Vagus Nerve Stimulation As an Adjunctive Neurostimulation Tool in Treatment-resistant Depression

The outcome measurements of VNS effectiveness usually include a decline in depressive symptoms, measured by observer-rated scales and/or self-rated assessment instruments [e.g., the Hamilton Depression Rating Scale, the Montgomery-Åsberg Depression Rating Scale, or the Beck Depression Inventory (BDI)]⁶.

The first trial that systematically examined VNS in patients with major depressive episodes was conducted by Rush et al.²⁴, who used VNS as an augmentation strategy with a 10 week combination VNS treatment alongside patients' normal medication. The average baseline 28 item Hamilton Depression Rating Scale score was 38.0. Forty percent of patients responded favorably, while 17% showed remission²⁴. Sackheim et al.²⁵ combined the initial study cohort with an additional 30 patients who were monitored for twelve weeks. The authors observed a response rate of 30% and a remission rate of 15%. Schlaepfer and colleagues examined the antidepressive effects of VNS in an open, uncontrolled European multicenter study. Response and remission rates were 37% and 17%, respectively²⁶. Notably, in long-term follow-ups, an increase in both response and remission rates could be observed (Figure 1)^23,27,28. In a sham-controlled, multicenter study, no significant difference between active and sham groups was observed²⁹. However, the authors only performed ten weeks of active or sham VNS. Again, follow-up observations with higher remission and response rates suggest that VNS treatment has a beneficial effect over time^30,31. The TRD registry followed the clinical course and outcome for TRD patients treated with and without adjunctive VNS²². In the observational study of unipolar or bipolar depression, the outcomes of 500 patients with VNS and 300 with treatment as usual (TAU) patients were compared. The subjects were permitted to choose between VNS and TAU at the time of screening. Registry results indicated that the adjunctive VNS group had better clinical outcomes than the TAU group. The 5 year cumulative response rate (67.6% compared to 40.9%) and remission rate (43.3% compared to 25.7%) were significantly higher in the VNS + TAU group than in the TAU-only group (Figure 2 and Figure 3)²². Moreover, the findings in the literature suggest decreased suicide attempts, reduced levels of suicidal ideation, and fewer hospitalizations due to depression in patients with VNS than in those with the same severity of disease but who are taking medications^22,32.

In the following, we present three clinical cases to highlight the potential benefits as well as possible difficulties of providing VNS therapy to patients with TRD.

A 53 year-old woman with a 20 year history of TRD was referred to our office. Her treatment included ineffective trials of tricyclic antidepressants, serotonin reuptake inhibitors, serotonin-norepinephrine reuptake inhibitors, and adjunctive treatments with antipsychotic agents and anxiolytics. She had also received a series of ECT. Although suicidality initially remitted, she remained depressed and required further ECT. ECT was continued, but her improved mood was unstable and did not result in a meaningful decrease in antidepressant medications. Her psychopathological characteristics included feelings of worthlessness, low self-esteem, and a daily lack of energy and motivation. The patient's long course of treatment-resistant major depressive episodes made her a candidate for VNS therapy. The patient underwent VNS therapy device implantation by a neurosurgeon during a 1.5 h procedure. The VNS was checked for functionality upon insertion. Stimulation was reactivated during an office visit within several weeks. At that time, the patient additionally received sertraline (200 mg/day). During the first stimulation visit, the vagus nerve stimulator was adjusted to 1.00 mA, 25 Hz, and 30 s on/5 min off. During the second office visit, the dosing was increased to 1.50 mA, and multiple 0.25 mA increases were performed. The patient did not report any side effects. Within three months, the current had been titrated up to an output of 1.5 mA, with a standard signal frequency of 30 Hz, pulse width of 500 µs, and signal time of 30 s on/5 min off. However, at that time, the patient experienced only minor improvements in mood and quality of life. The BDI showed a slight reduction of depression. The current output was set at 2.25 mA, and during the following two weeks, an improvement was observed. We then added lithium at a dosage of 900 mg/day to her medications and decreased the VNS off time to 2 min. During the next three months, the patient described an improvement of her mood and energy levels. Her BDI sum score showed a remission (i.e., the patient no longer met the criteria for major depression). Fortunately, she did not experience adverse events or side effects. Her medication was reduced to sertraline (100 mg/day). Perhaps the most significant improvement since the device adjustment was the reduction in medication use. The patient is now maintained on antidepressant therapy with sertraline and lithium. The device settings have not been changed again. After successful adjustment of the vagus nerve stimulator, the patient now participates every six months for control visits.

A 58-year-old chronically depressed patient was referred to our office. She had experienced ongoing depressive episodes since the age of 15, requiring several courses of ECT. With each relapse of severe depression, she presented in a characteristic way, namely with periods of emotional non-response, unresponsiveness, an expressionless face, and staring eyes. When depressed, she would not care for herself, resulting in poor oral intake and rapid weight loss. She also socially disengaged from people and refused to take medication. She received several sessions of emergency ECT and showed a good response. During the last ECT session, the patient again showed marked improvement. Her BDI score was 46 before ECT and 18 afterwards. However, she refused to undergo maintenance ECT. She reported experiencing severe headache and memory loss after ECT. The patient's long and excellent response to ECT in terms of antidepressant effects made her a candidate for VNS therapy²². The patient underwent VNS therapy device implantation. The VNS was checked for functionality when it was inserted. Stimulation was reactivated during an office visit later that month. At the time of VNS initiation, the patient received citalopram (30 mg/day) and mirtazapine (60 mg/day). During this visit, the vagus nerve stimulator was adjusted to 1.00 mA. Within five months, the current had been titrated up to an output of 2 mA, standard signal frequency of 25 Hz, pulse width of 250 µs, and signal time of 30 s on/5 min off. Six months after implantation of the VNS device, the patient was taking several antidepressants, including lithium and citalopram. The device was implanted 13 years ago. Since then, the patient's depressive symptoms have remained stable. She has reported subjective well-being and no side effects of VNS.

A 62-year-old man was referred to our office with a 40 year history of TRD. Due to the severity of the disease, he had been a patient in psychiatric hospitals and had made several severe suicide attempts. His depressive symptoms included melancholic mood, anhedonia, loss of appetite with consecutive weight loss, decreased energy, hopelessness, nihilistic delusion, and suicidal intentions. Somatic comorbidities consisted of controlled hypothyroidism and controlled epilepsy. The first seizure occurred after a closed head trauma when he was 49 years old. The patient had on average 3 seizures per year (tonic-clonic) between his 49th and 53th birthdays. Over the previous years, the patient had received treatments with different combinations of 2–5 antidepressants (e.g., citalopram, paroxetine, fluvoxamine, imipramine, amitriptyline, and venlafaxine extended release) without complete remittance. Remittance also failed with antidepressant augmentation trials with lithium, aripiprazole, and olanzapine. He also had 19 courses of ECT (bilateral and unilateral) during the period from 1980 to 2005. The ECT effect was mostly short-lasting, as no sustained or long-term benefit from ECT was observed. He then underwent implantation with a VNS device for TRD in 2010. Previously, his prescribed medications (i.e., duloxetine, quetiapine, lithium, and topiramate) were unchanged. The vagus nerve stimulator was checked for functionality at the time of insertion. Stimulation was reactivated during an office visit later that month. During that visit, the vagus nerve stimulator was adjusted to 1.75 mA. At the second in-office visit, dosing was increased to 2.00 mA. The patient reported no side effects. Within 14 weeks, the current had been titrated up to 2.5 mA with a standard signal frequency of 30 Hz, pulse width of 500 µs, and signal time of 30 s on/5 min off. However, during an in-office visit, the patient reported an unusual globus sensation and hoarseness. Depressive symptoms markedly improved. The BDI score before implantation was 46, but only 15 at 14 weeks post-implantation. Because of the remarkable antidepressant effect of VNS in this patient, we only made minor changes to the VNS setup. Moreover, the patient did not have any seizures after implantation of the VNS device. During the following three visits, which occurred over 2 weeks, we adapted the output to 2 mA and signal frequency to 40 Hz and adopted an extraordinary off time of 8 min. With this setup, mild side effects occurred, but the antidepressant effect persisted. In accordance with the patient's wishes, we did not change the device settings.

These three patients, as well as our other patients receiving VNS, were prepared as required so that individual needs were addressed within the first six months after device implantation. After the first 6 months, follow-up appointments should be scheduled after 9, 12, and 24 months, or when an adjustment of dosing or device maintenance is necessary. The setup must be adapted to the patient's psychopathology, compatibility, psychological examination, somatic status (including history of side effects), laboratory, ECG, and (if necessary) imaging results.

Figure 1: Development of depression severity (p < 0.001) under VNS treatment compared to baseline (before VNS stimulation). The depression severity is coded as: 0 = no depression, 1 = mild depression, 2 = moderate depression, and 3 = severe depression. Treatment varied between 3 and 200 months (mean 104.9 months). This figure has been modified from Muller et al. (2017)²³ with permission. The sample size was 18 (6 female, 12 male; mean age 54). Hamilton Depression Rating Scale raw scores were translated into an ordinal classification of depression severity. Differences in depression severity (pre- and post-treatment) were compared using a paired sample t-test. Pearson's product-moment correlation coefficient was computed to assess the relationship between VNS stimulation parameters and depression severity. Please click here to view a larger version of this figure.

Figure 2: First-time response rate after 5 years based on the Montgomery-Åsberg Depression Rating Scale (MADRS). This figure has been modified with permission from Aaronson et al.²². A total of 765 patients (489 in the VNS group and 276 in the TAU group) was included in the efficacy analyses. Please click here to view a larger version of this figure.

Figure 3: First-time remission rate based on the MADRS (≤9). This figure has been modified with permission²². A total of 765 patients (489 in the VNS group and 276 in the TAU group) was included in the efficacy analyses. Please click here to view a larger version of this figure.

Table 1: Adverse events [n (%)] reported during the first and second years on VNS treatment. This table has been modified with permission from Berry et al.³³.

Table 2: Dosing recommendation for VNS. Normal-mode stimulation: stimulation 24 h/day and 7 days/week. Dosing is optimized for therapeutic effect and/or tolerability.