The Use of Repetitive Transcranial Magnetic Stimulation and Vagal Nerve Stimulation in the Treatment of Depression

Paul B. Fitzgerald; Zafiris J. Daskalakis

Curr Opin Psychiatry.  2008;21(1):25-29.  ©2008 Lippincott Williams & Wilkins

Posted 01/08/2008

Abstract and Introduction

Abstract

Purpose of review: Patients with depressive disorders often fail to respond to standard antidepressant medications and have few available treatment alternatives. Repetitive transcranial magnetic stimulation and vagal nerve stimulation have been developed and investigated over the last 10 years as potential treatment options for this and other psychiatric conditions. The aim of this paper is to review recent therapeutic trials of these techniques.
Recent findings: Recent studies appear to have confirmed that standard left-sided repetitive transcranial magnetic stimulation has antidepressant efficacy, but that the degree of clinical effect may be somewhat limited. Promising data are emerging suggesting that other approaches, including right unilateral repetitive transcranial magnetic stimulation and sequential bilateral stimulation, may have equal or potentially greater effects. The evidence for the effectiveness of vagal nerve stimulation remains restricted to the primary company-sponsored trials. Although limited, these data suggest that valuable treatment effects may develop over time.
Summary: Further repetitive transcranial magnetic stimulation research should actively investigate novel stimulation approaches before high-frequency left-sided stimulation is accepted as the standard approach. Given the invasive nature of vagal nerve stimulation and potential side effects, further research is urgently required. This should include the development of predictors of clinical response and definition of stimulation parameters with enhanced efficacy.

Introduction

Recent years have seen a marked escalation of interest in the use of invasive and noninvasive forms of brain stimulation for the treatment of mood disorders, especially for patients with treatment-resistant depression. Research into the use of transcranial magnetic stimulation (TMS) first commenced in earnest in the mid-1990s and since then a considerable number of randomized controlled trials have been published. The majority of trials published, however, have included relatively small samples of patients and treatment has been relatively restricted in duration and dose. Recently, trials have begun to include larger numbers of patients and have provided treatment for longer periods of time. As such there is a growing evidence base from which to judge the potential value of this technique. Vagal nerve stimulation (VNS) is supported by a more restricted evidence base, but has moved fairly rapidly into clinical practice in some countries. Therefore, there is considerable clinical interest in both of these techniques, and as such the focus of this article is to review recent studies that relate to their efficacy, safety and mechanism of action.

Repetitive Transcranial Magnetic Stimulation

TMS is a noninvasive means of stimulating nerve cells in superficial areas of the brain. During a TMS procedure, an electrical current passes through a wire coil placed over the scalp which induces a magnetic field. The magnetic field is applied to produce an electrical field in the brain that is capable of inducing depolarization of nerve cells resulting in the stimulation or disruption of brain activity. [1] High-frequency repetitive TMS (rTMS) usually refers to the application of rTMS at frequencies above 1 Hz and is the type of stimulation most commonly utilized in treatment studies. Physiological and human imaging experiments indicate that high-frequency rTMS increases local brain activity/excitability. [2,3*] In contrast, rTMS at 1 Hz is referred to as slow or low-frequency rTMS and this seems to be associated with the reduction of local brain activity. [3*,4,5]

Left Prefrontal Repetitive Transcranial Magnetic Stimulation

The majority of rTMS studies have used high-frequency stimulation targeting the left dorsal lateral prefrontal cortex. There have been two types of randomized trials conducted of this form of stimulation: trials comparing rTMS with sham and trials comparing it with electroconvulsive therapy (ECT). With regard to the former, several of the largest trials conducted have been reported recently. In one of these, Avery et al . [6**] randomized 68 patients with treatment-resistant depression to 15 sessions of active or sham rTMS. This was provided at a dose compatible with or somewhat greater than most previous studies (32 daily trains at an intensity of 110% of the resting motor threshold). The study reported statistically significant differences in active compared with sham treatment on the Hamilton Depression Rating Scale, analyzed dimensionally and also in terms of remission and response rates. The response rate in the active group was 30.6% compared with 6.1% in the sham group.

The second substantive trial of left-sided high-frequency rTMS is a large commercially sponsored trial conducted across 22 sites in the US, Canada and Australia. [7**] This trial used a dose of stimulation much greater than that utilized in previous research in regard to both the intensity (120% of the resting motor threshold) and number of stimuli applied per day (75 trains). The trial used an innovative solid core coil and a sham system that blinded both patient and treater. In total, 301 patients were initially randomized to an acute treatment protocol that could extend for up to 6 weeks with a further 3-week rTMS taper period. A significantly greater reduction in depression was achieved in the active than in the sham group at 4 and 6 weeks on most of the rating scales used in the study, although not in the primary outcome variable - scores on the Hamilton Depression Rating Scale. Response rates at 6 weeks were approximately 24% in the active compared with 12% in the sham group. It is notable that the increase in dose provided in this trial did not result in a substantially greater response rate than that seen in previous studies (e.g. [6**,8] ). Interestingly, the patients selected were somewhat less treatment resistant than in these previous studies (they were allowed to have failed no more than four medication trials). Analysis of data from this trial considering the issue of the level of treatment resistance has yielded some interesting findings (M. Demitrack, personal communication). When only those patients who had failed one antidepressant treatment course in the current episode were analyzed (164 patients, 76 sham, 88 active), there was a highly significant difference in response between the groups ( P = 0.002 at week 2 and P = 0.0006 at week 4 for the Montgomery Åsberg Depression Rating Scale score) and a more substantial difference in mean change in depression score (e.g. approximately 5 Montgomery Åsberg Depression Rating Scale points at week 4). On the primary outcome measure, standardized effect size calculations showed a 'large' effect size for this group (0.94) compared with a 'moderate' effect size for the overall population (0.39).

Several recent studies have also added to the literature comparing rTMS with ECT. A number of previous studies have randomized patients to one of these two conditions (e.g. [9-11] ) and all have found no efficacy differences except for a subgroup of patients with psychotic depression in a single study. [9] Despite somewhat similar methods, the two recent studies have produced conflicting results. Eranti et al . [12**] randomized 46 patients to either a fixed 15-day course of left-sided rTMS or a course of ECT that was flexible in duration and mode of administration (uni- or bilateral). At the end of acute treatment there was a significant treatment advantage for the ECT compared with the rTMS group. Notably, the group of patients in this study had a mean age of greater than 60. This is an important feature as the literature suggests that rTMS response is reduced in the elderly, [13] unless adjusted for scalp to cortex distance as done by Avery et al . [6**] . In addition, the study, like most of the rTMS-ECT comparisons published to date, did not allow for sequential trials of other forms of rTMS despite the ECT conditions being either variable or sequentially tried.

In the second study, 42 patients were randomized (and 35 completing subjects evaluated) to ECT (right unilateral followed by bilateral in nonresponders) or 4 weeks of left rTMS at 100% of the resting motor threshold (25 trains daily). [14**] No differences in response were found with either the total or completing group on any variable. A moderate response rate similar to most previous ECT rTMS trials (e.g. [9-11] ) was found.

Almost all previous rTMS studies have used a once per day, 5 days per week treatment schedule. Several studies have recently been conducted to analyze whether other treatment approaches are successful or potentially more effective. In the first of these Loo et al . [15*] randomized 38 patients to either sham stimulation or high-frequency left-sided rTMS provided twice daily over a 2-week period, with a possible further 4 weeks of unblinded once-daily treatment. Active treatment was significantly better than sham stimulation. This finding is of interest, but a comparison of twice- with once-daily stimulation in a substantial sample is obviously required to assess whether there are any advantages with this approach, especially given the resource implications of providing twice-daily treatment. In the opposite approach, a small study was conducted comparing 2 weeks of daily stimulation with treatment provided three times a week in week 1 and two times a week in week 2. [16*] No difference in response was seen between these groups, although the sample of 16 patients may have been insufficient to show subtle or moderate differences.

Other Approaches

In contrast to high-frequency left prefrontal rTMS, a number of previous studies have investigated the efficacy of low-frequency stimulation (usually 1 Hz) applied to the right prefrontal cortex (e.g. [8,17] ). A number of recent studies have further investigated this approach. In a small study of 27 patients, active right-sided stimulation at a low dose was found to be significantly better than placebo over a 4-week treatment course. [18*] A significantly larger study ( n = 130) compared 15 min of 1 Hz with 2 Hz rTMS to see whether there was a specific effect of frequency on response to right-sided stimulation. [19**] No difference was found between stimulation conditions. Interestingly, the overall response rate to right-sided stimulation was close to 50%, with a very high response rate in patients with a depressive phase of bipolar affective disorder. A small open-label pilot study suggested that 1 Hz right-sided rTMS may ameliorate both depressive symptoms and the symptoms of panic disorder in patients with comorbidity. [20]

A number of recent studies have also considered the value of a combination of low-frequency right-sided and high-frequency left-sided stimulation. One of these compared a substantial group of patients ( n = 50) randomized to either sequential bilateral stimulation or sham over up to a 6-week period. [21**] This study showed a substantial improvement of active over sham stimulation, with a progressive improvement in the active treatment group across the entire 6 weeks of the study. There was a response rate in the active group of approximately 50%. A second study randomized patients to sham stimulation, bilateral stimulation or bilateral stimulation with location derived from single photon emission computed tomography scanning. [22*] The last group received 1 Hz stimulation to an area of the brain that was relatively overactive and high-frequency stimulation to an area of the brain that was underactive, considering bilateral frontal and temporoparietal regions. Overall, active treatment did improve depression to a greater degree than sham. There was, however, no additional advantage of single photon emission computed tomography targeting. Despite this negative conclusion, this remains an interesting approach and the numbers of subjects in this study may have just been insufficient to show significant differences, especially as the targeted group did improve to a greater degree than the nontargeted active group.

Safety

A number of studies have recently addressed safety issues with rTMS. No cases of seizure have been reported in the growing number of depression trials in recent years. A seizure was, however, reported in a chronic pain protocol where 10 Hz stimulation was applied for 10 s - a train length in excess of guidelines. [23] A study by Anderson et al . [24*] reported interesting safety data on normal subjects given high doses (almost 13 000 pulses per day) over 3 days. Despite this large exposure, no side effects were noted. [24*] There was also a report of two cases of altered visual function (one improved, one worsened) in two patients receiving left prefrontal rTMS (1 and 5 Hz) in depression trials. Both patients had some form of baseline visual change and the TMS-induced effects resolved over days or weeks. [25] Finally, Xia et al . [26*] have summarized 13 reported cases of rTMS-related switch to mania/hypomania. Most of these occurred in patients with bipolar disorder. The authors conclude that the rate of switch may not vary substantially from that induced by sham, that it is not related to one specific rTMS parameter and that it resolves with usual antimanic treatment.

Summary

Recent evidence tends to confirm previous findings that high-frequency left prefrontal rTMS has antidepressant properties; however, studies of this type of stimulation published to date, even newer studies with larger patient samples, higher doses and longer durations of treatment, tend to demonstrate only quite moderate effect sizes. The lack of a dramatic effect of this form of stimulation is also mirrored in the TMS-ECT comparative research where equivalence of these techniques is seen only when the ECT response rate is more modest than evident in other research. Importantly, at least in part supported by inconsistencies in the research base suggesting that high-frequency left prefrontal rTMS may be the most valuable approach (e.g. [27] ), research is now moving beyond this initial paradigm and exploring other alternatives. Unilateral right-sided rTMS, sequential bilateral approaches and innovative variations such as those using targeting based on imaging techniques or completely novel stimulation parameters require further attention. Some of these approaches have a promising research base, but require large confirmatory studies, and there remains considerable room for the testing of innovative approaches.

Vagal Nerve Stimulation

VNS is a technique that has been more recently trialed for the treatment of patients with resistant depression. VNS involves the implantation of a pulse generator, similar to a pacemaker, in the chest. This is connected to a stimulating electrode attached to the vagus nerve in the neck. [28,29] The commercially available VNS therapy system has a programmable pulse generator that is able to provide chronic low-frequency intermittent pulsed electrical signals. These are provided through bipolar nerve-stimulating electrodes that are subcutaneously channeled and connected to the left cervical vagus nerve. Stimulation parameters are adjusted using a telemetric wand and personal digital assistant with variation in stimulation intensity, frequency, pulse width and duty cycle possible. Over the past 10 years, VNS has become a relatively routine procedure for the treatment of refractory partial-onset seizures with approval in Europe in 1994 and the US in 1997. The use of VNS for the treatment of depression was suggested by the observation of improved mood in patients receiving VNS for epilepsy, the observation that anticonvulsant drugs may have antidepressant activity, and the finding that VNS affects mood-relevant brain systems such as the locus coeruleus and raphe nucleus. [28,30*]

The initial pilot trials of VNS were published in the early 2000s and the substantive pivotal trials used for commercial registration in 2005. [31] These showed no substantial efficacy over the 10-week acute treatment period, [32] but a pattern of later and sustained response that was better than with usual treatment. [33] There was a doubling of response rates between 3 and 12 months, suggesting progressive benefit. There have been a few additions to this literature in the last year. The most substantial of these studies looked at the persistence of response to VNS and the relationship of this to the timing of its onset. Data were analyzed from both the initial pilot and pivotal registration trials. [34**] Data from the larger registration trial found an early response rate of 14.6% plus a late response rate of 19.5%. Of both responder groups, more than 60% maintained response over 24 months despite the inclusion of patients with substantial treatment resistance. Response was quite stable and did not appear to relate to altered medication treatment. This suggests that, although only a minority of patients respond to VNS, these patients have a good possibility of achieving a stable and persistent clinical improvement.

A smaller study reported outcomes for 11 patients. [35*] There was an impressive increase in response rate from one of 11 at 3 months to six of 11 at 12 months. A number of serious adverse events were, however, reported. These included a suicide in a nonresponder, a case of the recurrence of pre-VNS pulmonary emboli and two patients who experienced persistent vocal cord palsies, lasting 2 months in one patient and 6 months in the other. In addition to these reports, one study has made an interesting investigation into the potential effects, mediated through the vagal nerve, of VNS on appetite regulation. [36] Although no overall effects on weight were reported in the registration trials, a number of lines of evidence suggest VNS may alter appetite and weight. [36] In this study, VNS was found to alter cravings for sweet food, with around half of subjects experiencing an increase and the rest a decrease, a difference partially explained by the intensity of stimulation and baseline weight. Finally, there is a report of 14 patients who received ECT whilst a VNS device was in place. [37*] This appears quite safe, and the research suggests that sequential or concurrent VNS and ECT is practical and may be effective.

Summary

The evidence for the effectiveness of VNS remains limited to the primary company-sponsored trials. This database is shallow, but suggests that valuable treatment effects may develop over time. Given the invasive nature of the procedure and potential side effects, further research is urgently required to try and develop predictors of clinical response, to find stimulation parameters with enhanced efficacy and to better understand its mechanism of action.

Conclusion

Despite considerable attention provided to this field in recent years we remain early in the development of alternative brain stimulation techniques. The development of large multisite trials of techniques such as rTMS and VNS including those independently funded is warranted, and given the considerable problems associated with treatment-resistant depression and the lack of progress in developing pharmaceutical options for these patients, this would seem to be an appropriate area for priority funding. It is timely to consider the development of decent studies in relatively neglected areas such as the use of rTMS as a maintenance/longer term treatment. Large-scale studies comparing ECT and rTMS are warranted, but with a better balance of the treatment conditions than has been the case with studies to date. It is also interesting to consider the possibilities for future trials that directly compare rTMS and VNS in head to head studies.

References

* of special interest
** of outstanding interest

Funding Information

P.B.F. and this work were supported by a Practitioner Fellowship grant from NHMRC and a NHMRC project grant (436710). P.B.F. and Z.J.D. are each supported by a NARSAD Young Investigator award.

Reprint Address

Professor Paul B. Fitzgerald, MBBS, MPM, PhD, FRANZCP, Alfred Psychiatry Research Centre, The Alfred, First Floor Old Baker Building, Commercial Road, Melbourne, Victoria 3004, Australia Tel: +61 3 9076 6552; fax: +61 3 9076 6588; E-mail: paul.fitzgerald@med.monash.edu.au

Paul B. Fitzgerald , a Zafiris J. Daskalakis b

a The Alfred and Monash University School of Psychology, Psychiatry and Psychological Medicine, Victoria, Australia
b Centre for Addiction and Mental Health Clarke Division, Toronto, Ontario, Canada

Disclosure: Both authors have received direct research funding for participation in a clinical trial for Neuronetics Inc.