The Blue Cross and Blue Shield Association (BCBSA) Technology Evaluation Center (TEC) published an assessment of repetitive TMS (rTMS) for depression in 2009. The TEC Assessment concluded that the available evidence did not permit conclusions regarding the effect of TMS on health outcomes. Limitations of the evidence included:
- Equivocal efficacy in the largest sham-controlled trial of TMS,
- Uncertain clinical significance of the short-term anti-depressant effects found in meta-analyses,
- A lack of information beyond the acute period of treatment, and
- Lack of comparison with standard therapy (a second course of antidepressant therapy) in the population for whom TMS is indicated (patients who have failed one 6-week course of antidepressant medication).
BCBSA published an updated TEC Assessment of TMS for depression in 2011. Included were six recent meta-analyses, the largest of which evaluated 30 double-blind sham-controlled trials with a total of 1,383 patients. Recent clinical trials were also reviewed. The 2011 TEC Assessment reached the following conclusions:
- The meta-analyses and recent clinical trials of TMS generally show statistically significant effects on depression outcomes at the end of the TMS treatment period. However, the clinical significance and durability of the effect are not well-characterized.
- The largest randomized clinical trial showed a greater effect in patients with only one prior treatment failure, with possibly minimal or no effect in patients with greater than one prior treatment failure. Based on current evidence, it cannot be determined whether TMS after one treatment failure would be as effective as the current standard of a second course of antidepressant therapy.
- Also identified as gaps in current knowledge are whether TMS is effective as an adjunctive treatment and whether retreatment is effective.
Following is a summary of the key literature to date, focusing on randomized controlled trials (RCTs). Studies published prior to 2008 are included if the study design was a randomized sham-controlled double-blind trial that enrolled at least 40 subjects; refer to the 2008 meta-analysis by Schutter for a summary of study characteristics and stimulation parameters used in these trials. The evidence review is divided by key differences in treatment protocols, specifically high-frequency left dorsolateral prefrontal cortex stimulation (DLPFC), low-frequency (1–2 Hz) stimulation of the right dorsolateral prefrontal cortex, or combined high-frequency and low-frequency stimulation.
Note that over the last decade, there has been a trend to increase the intensity, trains of pulses, total pulses per session, and number of sessions (Gross et al.). Unless otherwise indicated in the trials described below, stimulation was set at 100% to 120% of motor threshold, clinical response was defined as an improvement of 50% or more on the Hamilton Depression Rating Scale (HAM-D), and remission was considered to be a score of 7 or less on the HAM-D.
High Frequency rTMS of the Left Dorsolateral Prefrontal Cortex (DLPFC) for Treatment-Resistant Depression (TRD)
Lam and colleagues conducted a meta-analysis of 24 randomized controlled trials comparing active versus sham rTMS in patients with TRD, although there were varying definitions of TRD. This analysis calculated a number needed to treat of six, with a clinical response in 25% of active rTMS and 9% of sham rTMS patients. Remission was reported for 17% of active rTMS and 6% of sham rTMS patients.
The largest study (23 study sites) included in the meta-analysis was a double-blind multicenter trial with 325 TRD patients randomized to daily sessions of high frequency active or sham rTMS (Monday to Friday for six weeks) of the left DLPFC (O’Reardon et al.). TRD was defined as failure of at least one adequate course of antidepressant treatment. Patients had failed an average of 1.6 treatments in the current episode, with about half of the study population failing to benefit from at least two treatments. Loss to follow-up was similar in the two groups, with 301 (92.6%) patients completing at least one post-baseline assessment and an additional 8% of patients from both groups dropping out before the four-week assessment. Intent-to-treat analysis showed a trend favoring the active rTMS group in the primary outcome measure (two points on the Montgomery-Asberg Depression Rating Scale; p = 0.057) and a modest (two-point) but significant improvement over sham treatment on the HAM-D. The authors reported that after six weeks of treatment the subjects in the active rTMS group were more likely to have achieved remission than the sham controls (14% vs. 5%), although this finding is limited by loss to follow-up.
In 2010, George et al. reported a randomized sham-controlled trial that involved 199 patients treated with left-prefrontal rTMS. This was a multi-centered study involving patients with a moderate level of treatment resistance. The response rate using an intention-to-treat analysis was 14% for rTMS and 5% for sham (p=0.02). In this study, the site for stimulation was determined through pre-treatment magnetic resonance imaging (MRI). Results from Phase 2 (open treatment of non-responders) and Phase 3 (maintenance and follow-up) will be reported in the future.
Another randomized sham-controlled double-blind trial was conducted in 68 patients who had failed at least two courses of antidepressants (Avery et al.). Three patients in each group did not complete the 15 treatment sessions or were excluded due to a change in medication during treatment, resulting in 91% follow-up. Independent raters found a clinical response in 31% (11 of 35) of the active rTMS patients and 6% (2 of 33) of the sham group. The average change in HAM-D was 7.8 for the active group and 3.7 for the control group. The Beck Depression Inventory (BDI) decreased by 11.3 points in the active rTMS group and 4.8 points in controls. Remission was observed in seven patients (20%) in the active rTMS group and one patient (3%) in the control group. Regarding effectiveness of blinding, 15% of subjects in each group guessed that they were receiving active TMS after the first session. After the 15th session, 58% of the rTMS group and 43% of the sham group guessed that they had received active TMS; responders were more likely than non-responders (85% vs. 42%) to think that they had received the active treatment. The 11 responders were treated with antidepressant medication and followed up for six months. Of these, one was lost to follow-up, five (45%) relapsed, and five (45%) did not relapse.
Rossini and colleagues randomized 54 patients who had failed at least two adequate courses of antidepressants to sham control or active rTMS at 80% or 100% of motor threshold for 10 sessions over a two-week period. Double-blind evaluation found an intensity-dependent response with 6% (1 of 16) of the sham, 28% (5 of 18) of the 80% MT, and 61% (11 of 18) of the 100% motor threshold groups showing improvement of 50% or more over a five-week evaluation. All of the patients reported that they were unaware of the differences between sham and active stimulation.
In a 2008 report, Mogg et al. randomized 59 patients with major depression who had failed at least one course of pharmacotherapy for the index depressive episode. In this study population, 78% of the patients had failed two treatment courses and 53% had failed three. The sham coil, which was provided by Magstim, was visually identical to the real coil and made the same clicking sound, but did not deliver a magnetic field to scalp or cortex. Blinded assessments were measured two days after the fifth and final (tenth) sessions (97% follow-up), with additional assessments at six weeks (90% follow-up) and four months (83% follow-up). The mean group difference was estimated to be 0.3 points in HAM-D scores for the overall analysis. Interpretation of this finding is limited since seven sham patients (23%) were given a course of real rTMS after the six-week assessment and analyzed as part of the sham group in the intent-to-treat analysis. The study was powered to detect a difference of 3.5 points in the HAM-D between the active and sham groups, and the 2.9-point group difference observed at the end of treatment was not significant. A higher percentage of patients in the active rTMS group achieved remission criteria of ≤8 points on the HAM-D (25% vs.10% control), and there was a trend for more patients to achieve clinical response in the active rTMS group (32% vs.10%, p = 0.06). All of the 12 patients who met the criterion for clinical response (nine active and three sham) thought that they had received real rTMS, with more patients in the active group (70%) than the sham group (38%) guessing that they had received the real treatment. Interpretation of this finding is also limited, since the reason the subjects guessed that they had active treatment was not reported, and the subjects were not asked to guess before they began to show a clinical response.
A small double-blind randomized trial from 2009 suggests that specific targeting of Brodman areas 9 and 46 may enhance the anti-depressant response compared with the standard targeting procedure, i.e., measuring 5 cm anterior from the motor cortex (Fitzgerald et al., 2009, p.1255-62). Fifty-one patients who had failed at least two 6-week courses of antidepressant medication (average 5.7 failed courses) were randomized to a standard localization procedure or to structural MRI-aided localization for three weeks (with one-week extension if >25% reduction on the Montgomery-Asberg Depression Rating (MADRS). Six patients in the targeted group and 10 in the standard group withdrew due to lack of response. A single patient in the targeted group and five in the standard group withdrew for other reasons, resulting in 17 patients in the targeted group and 12 in the standard group continuing for the full four weeks of treatment. To adjust for the imbalance in discontinuation rates, a mixed model statistical analysis was used. There was a significant difference between the groups in the overall mixed model analysis, and planned comparisons showed significant improvement in MADRS scores for the targeted group at four weeks. Response criteria were met by 42% of the targeted group and 18% of the standard group. Remission criteria were met by 30% of the targeted group and 11% of the standard group. Although encouraging, additional trials with a larger number of subjects are needed to evaluate this procedure.
Several studies have compared the outcomes of rTMS with those from ECT. In one study, 40 patients with nonpsychotic major depression were treated over the course of a month (20 total sessions) and evaluated with the HAM-D, in which a response was defined as a 50% decrease with a final score of less than or equal to 10 (Grunhaus et al.). There was no difference in response rate between the two groups; 12 of 20 responded in the ECT group compared to 11 of 20 in the magnetic stimulation group. A United Kingdom National Institute for Health Research health technology assessment compared efficacy and cost effectiveness of rTMS and ECT (McLoughlin et al.). Forty-six patients who had been referred for ECT were randomized to either ECT (average of 6.3 sessions) or a 15-day course (five treatments per week) of rTMS of the left DLPFC. ECT resulted in a 14-point improvement in the HAM-D and a 59% remission rate. rTMS was less effective than ECT (five-point improvement in HAM-D and a 17% remission rate). Another study reported no significant difference between ECT and rTMS in 42 patients with TRD; however, response rates for both groups were low (Rosa et al.). The number of remissions (score of ≤7 on the HAM-D) totaled three (20%) for ECT and two (10%) for rTMS.
Janicak and colleagues reported on assessment of relapse during a multisite, open-label study. In this study, patients who met criteria for partial response during either a sham–controlled or open-label phase of a prior study were tapered from rTMS and simultaneously started on maintenance antidepressant monotherapy. They were then followed for 24 weeks. Ten of 99 patients relapsed. Thirty-eight patients had symptom worsening, and 32 of these (84%) had symptomatic benefit with adjunctive rTMS. Additional data are needed related to durability of effect and to maintenance phases.
Low Frequency rTMS of the Right Dorsolateral Prefrontal Cortex (DLPFC) or Bilateral Stimulation for Treatment-Resistant Depression (TRD)
Fitzgerald et al. (2003) randomized 60 patients who had failed a minimum of at least two 6-week courses of antidepressant medications into 1 of 3 groups; high frequency left rTMS, low frequency right rTMS, or sham stimulation over 10 sessions. All patients who entered the study completed the double-blind randomized phase, which showed no difference between the two active treatments (left: 13.5% reduction; right: 15% reduction) and greater improvements in the MADRS scores compared to the sham group (0.76% reduction). Only one patient achieved 50% improvement during the initial two weeks. Then, only the subjects who showed at least 20% improvement at the end of the 10 sessions (15 active and two sham) continued treatment. Patients who did not respond by at least 20% were switched to a different active treatment. From week 2 to week 4 there was greater improvement in the low frequency right rTMS group compared with the high frequency left rTMS group (39% vs. 14% improvement in MADRS). Seven patients (18% of 40) showed a clinical response of >50% by the end of the four weeks.
In a subsequent study Fitzgerald and colleagues (2006) randomized 50 patients with TRD to sequential bilateral active or sham rTMS. After two weeks of treatment, three subjects had dropped out of the sham treatment group and there was a slight but non-significant improvement favoring the active group for the MADRS (26.2 vs. 30.9) and the BDI (18.3 vs. 21.6). At this time point, 60% of subjects receiving active rTMS and 50% of subjects receiving sham treatment guessed that they were in the active group. The clinical response was reported by subjects as the major reason for their guess, with 11 of 13 responders (nine active and two sham) guessing that they were in the active group. As in the earlier study, only the subjects who showed at least 20% improvement at the end of each week continued treatment. Treatment on week three was continued for 15 subjects in the active group and seven subjects in the sham group. By week six, 11 subjects in the active rTMS remained in the study, with no control subjects remaining. Final ratings for the 11 subjects who continued to respond through week six were 8.9 on the MADRS and 9.2 on the BDI.
Another multicenter double-blind trial randomized 130 patients with TRD to five sessions per week of either 1- or 2-Hz rTMS over the right DLPFC (Fitzgerald et al., 2009, p.655-66). Sixty-eight patients (52%) completed four weeks of treatment; there was an approximate 30% improvement in depression scales, with no differences between the 1- or 2-Hz groups. Due to the potential for placebo effects for this type of intervention, the absence of a sham control group limits interpretation.
A small randomized, sham-controlled trial was published in 2010 that involved either right or left rTMS in 48 patients with TRD (Triggs et al.). Overall reductions in the HAM-D-24 from baseline to three months were not significantly different between rTMS and sham treatment groups. In this small study, right cranial stimulation was significantly more effective than left cranial stimulation (sham or rTMS).
rTMS as an Adjunctive Treatment for Moderate to Severe Depression
Schutter conducted a meta-analysis of 30 double-blind randomized sham-controlled trials (1,164 patients) of high frequency rTMS over the left DLPFC in patients with major depression. The pooled weighted mean effect size for treatment was calculated with Hedges g (a standardized mean difference that adjusts for sampling variance) to be 0.39 (95% confidence interval 0.25–0.54), which is considered moderate. For 27% of the population rTMS was used as a primary/adjunctive treatment; three trials were included that used rTMS as a primary/adjunctive treatment for depression and enrolled more than 40 subjects (Koerselman et al., 2004; Rumi et al., 2005; Herwig et al., 2007). rTMS has also been examined in patients with clinical evidence of cerebrovascular disease and late-life depression (Jorge et al.). Additional research on whether adjunctive rTMS can improve response to pharmacologic treatment as a first-line therapy is also needed.
The largest area of TMS research outside of depressive disorders appears to be treatment of auditory hallucinations in schizophrenia resistant to pharmacotherapy. In 2011, BCBSA TEC published an assessment of TMS as an adjunct treatment for schizophrenia. Five meta-analyses were reviewed, along with RCTs in which measurements were carried out beyond the treatment period. A meta-analysis of the effect of TMS on positive symptoms of schizophrenia (hallucinations, delusions, and disorganized speech and behavior) did not find a significant effect of TMS. Four meta-analyses that looked specifically at auditory hallucinations showed a significant effect of TMS. It was noted that outcomes were evaluated at the end of treatment, and the durability of the effect is unknown. The TEC Assessment concluded that the available evidence is insufficient to demonstrate that TMS is effective in the treatment of schizophrenia.
Other Psychiatric/Neurologic Disorders
Two small (n=18 and 30) randomized sham-controlled trials found no evidence of efficacy for treatment of obsessive compulsive disorder (OCD), although another small sham-controlled trial (n=21) reported promising results with bilateral stimulation of the supplementary motor area (Sachdev et al.; Mantovani et al.; Mansur et al.).
In 2011, Short et al. evaluated the efficacy of adjunctive rTMS as a treatment for fibromyalgia pain in a small randomized controlled pilot study. Twenty patients with fibromyalgia, defined by the American College of Rheumatology criteria, were randomized to 10 sessions of left prefrontal rTMS or sham TMS along with their standard medications. At two weeks after treatment, there was a significant change from baseline in average visual analog scale (VAS) for pain in the rTMS group (from 5.60 to 4.41) but not in the sham-treated group (from 5.34 to 5.37). There was also a significant improvement in depression symptoms in the active group compared to baseline (from 21.8 to 14.10) but not in the sham group (from 17.6 to 16.4). There were no statistically significant differences between the groups in this small trial. Additional study with a larger number of subjects is needed.
Ahmed et al. randomized 45 patients with probable Alzheimer’s disease to five sessions of bi-lateral high-frequency rTMS, bi-lateral low-frequency rTMS, or sham TMS over the dorsolateral prefrontal cortex. Thirty-two patients had mild to moderate dementia and 13 had severe dementia. There were no significant differences between groups at baseline. Measures of cortical excitability immediately after the last treatment session showed that treatment with high-frequency rTMS reduced the duration of transcallosal inhibition. At three months after treatment, the high-frequency rTMS group improved significantly more than the other two groups in standard rating scales, and subgroup analysis showed that this was due primarily to improvements in patients with mild/moderate dementia. Patients in the subgroup of mild to moderate dementia who were treated with high-frequency rTMS improved from 18.4 to 22.6 on the Mini Mental State Examination (MMSE), from 20.1 to 24.7 on the Instrumental Daily Living Activity (IADL) scale and from 5.9 to 2.6 on the Geriatric Depression Scale (GDS).
In 2008, Walpoth et al. reported no evidence of efficacy of rTMS in a small trial (n=14) of patients with bulimia nervosa.
Practice Guidelines and Position Statements
The Canadian Network for Mood and Anxiety Treatments (CANMAT) updated their clinical guidelines on neurostimulation therapies for the management of major depressive disorder in adults (Kennedy et al.). The evidence reviewed supported ECT as a first-line treatment under specific circumstances; when used in patients who have failed to respond to one or more adequate antidepressant medication trials, ECT response rates have been estimated to be 50-60%. The guidelines considered rTMS to be a safe and well-tolerated treatment, with no evidence of cognitive impairment. Based on the 2008 meta-analysis by Lam et al., response (25%) and remission (17%) rates were found to be greater than sham but lower than for other interventions for TRD, leading to a recommendation for rTMS as a second line treatment. The guidelines indicated that there is a major gap in the evidence base regarding maintenance rTMS, as only one open-label case series was identified.
The Movement Disorder Society published an evidence-based review of treatments for the non-motor symptoms of Parkinson’s disease in 2011 (Seppi et al.). The review found insufficient evidence to make adequate conclusions on the efficacy rTMS for the treatment of depression in Parkinson’s disease.
In 2010 guidelines, the American Psychiatric Association stated that data are insufficient to recommend rTMS as initial therapy in major depressive disorder; for patients with inadequate response to pharmacotherapy, ECT remains the most effective form of therapy and should be considered, but TMS may be an option.
In 2007 guidelines, the National Institute for Health and Clinical Excellence stated that, due to lack of sufficient data regarding clinical efficacy, rTMS should be utilized in research studies only to provide further analysis of factors such as treatment duration and frequency and intensity of application; no major safety concerns have been identified with the use of TMS in severe depression.
Evidence on rTMS for depression and other psychiatric and/or neurologic disorders is insufficient to permit conclusions regarding the effect of this technology on health outcomes. The literature on rTMS for treatment-resistant depression is the most developed and includes a number of double-blind randomized sham-controlled short-term trials. Results of these trials show mean improvements of uncertain clinical significance across groups as a whole. The percentage of subjects who show a clinically significant response is reported at approximately 2-3 times that of sham controls, with approximately 15% to 25% of patients meeting the definition of clinical response (Mantovani et al.).
The treatment protocols are time intensive, and the patients who are most likely to benefit from treatment are not currently known. Importantly, a number of open issues need to be addressed before this procedure is widely implemented. The available studies do not establish that rTMS is as good as available alternatives, as the vast majority of the trials do not compare rTMS to alternative active treatments. Alternative treatments include a variety of different medication regimens and psychological talk therapy, both of which have demonstrated efficacy. In addition, further research is needed to determine which of the locations and treatment parameters examined to date are most effective to guide the number of sessions needed to elicit a clinically significant response, to determine whether the response is durable with or without anti-depressant medications, and to provide some information about whether maintenance treatments are needed, and which types of maintenance treatment are most effective..
A search for transcranial magnetic stimulation on the online site ClinicalTrials.gov indicates that these issues are being actively investigated. Given the number of important questions that remain for this novel treatment approach, the current state of knowledge is not sufficient to expand utilization outside of the research setting. Therefore, rTMS is considered experimental, investigational and unproven for the treatment of depression and other psychiatric and/or neurologic disorders.
A search of published literature about navigated transcranial magnetic stimulation (nTMS) located only small case series and reviews of records. No random controlled trials were located.
Picht et al. (37) conducted an assessment of the influence of nTMS on surgical planning for tumors in or near the motor cortex, to evaluate how much influence, benefit, and impact nTMS has on the surgical planning for tumors near the motor cortex. This study reviewed the records of 73 patients with brain tumors in or near the motor cortex, mapped preoperatively with nTMS. The surgical team prospectively classified how much influence the nTMS results had on the surgical planning. Stepwise regression analysis was used to explore which factors predict the amount of influence, benefit, and impact nTMS has on the surgical planning. The influence of nTMS on the surgical planning was as follows: it confirmed the expected anatomy in 22% of patients, added knowledge that was not used in 23%, added awareness of high-risk areas in 27%, modified the approach in 16%, changed the planned extent of resection in 8%, and changed the surgical indication in 3%. The study concluded that nTMS had an objective benefit on the surgical planning in one fourth of the patients and a subjective benefit in an additional half of the patients. It had an impact on the surgery itself in just more than half of the patients. By mapping the spatial relationship between the tumor and functional motor cortex, nTMS improves surgical planning for tumors in or near the motor cortex.
Saisanen et al. (38) conducted a study that characterized the muscle responses from human primary motor cortex system by navigated TMS to provide normative values for the clinically relevant TMS parameters on 65 healthy volunteers aged 22 to 81 years. They delivered focal TMS pulses on the primary motor area, and recorded muscle responses on thenar and anterior tibial muscles. Motor threshold, latencies and amplitudes of motor-evoked potentials, and silent period duration were measured. The correction of the motor-evoked potential latency for subjects' height was provided. The authors concluded that they provided a modified baseline of TMS-related parameters for healthy subjects, and stated that earlier such large-scale baseline material has not been available.
Vajkoczy et al (39) conducted a comparative evaluation of navigated brain stimulation and direct electrocortical stimulation in the first 30 patients presenting with rolandic tumor. Their objective was evaluation of the concordance between results of non-invasive mapping with the NBS System (Nexstim Oy, Finland) and direct electrocortical stimulation (DCS), and evaluation of clinical utility of NBS mapping data through changed or improved surgical strategy. This was a series of 30 patients for whom the authors found preoperative mapping of the motor cortex with NBS to be as accurate as direct cortical stimulation (DCS). NBS mapping has the potential to be of higher clinical utility for preoperative motor mapping than functional magnetic resonance imaging (fMRI).
The US National Institutes of Health has a trial underway on nTMS in monitoring stroke recovery (NCT01005394). The purposes of this research study are to: 1) determine changes in brain activity during the first 6 months after stroke (to determine how the brain "re-wires"); 2) compare changes in recovery of motor function with changes in brain re-wiring; 3) determine the ability of TMS to "predict" functional outcome in the first 6 months after stroke. The primary hypotheses are: 1) functional recovery will be correlated with TMS changes (as measure motor threshold (MT), motor evoked potentials (MEPs) and recruitment curves; 2) baseline TMS will predict future functional outcomes at 3 and 6 months. This trial is currently recruiting participants. (40)
Disclaimer for coding information on Medical Policies
Procedure and diagnosis codes on Medical Policy documents are included only as a general reference tool for each policy. They may not be all-inclusive.
The presence or absence of procedure, service, supply, device or diagnosis codes in a Medical Policy document has no relevance for determination of benefit coverage for members or reimbursement for providers. Only the written coverage position in a medical policy should be used for such determinations.
Benefit coverage determinations based on written Medical Policy coverage positions must include review of the member’s benefit contract or Summary Plan Description (SPD) for defined coverage vs. non-coverage, benefit exclusions, and benefit limitations such as dollar or duration caps.