS’s initial United States Food and Drug Administration (FDA) approval ( ). The role of anticonvulsant medications, including carbamazepine, lamotrigine, and valproate, in the treatment of mood disorders further hinted at a common mechanism underlying treatments for epilepsy and depression. Perhaps, researchers suspected, VNS would provide the same shared benefit.
Initially, this hypothesis gained strength through undercontrolled clinical observations in patients whose mood improved after receiving VNS for epilepsy ( ; ). These observations developed into more rigorously designed interventions exploring mood in epilepsy patients who had received VNS—several of which supported VNS’s mood-improving effects ( ; ; ).
The first VNS device was implanted specifically for depression in 1998 at the Medical University of South Carolina as part of a multicenter pilot trial which further established VNS’s antidepressant effects ( ). On the basis of four clinical trials in patients who had failed two to six treatments for their unipolar or bipolar depression (considered to be treatment-resistant depression (TRD)) which included a feasibility trial ( ); a randomized, sham-controlled 3 month-clinical trial ( ); a long-term open-label extension of the 3-month randomized trial ( ); and a long-term observational study comparing subjects receiving standard-of-care treatments with subjects receiving VNS ( ), the United States Federal Drug Administration (FDA) granted approval for the use VNS in treatment-resistant depression in 2005 even though the 3-month randomized controlled trial did not show significance in the primary depression outcome measure. This unusual determination was largely due to the absence of proven effective alternatives for this treatment-resistant population.
Subsequently, in 2007, the United States Center for Medicare and Medicaid Services (CMS) decided against coverage of VNS for TRD. This decision was based on the failure of the initial randomized controlled trial (RCT) to demonstrate separation in response between the active VNS group and the sham VNS group during the first 12 weeks of treatment. Subsequent years of research would reveal that the length of stimulation time required for VNS to achieve clinical efficacy is far longer than this 12-week endpoint—closer, in fact, to 6 months to 1 year ( ; ). Nevertheless, the adoption of vagus nerve stimulation since its initial approval for depression has been impeded by this early trial as well as by a later dose-finding trial in which subjects were randomized to low-, medium-, or high-dose stimulation ( ). While this study built on the results of the earlier RCT and made the primary end point 6 months after implantation, the efficacy of even modest, low-dose stimulation hampered demonstration of significantly improved efficacy of higher doses, though each group showed remarkable improvement. Most private insurers have followed the CMS decision and also do not cover the significant costs associated with the VNS device and surgical implantation. Costs of the device and implantation have been estimated at about $28,000; however, further economic analysis has suggested that, 8 years after device implantation, savings associated with VNS for TRD may be in the range of $23,000–$41,000 ( ). Despite these potential savings, the noncoverage determination placed VNS largely out of reach of the typical patient suffering from treatment-resistant depression during the first decade of its FDA approval ( ).
Researchers have continued investigating VNS’s efficacy for TRD in intervening years. To date, some 5000 patients have been implanted with VNS devices for treatment of depression in various clinical trials ( ). Notable among this body of literature is the VNS for TRD registry, which was established by LivaNova, the manufacturer of the VNS device, as part of the FDA’s initial approval in 2005 ( ). This registry followed 500 patients who had received VNS for TRD and 300 who received treatment as usual for over 5 years. report on these patients suggested improvements in depression in those with VNS compared to those receiving treatment as usual. As a result of these findings, CMS reopened its noncoverage decision in 2018. Currently, VNS for TRD falls under the “coverage with evidence development” category, indicating CMS’s participation in funding a new randomized controlled trial to further development of the VNS for TRD’s evidence base ( ).
The VNS device and its mechanism of action
The vagus nerve comprises a complex web of efferent and afferent fibers involved in diverse bodily functions, including autonomic control, motor action, and various sensory modalities. VNS targets the afferent fibers of the left cervical vagus nerve, which carry visceral, somatic, and gustatory sensation to the brainstem ( ). While the precise mechanism of action of VNS in depression is not entirely understood, there are several anatomical and neurobiological theories that help elucidate its antidepressant effects.
The VNS device
VNS entails surgical placement of electrode coils around the left vagus nerve. The device itself includes the stimulating electrodes, a pulse generator which serves as a battery and provides current to the electrodes, and electrode extenders which connect the electrodes to the pulse generator ( ; ; ) ( Fig. 21.1 ).
The vagus nerve
The anatomy of the vagus nerve itself also facilitates central propagation of the stimulation signal. The vagus nerve primarily comprises afferent fibers with low stimulation threshold ( ). What efferent fibers are present (approximately 20% of the total nerve fibers) have a higher stimulation threshold than the afferents. The preponderance of afferent fibers combined with their lower stimulation threshold helps direct stimulated impulses primarily toward the brain and away from visceral targets. Nevertheless, VNS does have effects on downstream organs. For example, VNS can cause bradycardia—an effect that is mitigated by stimulating the left vagus nerve instead of the right ( ). The left vagus nerve supplies parasympathetic input to the sinoatrial (SA) node while the right vagus nerve supplies the atrioventricular (AV) node. Since the AV node sits within the cardiac conduction pathway, right vagus nerve stimulation would potentially cause intracardiac conduction abnormalities. Left vagus stimulation would avoid such disruption, since the SA node lies at the very beginning of the pathway ( ). VNS is also known to cause GI side effects including dyspepsia, putatively a result of parasympathetic overstimulation of the enteric system.
Stimulation parameters
VNS combines four distinct stimulation parameters—including current (mA), frequency (Hz), pulse width (s), and duty cycle (the repeating period (in seconds) of active stimulation followed by a period of no stimulation)—which are independently manipulated to obtain a target “dose.” After 2 weeks of postoperative healing, the device is activated and the dose adjusted, typically over a 2-week titration period. Starting at 0.25–0.75 mA, the current in gradually increased over this time, most often to between 1.0 and 2.0 mA, to ensure tolerability and to monitor for side effects. The typical frequency ranges between 20 and 30 Hz, with frequencies above 50 Hz potentially causing permanent damage of the vagus nerve ( ). A pulse width of 250 s is common, and the duty cycle starts at 30 s on and 300 s off, so 10% of the time it is sending current.
Neurobiological effects of VNS
The essential brainstem pathway involved with VNS’s antidepressant effect includes the tractus solitarius, which links vagal afferents from their CNS input in the medulla to the nucleus tractus solitarius. These fibers travel upstream to the pontine parabrachial nucleus, the raphe nuclei, and the locus coeruleus ( ). Stimulation of the raphe nucleus, which contains serotonergic cell bodies, and the locus coeruleus, which contains noradrenergic cell bodies, is thought to play an essential role in VNS’s antidepressant effects. Decussating afferent fibers ensure bilateral central nervous system effects despite unilateral peripheral vagal stimulation.
Higher level brain structures, including the hypothalamus, thalamus, nucleus accumbens, amygdala, and stria terminalis, are also implicated in VNS’s mechanism of antidepressant action ( ). These regions each receive fibers directly from the nucleus tractus solitarius, and are known to mediate mood and depression. Cortical regions of affective regulation, including the anterior insula, lateral prefrontal cortex, and infralimbic cortex, also receive vagal afferent projections ( ). Furthermore, the insula communicates with other cortical structures, including the ventrolateral and orbital prefrontal cortex, which are also considered important targets of downstream VNS effects ( ) ( Fig. 21.2 ).
Brain imaging studies such as , have used fMRI to identify multiple regions, including the bilateral orbitofrontal cortex, parietooccipital cortex, hypothalamus, the left temporal cortex, and amygdala, that were activated while patients received VNS. Importantly, these brain regions are all known to be associated with major depression. Other studies have investigated the impact of various stimulation parameters, including stimulation frequency and pulse width, on regional brain activation in the subacute ( ) and chronic ( ) stages of treatment.
The timecourse of VNS’s antidepressant effects have also been elucidated in brain imaging studies. and , for example, found alterations in brain response to VNS after 30 weeks of stimulation compared with the period immediately after initiation of stimulation ( ; ). These changes, including decreased cerebral metabolic glucose (CMRGlu) in the right dorsolateral prefrontal cortex, correlate with the timing of VNS’s antidepressant effect. also noted that increased CMRGlu in the left substantia nigra and ventral tegmental area at 12 months of stimulation correlated with VNS response, while those who did not respond to VNS had decreased CMRGlu in this region. As the primary dopaminergic brainstem region, the VTA’s activation here supports the hypothesis that VNS’s antidepressant effects are also mediated in part by dopamine.
In an effort to establish potential biological predictors of VNS response, performed prestimulation FDG-PET scans on patients who went on to receive a course of VNS. They found that the combination of low anterior insular cortex (AIC) CMRGlu and high orbitofrontal cortex (OFC) CMRGlu predicted VNS response after 12 months of treatment.
Animal models of VNS support a unique antidepressant mechanism of action for VNS compared to antidepressant medications. In several studies, sustained VNS has been shown to increase the base firing rate of noradrenergic neurons in the locus coeruleus while suppressing inhibitory GABAergic interneurons ( ; ; ). Serotonergic firing in the raphe nucleus is also increased by sustained VNS after 2 weeks of treatment ( ; ). Antidepressant medications, on the other hand, primarily act via effects on neurotransmitter release and inhibitory autoreceptor desensitization ( ). VNS has not been found to cause desensitization of serotonergic and noradrenergic receptors, as is the case in antidepressant medications ( ).
Antiinflammatory effects of VNS
Other studies have considered an antiinflammatory mechanism of VNS’s role in TRD ( ). The vagus nerve is known to modulate inflammatory pathways ( ), while inflammation is increasingly acknowledged in the pathophysiology of depression ( ). Proinflammatory cytokines are known to be elevated levels in depressed patients, and are implicated in behavioral changes associated with depression ( ). One small clinical trial of VNS in TRD demonstrated increased circulating levels of antiinflammatory cytokines in patients after 3 months of VNS ( ). VNS’s potential antiinflammatory mechanism is further supported by its efficacy in various inflammatory diseases, including rheumatoid arthritis and Crohn’s disease ( ). A clinical trial in patients receiving VNS for rheumatoid arthritis demonstrated that VNS inhibited proinflammatory cytokine production ( ).
Combined, these studies paint VNS’s complex role in mediating brain areas crucial to affect, starting in the peripheral nervous system, passing through the brainstem and midbrain, and on to higher cortical structures directly involved in mood. VNS’s relative lack of invasiveness (compared with therapeutics which directly involve the central nervous system, like deep brain stimulation) is striking in that it demonstrates how a peripheral intervention can broadly modulate upstream targets. More research is required to clarify the relative importance of the various pathways thought to play a role in VNS’s antidepressant effects. Nevertheless, VNS’s effect on several targets known to play a role in depression, as well as the correlation of onset of metabolic changes in these targets with onset of VNS’s antidepressant effects, suggest that researchers are on the right path.
Indications
VNS is considered in patients suffering from treatment-resistant depression, which the United States FDA defines as four antidepressant treatment failures. The VNS registry, created after VNS for TRD gained FDA approval in 2005, included patients who met this criterion ( ). The registry included patients with both unipolar and bipolar depression, patients who had failed electroconvulsive therapy in the current episode, patients who had been depressed for more than 3 years, and patients who suffered with co-morbid anxiety—features which typically exclude such patients from other clinical trials ( ).
Patients in the VNS registry had 20 years of depressive illness on average. They had an average of 8.2 prior treatment failures, and MADRS scores averaging 33.1 on average—a number consistent with marked-to-severe depression ( ). These metrics indicate that VNS should be considered for those patients with the very most severe burden of depressive illness.
As discussed in , VNS may have a unique role in treatment of bipolar depression, where FDA approved treatments are scarce. Large trials of VNS in depression have included patients with both unipolar and bipolar depression, and results demonstrate similar efficacy between these groups ( ; ).
also discuss the relationship between personality disorders and TRD, noting that VNS should perhaps be avoided in patients with a strong history of personality pathology, including those with a history of severe self-injurious and parasuicidal behaviors. Patients with unrealistic expectations and poor frustration tolerance may not adapt well to VNS, particularly given the invasiveness of this treatment and the possible need for explanation should treatment not be tolerated. Because VNS requires months to demonstrate clinical efficacy, implantation should also be avoided in acutely suicidal patients. The process of pursuing implantation should not delay addressing the acute safety needs of the patient.
Clinical studies
Early studies of VNS for depression
Early studies of VNS for TRD were of mixed quality and demonstrated mixed efficacy. Of the 33 studies included in one systematic analysis, just six were conducted with double-blind, randomized, sham-controlled designs ( ). Follow up periods, primary outcomes, and stimulation parameters were heterogeneous across studies, making it more difficult to reach broad conclusions from VNS’s early trials.
The first study of VNS in patients with depression was an open label, nonrandomized, single-arm study conducted by . Thirty patients with both unipolar and bipolar depression received 10 weeks of VNS, resulting in a 40% response rate based on both HAM-D and CGI-I scores and 50% response rate based on MADRS scores.
The first randomized controlled trial of VNS for TRD was performed in 235 patients with nonpsychotic depression by . Of these patients, all of whom underwent surgical implantation of the VNS device, 112 were randomized to active treatment while 110 received sham treatment. The primary outcome measure—change in Hamilton Depression Rating Scale after 10 weeks of treatment—demonstrated no statistically significant difference between active and control groups. Of the three secondary outcome assessments (MADRS, CGI-I, and IDS-SR), only the IDS-SR demonstrated a statistically significant difference between active and sham treatment. Subsequent studies would reveal that the short follow-up period likely accounted for the largely negative results of this trial ( ). Moreover, the first 2 weeks of the active stimulation period were spent on dose titration, meaning patients only received full treatment for 8 weeks. Despite the limitations of this study’s design, CMS made their noncoverage decision for VNS in TRD in 2007 despite the negative results from this trial.
Patients from this study were subsequently included in a long-term, noncontrolled efficacy trial, during which patients from both the active and sham group of the acute phase could receive active VNS for 9 months ( ). This study, which used the change in Hamilton Depression Rating Scale over the 9-month treatment period as its primary outcome, demonstrated a 27.2% response rate and a 15.8% remission rate. Similar results were achieved in the secondary outcomes, including MADRS scales. Response rates doubled from the beginning of the long-term phase of trial to its end—a result that contributes to the current understanding that VNS requires between 6 and 12 months of active treatment to achieve clinical efficacy. As the authors acknowledged, the lack of a control group in this chronic study prevents definitive assertion that the improvements are attributable to VNS.
VNS dose-finding study
In the only “dose-finding” trial for VNS to date, compared the effects of low-, medium-, and high-intensity stimulation by varying output current and pulse width. Duty cycle and pulse frequency were uniform across the groups. This multicenter trial assessed 316 patients during a 22-week acute phase, followed by a 28-week long-term phase to assess durability of response. A total of 102 patients received low-dose stimulation (0.25 mA, 130 μs), 101 medium-dose (0.5–1.0 mA, 250 μs), and 107 high-dose (1.25–1.5 mA, 250 μs). Results from both the acute and chronic phases did not demonstrate statistically significant separation between the low, medium, and high-dose groups in the primary or secondary outcome assessments. Nevertheless, all three groups did demonstrate overall significant improvements in depression ratings compared to baseline scores. This outcome suggests that even low-dose stimulation is effective in treating depression, and is considered the main reason for the lack of separation. More patients in the medium- and high-dose groups (9%–11%) did experience remission by the end of the acute phase compared with the low-dose group (5%–6%). Also notable was the finding that patients in the low-dose group were more likely to have a depressive relapse at 50 weeks compared with those in the medium- and high-dose groups. Thus, while low-dose stimulation was effective in treating depression, there were other important clinical differences between the groups.
The VNS registry study
In the largest study of VNS in TRD to date, , compared 489 VNS patients with 276 patients receiving treatment as usual. This was a 5-year, prospective, open-label, nonrandomized, and observational registry study established by the manufacturer of the VNS device after it gained FDA approval in 2005. Patients in the VNS arm were among the most depressed and treatment resistant, with 8.2 failed antidepressant treatments on average, 1.8 lifetime suicide attempts on average, mean baseline MADRS scores indicating moderate to severe depression, and 57% of patients with prior ECT exposure. The results of this study demonstrated VNS’s striking potential to help such patients.
The primary outcome looked at antidepressant response based on MADRS scores at 5 years of follow-up. On this measure, 67.6% of VNS patients achieved antidepressant response, while 40.9% of TAU patients responded, as demonstrated in Fig. 21.3 ( Fig. 21.4 ).
Related posts:
Clinical and epidemiological predictors of treatment-resistant depression
Monoamine oxidase inhibitors for treatment-resistant depression
Inflammation and treatment resistance: Mechanisms and treatment implications
Treatment-resistant psychotic depression
Treatment-resistant bipolar depression
Drugs under investigation for treatment-resistant depression
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
