Repetitive Transcranial Magnetic Stimulation for Treatment-Resistant Depression in Adult and Youth Populations: A Systematic Literature Review and Meta-Analysis

Abstract

Background:

Between 30% and 60% of individuals with major depressive disorder will have treatment-resistant depression (TRD): depression that does not subside with pharmaceutical treatment. Repetitive transcranial magnetic stimulation (rTMS) is an emerging treatment for TRD.

Objective:

To establish the efficacy and optimal protocol for rTMS among adults and youth with TRD.

Data Sources:

Two systematic reviews were conducted: one to determine the efficacy of rTMS for adults with TRD and another to determine the effectiveness of rTMS for youth with TRD. For adults, MEDLINE, Cochrane Central Register of Controlled Trials, PubMed, EMBASE, PsycINFO, Cochrane Database of Systematic Reviews, and Health Technology Assessment Database were searched from inception until January 10, 2014 with no language restrictions. Terms aimed at capturing the target diagnosis, such as depression and depressive disorder, were combined with terms describing the technology, such as transcranial magnetic stimulation and rTMS. Results were limited to studies involving human participants and designed as a randomized controlled trial. For youth, the search was altered to include youth only (aged 13–25 years) and all study designs. When possible, meta-analysis of response and remission rates was conducted.

Study Selection:

Seventy-three articles were included in this review: 70 on adult and 3 on youth populations.

Results:

Meta-analysis comparing rTMS and sham in adults found statistically significant results favoring rTMS for response (RR: 2.35 [95% CI, 1.70–3.25]) and remission (RR: 2.24 [95% CI, 1.53–3.27]). No statistically significant differences were found when comparing high- and low-frequency, unilateral and bilateral, low- and high-intensity rTMS or rTMS and electroconvulsive therapy (ECT). While meta-analysis of results from the youth literature was not possible, the limited evidence base suggests that rTMS may be effective for treating TRD in youth.

Conclusions:

The evidence available on the use of rTMS for adults with TRD indicates that rTMS is approximately twice as effective as a sham procedure, although the optimal rTMS protocol remains unclear. Evidence also indicates that rTMS is as effective as ECT and appears promising as a treatment for youth with TRD; however, the evidence base is underdeveloped.

Clinical Points

• ■ While the optimal treatment protocol is yet to be established, repetitive transcranial magnetic stimulation (rTMS) is an effective treatment compared to sham with minor side effects.

• ■ The performance of rTMS in comparison to electroconvulsive therapy is not well understood.

Globally, more than 350 million people of all ages suffer from depression.¹ Symptoms of depression include loss of focus, lack of energy, complaints of physical illness with no cause, and thoughts of suicide.² On the basis of the epidemiologic data available,^3,4 between 30% and 60% of people with major depressive disorder will have treatment-resistant depression (TRD), a type of depression in which the patient does not experience sufficient relief after adequate rounds of medication. The definition of TRD has not been standardized, and, in practice, varies from lack of response to 1 antidepressant to lack of response to at least 3 adequate antidepressant trials.⁴

Repetitive transcranial magnetic stimulation (rTMS) is a noninvasive procedure in which cerebral electrical activity is influenced by a rapidly changing magnetic field.⁵ The magnetic field is created by a plastic-encased coil that is placed over the patient’s scalp. This magnetic field can be directed onto specific areas of the brain and modulates cerebral activity using low or high frequencies. In contrast to electroconvulsive therapy (ECT), rTMS can induce cortical electrical activity without causing a seizure; it is subconvulsive and, therefore, does not require anesthesia.⁶

Recent systematic reviews and meta-analyses^7–9 on rTMS have established the efficacy of rTMS for adults with TRD compared to sham. However, the optimal rTMS protocol, the efficacy of rTMS compared to ECT, and the effectiveness of rTMS in youth populations remain unknown.

Thus, the objective of this study was to determine the efficacy, safety, and optimal protocol of rTMS for TRD in adults and youth in comparison to sham procedures and ECT. This systematic review and meta-analysis provides a complete synthesis of the effectiveness of rTMS, the delivery of rTMS, and its place compared to relevant alternatives across all populations.

METHOD

Two systematic reviews were completed: one on the efficacy of rTMS for adults with TRD and one on the effectiveness of rTMS for youth with TRD. For adults, MEDLINE, Cochrane Central Register of Controlled Trials, PubMed, EMBASE, PsycINFO, Cochrane Database of Systematic Reviews, and Health Technology Assessment Database were searched from inception until January 10, 2014 with no language restrictions. Terms aimed at capturing the target diagnosis, such as depression and depressive disorder, were combined with terms describing the technology, such as transcranial magnetic stimulation and rTMS. Results were limited to studies involving human participants and designed as a randomized controlled trial (RCT). For youth, the search was altered to include only subjects aged 13–25 years and all study designs.

All abstracts and full-text studies were screened by 2 reviewers/authors (L.E.L. and S.C.) using the inclusion and exclusion criteria outlined in Table 1. Abstracts selected for inclusion by either reviewer proceeded to full-text review; this abstract review was intentionally broad to ensure that all relevant literature was captured. Studies included after abstract review proceeded to full-text review. Studies were included if they met all inclusion criteria and failed to meet any of the exclusion criteria presented in Table 1. Any discrepancy between reviewers was resolved through consensus. After full-text review, published systematic reviews and meta-analysis on rTMS were hand-searched to ensure that all relevant articles were captured in the literature search.

Table 1.

Inclusion and Exclusion Criteria for Articles Used in the Study

Inclusion Criteria
Youth
Report TRD or patients who have had ≥ 2 previous treatments
Youth and young adult population aged 13–25 y
Report on 1 of the following:
Effectiveness of rTMS in comparison to placebo, pharmacologic therapy, cognitive therapy, or ECT
Effectiveness of 1 type/protocol of rTMS treatment in comparison to another type/protocol of rTMS treatment
Bipolar or unipolar depression
Patients who have not been treated with rTMS prior to the study
Controlled clinical trials, RCTs, cohort studies
Adults
Report TRD or patients who have had ≥ 2 previous treatments
Adult population ≥ 18 y
Report on 1 of the following:
Efficacy of rTMS in comparison to placebo, pharmacologic therapy, cognitive therapy, or ECT
Efficacy of 1 type/protocol of rTMS treatment in comparison to another type/protocol of rTMS treatment
Bipolar or unipolar depression
Report remission and/or response rates using the Hamilton Depression Rating Scale, Montgomery-Asberg Depression Rating Scale, or Beck Depression Inventory
Patients who have not been treated with rTMS prior to the study
RCT study design
Exclusion Criteria
Not TRD or do not report whether patients have TRD
Nonoriginal data
Not rTMS
Not unipolar or bipolar depression
Animal models
Preclinical and biological studies
Studies reported only in abstract or as poster presentations
Not reporting on efficacy/effectiveness of rTMS
Studies including patients who have not responded to rTMS in previous treatments
Case studies, case series

Abbreviations: ECT = electroconvulsive therapy, rTMS = repetitive transcranial magnetic stimulation, RCT = randomized controlled trial, TRD = treatment-resistant depression.

For all studies, year of publication, country, patient selection, patient characteristics, definition of treatment resistance, description of technologies, protocols for control and treatment, outcomes measured, instruments used, definition of response, definition of remission, and follow-up time were extracted in duplicate using data extraction forms. The primary outcomes, response, and remission rates postintervention were also extracted from each study. During data extraction, each included study was assessed for quality using the Cochrane Risk of Bias Checklist¹⁰ for RCTs or the Downs and Black Checklist¹¹ for all other study designs. Quality assessment was completed in duplicate, and discrepancies were resolved through discussion.

Meta-analysis was conducted when possible, with response and remission rates as the primary outcomes considered. The definitions of response and remission as defined by authors of the included articles were used. For each study, the number of participants who experienced remission and response were compared between the rTMS group and the comparator group. Meta-analyses were conducted using relative risk (RR) to express the efficacy of rTMS in relation to other comparators, and a random-effects model was used. Begg funnel plots were completed to assess the risk of publication bias. All analyses were completed using Stata/IC 13.1 (StataCorp, College Station, Texas).

RESULTS

Adult Populations

The search strategy identified 786 citations for the review of rTMS for adult populations. Of these, 602 were excluded and 184 articles proceeded to full-text review. An additional 114 articles were excluded during full-text review, resulting in 70 articles included in the final analysis (Figure 1). Although 5 published systematic reviews and meta-analyses^8,12–15 were hand-searched for articles not captured in the original search, no additional articles were identified. On the basis of the Cochrane Risk of Bias Checklist, the included studies were determined to be of moderate quality, with areas of low, unclear, and high risk of bias (Figure 2).

Figure 1.

Flowchart of Studies Included in the Review of Repetitive Transcranial Magnetic Stimulation (rTMS)

Figure 2.

Percentage of Studies Assessed as Having Low, High, or Unclear Risk of Bias Based on the Cochrane Risk of Bias Checklist

The 70 articles on rTMS in adult populations were further separated into 6 categories based on comparator: rTMS versus sham (45),^16–61 high-frequency versus low-frequency rTMS (14),^{33–35,38,49,62–70} bilateral versus unilateral rTMS (5),^{16,40,61,71,72} high-intensity versus low-intensity rTMS (3),^51,53,54 other rTMS protocols (13),^{30,37,44,46,58,73–80} and rTMS versus ECT (6).^81–86 Seventeen of the included studies^{16,27,30,33–35,37,38,40,41,44,46,49,53,54,58,68} had 3 comparator arms (2 rTMS arms and a sham arm). These 17 studies were included in both the rTMS versus rTMS and the rTMS versus sham categories. The results for each category are summarized in Table 2.

Table 2.

Meta-Analysis Results of Included Studies

	Outcome Measure	Included Studies	Total Included Patients, n (range)	Risk Ratio (95% CI)	I2 (%)	Findings
Adult
rTMS vs sham	Response	Avery et al,18 Bakim et al,54 Bares et al,56 Berman et al,20 Blumberger et al,16 Boutros et al,19 Chen et al,47 Fitzgerald et al,38,40 Garcia-Toro et al,44 George et al,21 Hernández-Ribas et al,45 Holtzheimer et al,60 Jorge et al,23,24 Kauffmann et al,25 Loo et al,41,42 Manes et al,28 Mantovani et al,29 McDonald et al,30 Mosimann et al,32 O’Reardon et al,36 Padberg et al,51 Paillère Martinot et al,58 Peng et al,48 Rossini et al,53 Stern et al,35 Su et al,49 Zheng et al50	rTMS: 685 (7–155) Sham: 650 (5–146)	2.35 (1.70–3.25)	36.1	rTMS is an effective treatment; patients undergoing rTMS are twice as likely to achieve either clinical response or remission than patients undergoing a sham procedure
	Remission	Avery et al,18 Bakim et al,54 Bares et al,56 Blumberger et al,16 Fitzgerald et al,39 George et al,21 Jorge et al,23,24 Kauffmann et al,25 Loo et al,41 Manes et al,28 Mantovani et al,29 McDonald et al,30 O’Reardon et al,36 Padberg et al,51 Rossini et al,53 Stern et al,35 Su et al49	rTMS: 536 (7–155) Sham: 501 (5–146)	2.24 (1.53–3.27)	1.1
High-frequency rTMS vs low-frequency rTMS	Response	Eche et al,62 Fitzgerald et al,38 Fitzgerald et al,63 Fitzgerald et al,64 Fitzgerald et al,65 Isenberg et al,66 Richieri et al,69 Rossini et al,70 Speer et al,34 Stern et al,35 Su et al4	High frequency: 226 (6–63) Low frequency: 239 (8–67)	1.19 (0.97–1.46)	0.0	The optimal frequency of rTMS is unclear; there is a trend toward high-frequency rTMS being more effective to achieve both clinical response and remission than low frequency rTMS; however, both 95% CIs cross 1.0, indicating that compared to low-frequency rTMS, high-frequency rTMS may be equivalent, more effective, or less effective
	Remission	Fitzgerald et al,39,63 Isenberg et al, 66 Speer et al,34 Stern et al,35 Su et al49	High frequency: 120 (10–63) Low frequency: 130 (8–67)	1.29 (0.75–2.22)	8.1
Unilateral rTMS vs bilateral rTMS	Response	Blumberger et al,16 Fitzgerald et al,40,61,71 Pallanti et al72	Unilateral: 232 (20–91) Bilateral: 298 (21–147)	1.15 (0.85–1.52)	45.8	The optimal location of treatment for rTMS is unclear;there is a trend toward bilateral rTMS being more effective to achieve both clinical response and remission than bilateral; however, both 95% CIs cross 1.0, indicating that compared to bilateral, unilateral rTMS may be equivalent, more effective, or less effective
	Remission	Blumberger et al,16 Fitzgerald et al,61 Fitzgerald et al71	Unilateral: 188 (26–91) Bilateral: 256 (21–147)	1.18 (0.71–1.96)	53.6
Low-intensity rTMS vs high-intensity rTMS	Response	Bakim et al,54 Padberg et al,51 Rossini et al53	Low intensity: 40 (10–18) High intensity: 39 (10–18)	1.15 (0.54–2.41)	57.3	The optimal intensity of rTMS is unclear; there is a trend toward high-intensity rTMS being more effective to achieve both clinical response and remission than low intensity; however, both 95% CIs cross 1.0, indicating that compared to low intensity, high-intensity rTMS may be equivalent, more effective, or less effective
	Remission	Bakim et al,54 Padberg et al,51 Rossini et al53	Low intensity: 40 (10–18) High intensity: 39 (10–18)	1.72 (0.89–3.33)	0.0
rTMS vs other rTMS protocols	Narrative summary	Image guidance: Fitzgerald et al,73 Garcia-Toro et al,44 Paillère Martinot58 Right vs left dorsolateral prefrontal cortex: Pascual-Leone et al,46 Triggs et al,37 Scheduling: Galletly et al,74 Turnier-Shea et al75 Electroencephalography timed: Price et al76 Combination protocols: Conca et al,77 Fitzgerald et al,78 Levkovitz et al,79 McDonald et al,30 Rybak et al80		NA	NA	Experimentation to identify the optimal rTMS protocol is ongoing with research exploring the impact of image-guided, scheduling, and combination therapy
rTMS vs ECT	Response	Grunhaus et al,81 Janicak et al,82 Rosa et al86	rTMS: 41 (8–20 ECT: 39 (8–20)	1.09 (0.79–1.48)	0.0	The effectiveness of rTMS compared to ECT is unclear; there is a trend toward rTMS being more effective to achieve clinical response but less effective to achieve remission; however, both 95% CIs cross 1.0, indicating that compared to ECT, rTMS may be equivalent, more effective, or less effective
	Remission	Grunhaus et al,81 Pridmore et al,84 Rosa et al86	rTMS: 44 (8–20 ECT: 44 (8–20)	0.97 (0.65–1.45)	0.0
Youth
rTMS		Bloch et al,87 Croarkin et al,88 Mayer et al89	rTMS: 24 (7–9)	NA	NA	The included studies suggest that rTMS may be an effective intervention for treatment-resistant youth and young adults; however, the evidence is too weak to draw conclusions

Abbreviations: NA = not applicable, rTMS = repetitive transcranial magnetic stimulation.

rTMS Compared to Sham

Of the 45 studies comparing rTMS and sham, 1 was conducted in Canada,¹⁶ 21 in the United States,^17–37 5 in Australia,^38–42 4 in Spain,^43–46 3 in China,^47–50 2 in Germany,⁵¹ 2 in Italy,^52,53 and the remaining 6 in various other countries (Turkey,⁵⁴ Belgium,⁵⁵ Czech Republic,⁵⁶ Denmark,⁵⁷ France,⁵⁸ Iceland⁵⁹). The studies were published between 1996⁴⁶ and 2013.^{29,34,45,47,55} Fifteen studies^{16,18,20,21,23,26,27,30,36,39,41,42,56,58,60} used an intention-to-treat analysis, 1⁵² used a per-protocol analysis, and the remaining did not report on the type of analysis that was conducted.

The number of participants included in each study varied between 6¹⁷ and 301,³⁶ with a total of 1,903 participants included across 45 studies. Frequency of rTMS varied from 1 Hz^{25,29,36,39,44,56} to 20 Hz,^{19,20,28,31–34,43,47,49,52,54,55} and motor threshold varied from 80%^{17,19,20,28,31} to 120%.^21,26,36,40 The number of rTMS sessions provided to each participant in the active arms varied from 5 to 30 over a period of 5 days to 6 weeks. The protocol used for the sham procedure was similar in all studies, with most using an rTMS machine turned on and at a 45° angle from the patient. Additional details can be found in Supplementary Table 1.

Thirty-one of the studies* comparing rTMS and sham provided adequate data on treatment response to permit pooling for meta-analysis and 18^{16,18,21,23–25,28–30,35,39,41,49,51,53,54,56} provided adequate data on treatment remission. As shown in Figures 3 and 4, the scale and threshold for both response and remission varied by article. The overall pooled RR for response was 2.35 (95% CI, 1.70–3.25) (Figure 3) and remission was 2.24 (95% CI, 1.53–3.27) (Figure 4). Thus, patients are twice as likely to achieve response or remission with rTMS than with a sham procedure. For response, there was evidence of publication bias (P value = .025), although visual inspection of the plot did not indicate any bias.

Optimal Protocol

Of the 35 studies comparing optimal protocols for rTMS, 14^{33–35,38,49,62–70} compared high-frequency and low-frequency rTMS, 5^{16,40,61,71,72} compared unilateral and bilateral rTMS, 3^51,53,54 compared high-intensity and low-intensity rTMS, and 13^{30,37,44,46,58,73–80} compared standard rTMS with various other rTMS protocols (eg, image guided, session scheduling, combination protocols). Additional details can be found in Supplementary Tables 2–5.

Eleven^{34,35,38,49,62–66,69,70} of the 14 studies assessing high-frequency versus low-frequency rTMS provided adequate data on treatment response, and 6 of the studies^{34,35,49,63,65,66} provided adequate data on treatment remission to permit pooling. Comparing high-frequency and low-frequency rTMS, the overall pooled RR for response was 1.19 (95% CI, 0.97–1.46) and remission was 1.29 (95% CI, 0.75–2.22). These results suggest that there is a tendency for high-frequency rTMS to result in more cases of response and remission, but neither result is statistically significant (Table 2).

Of the studies comparing unilateral and bilateral rTMS, 5^{16,40,61,71,72} provided adequate data on treatment response and 3^16,61,71 provided adequate data on treatment remission to permit pooling. The overall pooled RR for response was 1.15 (95% CI, 0.85–1.56) and remission was 1.18 (95% CI, 0.71–1.96). These results suggest that there is a tendency for bilateral frequency rTMS to result in more cases of response and remission, but neither result is statistically significant (Table 2).

All 3 of the studies comparing high- and low-intensity rTMS provided adequate data on treatment response and remission to permit pooling.^51,53,54 For treatment response, the overall pooled RR for high-intensity versus low intensity rTMS was 1.15 (95% CI, 0.54–2.41) and remission was 1.72 (95% CI, 0.89–3.33). These results suggest that there is a tendency for high-intensity rTMS to result in more cases of response and remission, but neither result is statistically significant (Table 2).

Thirteen RCTs^{30,37,44,46,58,73–80} assessing rTMS compared to various other rTMS procedures were included. Meta-analyses could not be performed due to the significant heterogeneity in rTMS protocol properties, duration of follow-up periods, reported outcome measures, control or comparison groups, and study quality and size. Three of these studies^44,58,73 investigated the use of image guidance in rTMS, 2^37,46 compared left and right cortex targeting, 2^74,75 compared the scheduling of rTMS sessions, 1⁷⁶ compared standard rTMS to rTMS using electroencephalogram activity, and 5^30,77–80 assessed the efficacy of combination protocols for rTMS treatment. Among the 13 RCTs^{30,37,44,46,58,73–80} that assessed the efficacy of rTMS compared to various other rTMS procedures, there was significant variability in the reporting of and criteria used to define treatment efficacy (ie, response and/or remission). The majority of studies were small (n ≤ 100 patients). Ten studies^{30,37,44,58,74,75–78,80} reported no differences in treatment efficacy (ie, rates of response and/or remission, mean change in outcome score) between groups, 3 studies^46,73,79 reported statistically significant differences in treatment efficacy between groups, and no studies reported worsening of TRD following any of the investigated rTMS treatments. Of the 3 studies^46,73,79 that reported statistically significant differences in outcomes between treatment groups, 1 study⁷³ reported higher rates of response and remission using a neuronavigational method for rTMS localization over the standard 5-cm method, 1 multiple crossover study⁴⁶ found that left-sided rTMS resulted in lower mean Hamilton Depression Rating Scale (HDRS) scores in comparison to right-sided, and 1 study⁷⁹ reported superior efficacy of high-intensity and left-sided (unilateral) rTMS compared to low-intensity and bilateral rTMS.

rTMS Compared to Electroconvulsive Therapy

Six RCTs^81–86 comparing rTMS with ECT were included. Two studies^83,84 were conducted in Australia, 1⁸⁶ was conducted in Brazil, 1 was conducted in Iran,⁸⁵ 1⁸¹ was conducted in Israel, and 1⁸² was conducted in the United States. Five of the 6 RCTs^{81,82,84–86} compared an rTMS arm to an ECT arm, while the remaining RCT⁸³ compared a combination of rTMS and ECT to ECT alone. Additional details can be found in Supplementary Table 6.

Three rTMS versus ECT studies^81,82,86 provided adequate data on treatment response to permit pooling. All 3 articles used the HDRS to determine treatment response, which was defined as a minimum 50% reduction in the HDRS score. The overall pooled RR for rTMS versus ECT is 1.09 (95% CI, 0.79–1.48) (Table 2). However, the results are not statistically significant different; rTMS may be less or more effective compared to ECT.

Three of the rTMS versus ECT studies^81,84,86 provided adequate data on treatment remission to permit pooling. While all 3 of these studies used the HDRS to define remission, 2^81,84 used a threshold score of 8 and 1⁸⁶ used a threshold score of 7.

The overall pooled RR for rTMS versus ECT remission rate is 0.97 (95% CI, 0.65–1.45). This result suggests that there is no statistically significant difference in remission rates of patients treated with rTMS compared to those treated with ECT.

Youth Populations

The systematic review of rTMS effectiveness with youth populations identified 140 abstracts. During abstract review, 114 studies were excluded, with 26 proceeding to full-text review. After being reviewed in full-text, 23 were excluded and 3 were included (Figure 1). The included studies were conducted in Israel,⁸⁷ the United States,⁸⁸ and Australia,⁸⁹ and all 3 were designed as prospective cohort studies, were small including 7^88,89 or 9⁸⁷ participants, and were conducted between 2008⁸⁷ and 2012.^88,89 Bloch et al⁸⁷ and Croarkin et al⁸⁸ recruited participants from medical centers and reported short-term outcomes after rTMS treatment (1 month posttreatment and 5 weeks posttreatment, respectively). Mayer et al⁸⁹ recruited participants from the Bloch et al⁸⁷ trial and reported patients’ long-term outcomes (6 years posttreatment) based on their original treatment allocation. Bloch et al,⁸⁷ Croarkin et al,⁸⁸ and Mayer et al⁸⁹ were given quality assessment scores of 17, 16, and 15, respectively, out of a possible total of 23 points on the modified Downs and Black Checklist. Areas where quality was most often lacking were related to whether an attempt was made to blind study subjects or those measuring intervention outcomes, respectively. Additional study details are available in Supplementary Table 7.

Bloch et al⁸⁷ used 10 Hz rTMS at 80% motor threshold for 14 sessions and assessed the outcomes at baseline; days 7, 10, and 14; and 1 month posttreatment. At 1 month posttreatment, 3 of 9 participants experienced clinical response (at least a 30% reduction in the Children’s Depression Rating Scale).⁸⁷ This study also found statistically significant reductions in depression measured using the Beck Depression Inventory at days 7 and 10 and 1 month posttreatment when compared to baseline (P < .05).⁸⁷ Using the Screen for Child Anxiety-Related Disorders Questionnaire, participants’ anxiety levels were significantly lower at the end of treatment and 1 month posttreatment (P < .05).⁸⁷ Statistically significant results were not found with the Suicide Ideation Questionnaire at any time point, indicating that there is no evidence to suggest that rTMS has an effect on suicidal ideations and behaviors.⁸⁷

Mayer et al⁸⁹ reported long-term (3-year) outcomes of the patients in the study by Bloch et al⁸⁷; no additional rTMS treatment was received. No statistically significant differences between long-term outcomes and outcomes at the end of treatment were reported. This finding suggests that participants experienced initial improvement in their depression severity and did not experience worsening or improvement in depression severity over time.⁸⁹

Croarkin et al⁸⁸ used 10 Hz rTMS at 120% motor threshold for 30 sessions over a period of 6 to 8 weeks and assessed outcome measures at baseline and at weeks 2, 4, and 6. Clinical response and remission were not reported. However, the mean Children’s Depression Rating Scale score was reduced from 69.3 (SD = 8.6) at baseline to 42.1 (SD = 10.7) (average of week 2, 4, and 6 outcomes reported). This study also reported increases in cortical activity over time when baseline is compared to week 5.

DISCUSSION

The 2 systematic reviews identified 70 relevant RCTs and 3 observational studies. Of these studies, 45 compared rTMS and sham, 35 compared optimal rTMS protocols, 6 compared rTMS and ECT, and 3 assessed the effectiveness of rTMS for youth populations. The body of evidence available on the use of rTMS for adults indicates that rTMS is approximately twice as effective as the sham procedure. However, the optimal frequency, location, and intensity of rTMS for adults remain unclear. There is likely no difference in efficacy between ECT and rTMS; rTMS may be more effective to achieve response but less effective to achieve remission; however, neither result is statistically significant. There is substantial experimentation still required to identify and improve the optimal rTMS protocol. Active research is ongoing with the use of image-guided techniques, scheduling of treatment, and timing of treatment. However, none of these research areas are developed enough to clarify the role of these variables in the effective use of rTMS.

Compared to the breadth of adult literature available on the use of rTMS for TRD, there is very little literature on youth and adolescent populations. Three articles on the treatment of youth with TRD were included in this research, all of which were prospective cohort studies. Using the Children’s Depression Rating Scale, both of the studies assessing short-term outcomes found a reduction in depression severity after rTMS treatment; 1⁸⁷ reported statistically significant reductions, while the other⁸⁸ did not conduct tests of significance. The former⁸⁷ found statistically significant reductions in depression severity at days 7 and 10 and 1 month posttreatment. Other outcome measures such as the Beck Depression Inventory and the Screen for Child Anxiety-Related Disorders Questionnaire also suggested statistically significant reduction in depression severity.

The results from the 3 included articles suggest that rTMS may be an effective method for alleviating severe depression in youth who have failed to respond to other treatments. However, the limited number and the low to moderate quality of the studies on this topic restrict the ability to draw generalized conclusions about the use of rTMS in this population. The rTMS protocols were heterogeneous among the included studies, precluding inference of the most appropriate protocol for this patient population. Furthermore, the small sample size in each article, with a total of 25 participants included in all 3 studies, does not provide a robust evidence base.

With depression affecting a sizeable number of youth and adolescents, finding acceptable, efficacious treatments for this patient population is of particular importance. The included literature suggests that rTMS may be an effective treatment option for youth and young adults with depression. However, with limited literature and data available, further studies, particularly large-scale high-quality studies with this patient population, are required before conclusive inferences can be drawn.

A number of limitations from both the adult and youth systematic reviews merit comment. Within the adult literature, there is significant heterogeneity in rTMS protocols, duration of follow-up periods, reported outcome measures, control or comparison groups, and study quality and size. When possible, studies were divided by comparator. However, it was not possible to further divide them on the basis of small protocol differences, and this may limit the robustness of our findings. Yet, given that no statistically significant differences were found between protocol differences such as high and low frequency, unilateral and bilateral treatment, and high and low intensity, it is unlikely that pooling the mixed protocols would have introduced significant bias.

Broadly, the included studies on the efficacy of rTMS for adult populations were of moderate quality, with most having a combination of unclear and low risks of bias and few having high risks of bias, as assessed by the Cochrane Risk of Bias Checklist. Blinding of participants and treatment providers was an area in which the included studies often suffered from a lack of clarity. Methods of random sequence generation were also largely unclear in the included studies. However, given that very few studies had a “high” risk in this area and most were “unclear,” these areas of bias could have been due to a lack of detail in method descriptions rather than an area of bias.

In both of the systematic reviews on effectiveness/efficacy (youth and adult), treatment response and remission were the primary outcomes assessed. These outcomes were selected because they were most frequently reported among the included studies and give a broad sense of patient improvement or worsening. Ideally, outcomes such as function and quality of life would be the primary outcomes assessed, as these outcomes would more closely determine the impact of rTMS treatment on a patient’s life. However, measures of function and/or quality of life are infrequently found in the literature, resulting in insufficient data to pool for meta-analysis. Thus, a major limitation is that the outcomes available in the literature are not directly measuring improvement in patient quality of life.

Both the adult and youth reviews were also limited by a lack of studies reporting long-term data. The vast majority of studies on rTMS assess only the short-term impact of treatment (4–6 weeks). Due to limited long-term data, it is not possible to draw conclusions about the length of treatment effect or the long-term safety of treatment and its impact on outcomes such as return to work or ability to complete daily tasks. Studies reporting long-term patient outcomes such as relapse, reoccurrence, participation in life, and major side effects are required.

Lastly, the majority of the included studies were conducted in the United States and Australia. With no reason to suspect that the patient mix and underlying etiology of major depressive disorder and TRD are substantially different in these countries, we anticipate that our findings are broadly generalizable to other countries.

In summary, the evidence available on the use of rTMS for adults with TRD indicates that rTMS is approximately twice as effective as a sham procedure and most likely as effective as ECT. However, the optimal rTMS protocol remains unclear, with no statistically significant differences between frequency, intensity, or location. rTMS appears promising as an effective treatment for youth with TRD, although the evidence base is underdeveloped.

Author contributions:

Design of the study (all authors), collection of data (Mss Leggett, Coward, and Lorenzetti), management of data (Mss Leggett, Coward, and Soril and Dr Clement), analysis of data (Mss Leggett, Coward, and Soril and Dr Clement), interpretation of data (Mss Leggett and Coward), preparation of the manuscript (all authors), review of the manuscript (all authors), and approval of the manuscript (all authors).

Potential conflicts of interest:

None reported.

Funding/support:

This work was supported by a financial contribution from Alberta Health.

Role of the sponsor:

The sponsor had no role in the design and conduct of the study; collection, management, analysis, and interpretation of data; or preparation of the manuscript. The sponsor did review and approve the manuscript content for publication.

Disclaimer:

The views expressed herein do not necessarily represent the official policy of Alberta Health.

Supplementary material:

See accompanying pages.