Review Paper

Repetitive Transcranial Magnetic Stimulation

for Treatment-Resistant Depression:
A Systematic Review and Metaanalysis

Raymond W Lam, MD, FRCPC1; Peter Chan, MD, FRCPC2;

Michael Wilkins-Ho, MD, FRCPC3; Lakshmi N Yatham, MBBS, MRCPsych (UK), FRCPC4

Objective: Systematic reviews show that repetitive transcranial magnetic stimulation

(rTMS) is superior to sham control conditions in patients with major depressive disorder,
but the clinical relevance is not clear. None have specifically examined outcomes in
patients with treatment-resistant depression (TRD).
Method: A systematic review was conducted by identifying published randomized
controlled trials of active rTMS, compared with a sham control condition in patients with
defined TRD (that is, at least one failed trial). The primary outcome was clinical response
as determined from global ratings, or 50% or greater improvement on a rating scale. Other
outcomes included remission and standardized mean differences in end point scores.
Metaanalysis was conducted for absolute risk differences using random effects models.
Sensitivity and subgroup analyses were also conducted to explore heterogeneity and
robustness of results.
Results: A total of 24 studies (n = 1092 patients) met criteria for quantitative synthesis.
Active rTMS was significantly superior to sham conditions in producing clinical response,
with a risk difference of 17% and a number-needed-to-treat of 6. The pooled response and
remission rates were 25% and 17%, and 9% and 6% for active rTMS and sham conditions,
respectively. Sensitivity and subgroup analyses did not significantly affect these results.
Dropouts and withdrawals owing to adverse events were very low.
Conclusions: For patients with TRD, rTMS appears to provide significant benefits in
short-term treatment studies. However, the relatively low response and remission rates, the
short durations of treatment, and the relative lack of systematic follow-up studies suggest
that further studies are needed before rTMS can be considered as a first-line monotherapy
treatment for TRD.
Can J Psychiatry 2008;53(9):621–631

Clinical Implications
· Adjunctive treatment with rTMS may be efficacious for patients with TRD.
· There are few adverse effects with rTMS and it is well-tolerated by patients.
· s and rTMS can be used in combination.

Limitations
· Unpublished studies were not sought in this systematic review.
· The identified randomized controlled trials had small sample sizes and short durations (1 to 4
weeks with a median of 2) of treatment.
· Follow-up after treatment discontinuation was usually not systematically conducted.

Key Words: depressive disorders, transcranial magnetic stimulation, treatment resistant

depression, metaanalysis, systematic review

The Canadian Journal of Psychiatry, Vol 53, No 9, September 2008 W 621

Review Paper
novel treatment for neuropsychiatric conditions such as
A depression is rTMS, a noninvasive procedure in which a
sequence of high-intensity magnetic pulses is applied to stim-
ulate cortical neurons.1 Advantages of rTMS, compared with
other somatic treatments such as ECT, include it being deliv-
erable in an office setting, it does not require anesthesia, and it
has fewer associated side effects.1
Several systematic reviews and metaanalyses of rTMS studies
in MDD have been conducted, including one for the Cochrane
Collaboration.2–8 These reviews have found statistically sig-
nificant effects of active rTMS over sham conditions. How-
ever, the clinical relevance of these findings is not as clear, as
the overall clinical response of rTMS (whether measured as
percentage improvement on depression rating scales, or as
dichotomous outcomes on global improvement scales) has
been rather limited. Hence, several of these systematic
reviews concluded that there was insufficient evidence to sup-
port the clinical use of rTMS in treating depression. In part,
this may have been owing to inclusion of earlier studies in
which stimulus parameters were still being investigated and
optimized. More recently, there has been consensus that
high-intensity HFL-rTMS, or LFR-rTMS are associated with
AD effects. Recent RCTs using these stimulation parameters
have shown greater clinical effects than previous studies. 8
Episodes of depression that are refractory to one or more AD
trials are referred to as TRD. Although there is still no clear
consensus for criteria for medication resistance,9 TRD (as typ-
ically defined by failure of 2 AD trials) may affect 15% or
more of patients treated for a major depressive episode.10
Given the large proportion of patients with TRD and the lim-
ited response to treatment, it is not surprising that TRD is asso-
ciated with a significant burden of illness and disability.11–13
Abbreviations used in this article
AD antidepressant
ECT electroconvulsive therapy
HDRS Hamilton Depression Rating Scale
HFL high-frequency to the left dorsolateral prefrontal cortex
LFR low-frequency to the right dorsolateral prefrontal cortex
MADRS Montgomery–Asberg Depression Rating Scale
MDD major depressive disorder
NNT number-needed-to-treat
RCT randomized controlled trial
rTMS repetitive transcranial magnetic stimulation
SMD standardized mean difference
TRD treatment-resistant depression
622
The management of TRD remains unclear as there is still lim-
ited ev id e n c e f o r o p tima l p h a r ma c o lo g ic an d
psychotherapeutic strategies for TRD.14–16 Given the limited
number of validated therapies for TRD, rTMS has been iden-
tified as an important investigational treatment.17 Indeed,
several recently published sham-controlled trials of rTMS
have focused on samples of patients with TRD. However,
many of the previous systematic reviews combined studies
with medication-resistant patients with those involving less
resistant samples and they have not specifically examined
efficacy of rTMS for TRD. The objective of this study was to
systematically review the published studies of rTMS in
patients with TRD, focusing on clinical outcomes that are rel-
evant to clinicians.
Methods
All relevant RCTs comparing rTMS with a sham control con-
dition were identified. Inclusion criteria for studies were:
diagnosis of MDD by the Diagnostic and Statistical Manual
of Mental Disorders or the International Classification of
Diseases criteria, an explicit definition of TRD that included
at least one failed trial of an AD, a sham control condition, an
outcome measure that included either a defined clinical
response or a continuous score on a depression rating scale,
and a description of the rTMS parameters used. Studies
involving depression comorbid to other medical conditions
(for example, Parkinson disease) were excluded.
The primary outcome was clinical response as defined by
either a defined percentage improvement on a continuous
score from a depression rating scale (for example, 50% or
greater improvement from baseline score to end of treatment
on the HDRS or the MADRS), or by a global rating scale (for
example, much improved or very much improved on the
Clinical Global Impression Scale). Secondary outcomes
included clinical remission, defined as either: a score on a
depression rating scale within the normal range (for example,
HDRS of 7 or less, or MADRS of 12 or less), or a global rat-
ing of not depressed or equivalent on the Clinical Global
Impression Scale; and SMD (also known as the effect size)
on end of treatment scores on depression rating scales.
Adverse effects were evaluated by examining the number of
dropouts owing to adverse effects, and total number of drop-
outs for any reason was examined as a proxy measure of treat-
ment acceptability.
Relevant RCTs were identified by electronically searching
MEDLINE (from 1966 onward), EMBASE (from 1980
onward), PsycINFO (from 1974 onward) and the Cochrane
Central Register of Controlled Trials databases up to May 15,
2008, using the search terms transcranial, magnetic, and
depress*. Two independent reviewers examined titles and
abstracts of the studies, and then checked full articles for
W La Revue canadienne de psychiatrie, vol 53, no 9, septembre 2008
 Repetitive Transcranial Magnetic Stimulation for Treatment-Resistant Depression: A Systematic Review and Metaanalysis
eligibility. To identify further trials, references of these
selected studies and of other review papers were also checked.
No language restriction was imposed.
Two independent reviewers extracted data and assessed the
quality of methodological reporting of selected studies, using
data extraction forms. The criteria for quality assessment were
based on recommendations in the Cochrane Handbook for
Systematic Reviews of Interventions,18 and included the qual-
ity of allocation concealment, blinding to treatment alloca-
tion, and plausibility of the sham condition. Any
disagreement was resolved by discussion with a third
reviewer and consensus among all 3 reviewers.
Posttreatment data were extracted from the end of blinded
acute treatment and after a no-treatment, blinded follow-up
period. When more than one active treatment (for example,
HFL and LFR stimulation) was compared with sham, the
results from the active treatment conditions were pooled
together. For crossover studies, only data from the first cross-
over sequence were used. In some studies, after a defined
period of blinded treatment, continuation treatment was
offered only to responders (leading to differential retention of
patients between conditions) or under unblinded, open-label
conditions. For these studies, we used only the data from the
blinded acute treatment period that included the original ran-
domized sample.
Data were double-entered into Review Manager 4.219 to
check for accuracy. Metaanalyses were then performed. For
dichotomous outcomes, an intent-to-treat analysis was
adopted. When dropouts were excluded from any assessment
in the studies (for example, those who never returned for
assessment after randomization), they were considered
nonresponders. Absolute risk differences, indicating the dif-
ference in response rates between the active and sham condi-
tions, and 95%CIs were calculated. We used a random effects
model rather than a fixed effects model because it is more
appropriate when heterogeneity is likely to be present.20 For
continuous outcomes, the SMD was calculated for end point
data of completers, again using a random effects model.
Results are reported as means and 95%CIs.
Heterogeneity, referring to variability among studies in a sys-
tematic review arising from clinical, methodological, or
statistical diversity, 21 was assessed by chi-square and
I-squared statistics.22 We also planned sensitivity and
subgroup analyses to examine robustness of findings, by lim-
iting to trials with stricter criteria for TRD (2 or more failed
trials of ADs), limiting to trials using high-intensity (motor
threshold of 90% or greater) HFL (10 Hz or greater) stimula-
tion, and limiting to trials with higher quality reporting (the
following rated at least adequate: description of allocation
concealment, double-blindness, and adequacy of sham
The Canadian Journal of Psychiatry, Vol 53, No 9, September 2008 W
treatment). Publication bias was examined using funnel plots
and visual inspection.
Results
Included Studies
A total of 32 RCTs of rTMS for TRD were identified. Among
these, 8 studies were excluded for the following reasons: no
sham treatment,23–25 no explicit definition for TRD,26,27
reporting cognitive or pain outcomes from another trial,28,29
and multiple crossovers with no data on the first treatment
sequence.30 Among the remaining studies, one did not report
clinical response or remission rates, or rating scale scores
after the first crossover sequence,31 leaving 23 studies avail-
able for quantitative synthesis. All studies used the HDRS or
MADRS as a primary outcome.
Table 1 lists the definitions of TRD and rTMS parameters
used in the 24 included trials. Nine studies used a definition
of TRD as failing one or more trials of ADs; the rest used the
more typical definition of failing 2 or more ADs. There was
reasonable homogeneity of rTMS variables, with HFL-rTMS
(10 Hz or higher), the most common active condition studied.
Most studies used 1 to 2 weeks of treatment, with only 3 stud-
ies using 3 weeks and only 2 studies using 4 weeks; almost all
studies limited treatment to weekdays (5 sessions for each
week). Most studies also included patients who were on vari-
ous medications, although doses were kept unchanged before
and during treatment. Only 4 studies included medication-
free patients. In 2 studies, rTMS was used concomitantly
with starting a new AD such as sertraline 50 mg32 and
escitalopram 20 mg.33
Only 8 studies systematically followed all patients,
untreated, after completing the double-blind phase (Table 2).
Among these, 2 studies also started patients on an AD at the
beginning of rTMS33,34 and 2 studies had only graphical pre-
sentation of group mean scores at follow-up.35,36 The remain-
ing 4 studies were available for quantitative synthesis.
Quantitative Synthesis
The chi-square and I-squared tests were significant (P < 0.05
or I2 > 30%) for all comparisons, indicating significant heter-
ogeneity in studies. The active rTMS condition had a signifi-
cant risk difference for clinical response of 17% (95%CI,
10% to 23%; n = 22 studies, 996 total patients) (Figure 1) and
for clinical remission of 14% (95%CI, 6% to 21%; n = 16
studies, 795 total patients) (Figure 2). The active rTMS also
resulted in a significant and medium-sized SMD of 0.48
(95%CI, 0.28 to 0.69; n = 21 studies, 899 total patients), com-
pared with sham conditions. The pooled response and remis-
sion rates were, respectively, 25% and 17% for active rTMS,
and 9% and 6% for sham conditions. The NNT for clinical
response was 6, and for remission was 7.
623
Review Paper

Table 1 Summary of clinical parameters of included studies

rTMS rTMS rTMS Duration, Sample
Study TRD definition placement frequency % MT weeks size Medications and comments

Anderson et al40 Failed 2 or more AD L DLPFC 10 110 4 29 Same medications

(79% of sample)
3 sessions per week
Avery et al41 Failed 2 or more AD, L DLPFC 10 110 3 68 Same medications
ATHF 3 or more
Berman et al42 Failed 1 or more AD, L DLPFC 20 80 2 20 Off all medications ´ 1 week
past or present
episode, 4 weeks
minimum
Boutros et al43 Failed 2 or more AD L DLPFC 20 80 2 22 Same medications
Bretlau et al33 Failed 1 or more AD, L DLPFC 8 90 3 45 All started on escitalopram
6 weeks minimum 20 mg
Fitzgerald et al44 Failed 2 or more AD; L DLPFC; L = 10; 100 2 60 Same medications
6 weeks minimum R DLPFC R=1
HFL, compared with LFR
Fitzgerald et al45 Failed 2 or more AD; R then L L = 10; L = 100; 2 50 Same medications
6 weeks minimum DLPFC R=1 R = 110
Sequential R then L
Garcia-Toro Failed 2 or more AD, L DLPFC 20 90 2 40 Same medications
et al32 6 weeks minimum
Garcia-Toro Failed 1 or more AD L DLPFC 20 90 2 28 All started on sertraline 50 mg
et al34 (57% of sample)
Also started on sertraline at
start of rTMS
Garcia-Toro Failed 2 or more AD L DLPFC; L = 20; 110 2 30 Same medications
et al46 R DLPFC R=1
HFL, compared with LFR
47
Kauffmann et al Failed 2 or more AD, R DLPFC 1 110 2 12 Same medications
8 weeks minimum LFR
48
Loo et al Failed 1 or more AD, Bilat 15 90 3 19 Same medications
adequate trial DLPFC
Bilateral L and R
49
Loo et al Failed 1 or more AD L DLPFC 10 110 2 40 Same medications
(55% of sample), but Excluded patients with typical
excluded if failed 2 or definition of TRD
more AD in current
episode
Manes et al50 Failed 1 or more AD, L DLPFC 20 80 1 20 Off all medications ´ 4 days
4 weeks minimum
Miniussi et al31 Failed 2 or more AD, at L DLPFC 17; 1 110 1 51 Same medications
least 2 classes HFL, compared with LFR
Moller et al51 Failed 1 or more AD L DLPFC 10 100 1 10 Same medications
36
Mogg et al Failed 2 or more AD L DLPFC 10 110 2 59 Same medications
(78% of sample)
Mosimann et al52 Failed 2 or more AD, L DLPFC 20 100 2 24 Same medications
adequate dose and
duration
O’Reardon et al39 Failed at least 1 but L DLPFC 10 120 4 301 No medications
less than 5 AD
Previous ECT was an exclusion
criterion

continued

624 W La Revue canadienne de psychiatrie, vol 53, no 9, septembre 2008

 Repetitive Transcranial Magnetic Stimulation for Treatment-Resistant Depression: A Systematic Review and Metaanalysis

Table 1 continued
rTMS RTMS RTMS Duration, Sample
Study TRD definition placement frequency % MT weeks size Medications and comments

Padberg et al53 Failed 2 or more AD, L DLPFC 10; 0.3 90 1 12 Same medications
including 1 TCA, HFL, compared with LFR
4 weeks minimum
Padberg et al54 Failed 2 or more AD L DLPFC 10 100 and 2 30 Same medications
90
Two MT conditions
Rossini et al35 Failed 2 or more AD L DLPFC 15 100 and 2 54 Same medications
with different 80 Two MT conditions
mechanisms,
6 weeks minimum
Stern et al37 Failed 1 or more AD L DLPFC; L = 10; 110 2 35 Off all medications ´ 14 days
R DLPFC R=1
LFR condition not included
55
Su et al Failed 2 or more AD, L DLPFC 20; 5 100 2 33 Same medications
6 weeks min
HFL, compared with LFR

ATHF = treatment history form; DLPFC = dorsolateral prefrontal cortex; L = left; MT = motor threshold; R = right; TCA = tricyclic antidepressant

Sensitivity analyses were done using only studies with a
stricter, but more usual, definition for TRD, namely, failure of
2 or more ADs within the current episode (Table 3). The risk
difference for active rTMS was still significantly higher for
clinical response at 16% (95%CI, 9% to 24%) and for clinical
remission at 16% (95%CI, 5% to 27%). Pooled response and
remission rates were, respectively, 24% and 17% for active
rTMS and 9% and 5% for sham. The NNT was 6 for both clini-
cal response and remission. The SMD was also significant at
0.37 (95%CI, 0.18 to 0.56). The sensitivity analyses of studies
using higher quality reporting methods showed similar find-
ings, as did studies using high-intensity, HFL stimulation
(Table 3).
Examination of funnel plots of risk differences and SMD
showed a single possible outlier for each outcome, but the
plots were reasonably symmetrical. There were no indications
of missing or unpublished negative studies, and only a sugges-
tion of missing or unpublished studies showing positive risk
difference effects for rTMS (data not shown).
There were very few withdrawals reported due to adverse
effects. Only 19 withdrawals occurred overall: 11 (2%) in
active rTMS conditions and 8 (1.5%) in sham conditions. The
number of dropouts for any reason were also very low (active
rTMS = 4%, n = 595; sham = 6%, n = 497), with a risk differ-
ence of –0.01 (95%CI, –0.04 to 0.01; n = 24 studies). The low
dropout rates, in part, may be due to the short durations of
treatment.
The 4 studies available for quantitative analysis that system-
atically followed all patients after the end of double-blind
treatment included only 1- to 2-week follow-up periods
(Table 2). The SMD was large at 1.04 (95%CI, 0.08 to 2.01;
n = 4 studies, 122 patients). However, there was significant
heterogeneity with one study by Stern et al37 potentially
skewing results with a large individual SMD (2.67). How-
ever, the SMD was still significant when the study was omit-
ted (0.58, 95%CI, 0.13 to 1.02). That study was also the only
one that reported on responder and remitter status at
follow-up; both rates were stable or increased at follow-up in
the active rTMS conditions, but not in the sham condition.37
Discussion
Previous systematic reviews of rTMS have focused on
weighted mean differences, SMDs or relative risks between
active and sham treatments, but absolute risk differences and
NNTs from categorical response and remission outcomes are
more relevant and understandable outcomes for clinicians.
Similarly, most systematic reviews pooled all patient groups
irrespective of treatment resistance, or included treatment
resistance as a subgroup analysis without examining this
important population in greater detail.
In our systematic review, studies were selected based on
explicit criteria for TRD. Additionally, a clinical response
measure (as typically defined, and which was used by all of
the studies in this review, as 50% or greater reduction in
HDRS or MADRS scores) was chosen as the primary out-
come, with clinical remission (end point scores within the

The Canadian Journal of Psychiatry, Vol 53, No 9, September 2008 W 625

Review Paper

Table 2 Studies with systematic, no-treatment follow-up after acute double-blind treatment
Study Sample size Period of treatment and follow-up Outcome measure(s) available

Bretlau et al33 45 3 weeks double-blind treatment HDRS end point scores

9 weeks follow-up Note that patients were also started on escitalopram
(20 mg) at beginning of rTMS and continued on the drug
during follow-up. Although the active rTMS group
maintained improvement over the first 5 weeks of follow-up,
by 9 weeks there was no difference in scores.
Garcia-Toro et al32 35 2 weeks double-blind treatment HDRS end point scores
2 weeks follow-up
Garcia-Toro et al34 22 2 weeks double-blind treatment HDRS end point scores
2 weeks follow-up Note that patients were also started on sertraline (50 mg)
at beginning of rTMS and continued on the drug during
follow-up.
Garcia-Toro et al46 37 2 weeks double-blind treatment HDRS end point scores
2 weeks follow-up
Manes et al50 20 5 days double-blind treatment HDRS end point scores
1 week follow-up
Mogg et al36 59 2 weeks double-blind treatment Only graphical presentation of mean HDRS scores. Scores
increased in the active rTMS group while scores decreased
6 weeks follow-up
in the sham group during the follow-up; at
6 weeks there was no difference in scores.

Rossini et al35 54 2 weeks double-blind treatment Only graphical presentation of mean HDRS scores. Scores
appeared to be stable, and authors commented that
3 weeks follow-up
responders showed sustained response, during follow-up,
but no data reported.

Stern et al37 35 2 weeks double-blind treatment HDRS end point scores and response and (or)
remission rates
2 weeks follow-up

Table 3 Sensitivity and subgroup analyses of rTMS study parameters

Clinical response Clinical remission
Parameter Risk difference (95%CI) Risk difference (95%CI) SMD (95%CI)

All studies 17% (10% to 23%)a 14% (6% to 21%)a 0.48 (0.28 to 0.69)a
n = 22 studies, 996 patients n = 16 studies, 795 patients n = 21 studies, 899 patients
Strict definition of TRD 16% (9% to 24%)a 16% (5% to 27%)a 0.37 (0.18 to 0.56)
n = 14 studies, 795 patients n = 10 studies, 641 patients n = 11 studies, 607 patients
Higher qualityb 15% (7% to 22%)a 11% (2% to 20%)a 0.39 (0.23 to 0.55)
n = 10 studies, 713 patients n = 8 studies, 623 patients n = 9 studies, 625 patients
High-intensity, LFR stimulation 19% (10% to 27%)a 16% (6% to 27%)a 0.48 (0.21 to 0.76)a
n = 15 studies, 782 patients n = 11 studies, 630 patients n = 13 studies, 648 patients
a
Significant heterogeneity of results (c2, P < 0.05; I2 > 30%)
b
Adequate reporting of allocation concealment, blinding, and adequacy of sham treatment

626 W La Revue canadienne de psychiatrie, vol 53, no 9, septembre 2008

 Repetitive Transcranial Magnetic Stimulation for Treatment-Resistant Depression: A Systematic Review and Metaanalysis

Figure 1 Metaanalysis results for clinical response showing risk difference between
active and sham rTMS in all studies

normal range) and SMDs as secondary outcomes. Our results
show that active rTMS is significantly superior to sham in
short-term acute treatment of TRD. The risk difference of
17% and the NNT of 6 are of clinically significant magnitude
for these difficult to treat patients, and similar to those seen
with medication-placebo comparisons in TRD. In addition,
the results were similar with sensitivity analyses of studies
using the stricter definition of TRD and high-intensity, HFL
stimulation. The rTMS treatments were well-tolerated with
very few dropouts owing to adverse effects.
Only 2 other metaanalyses of rTMS have specifically
addressed TRD, both within sensitivity analyses of mixed
sample studies. Herrmann and Ebmeier 38 examined
medication resistance in a larger systematic review of rTMS
trials. They included 18 trials (of which 15 were included in
our metaanalysis) with a treatment-resistant sample and
found a weighted ES (Cohen’s d) of 0.66 (95%CI, 0.47 to
0.86), not significantly different from the ES in trials without
treatment-resistant samples (Cohen’s d = 0.61). In contrast,
the SMD (ES) in end point scores in our review was slightly
lower at 0.48 (95%CI, 0.28 to 0.69). The discrepancy
between results cannot be further explored as there were no
details in the Herrmann and Ebmeier38 review on the TRD
studies included and their criteria for extracting data.
Schutter8 reviewed 25 studies using HFL, with an overall
weighted mean ES in change scores of 0.39. There was no
difference by ANOVA between medication-resistance

The Canadian Journal of Psychiatry, Vol 53, No 9, September 2008 W 627

Review Paper
Figure 2 Metaanalysis results for clinical remission showing risk difference between
active and sham rTMS in all studies
studies (n = 17), defined as greater than 2 failed medication
trials or history of failure to ECT, and nonmedication resis-
tance studies (n = 8). This is comparable to our sensitivity
analysis of high-intensity HFL studies (n = 13), in which the
SMD for end point scores was 0.48 (95%CI, 0.21 to 0.76).
Despite these significant findings, the included rTMS studies
have major limitations in methodology that temper the clinical
significance of the results. In particular, these studies had very
short treatment durations, with most studies using only 1 to 2
weeks. The pooled response and remission rates of 25% and
17%, respectively, are low, compared with most medication
studies of TRD. The clinically relevant risk difference of 17%
is owing to the fact that the placebo response rates (averaging
about 7.5%) are also very low. Although one can argue that
patients with TRD should be expected to have lower placebo
response, medication studies in TRD (which usually use lon-
ger treatment periods) have generally found higher placebo
response rates.
Several rTMS studies reported longer treatment periods with
higher response rates at the end of treatment. Unfortunately,
628
these studies used a nonstandard methodology in which
nonresponders at each week were dropped from the study and
only partial responders continued with blinded treatment.
The differential retention of patients in active and sham con-
ditions precludes using the data to examine efficacy. How-
ever, the results suggest that longer duration of rTMS (4 or
more weeks) may lead to higher clinical response. In this
regard, the results of the large-sample, 4-week study by
O’Reardon and colleagues,39 which included 301 patients
randomized to sham or active rTMS treatment in a
multicentre design, should be noted. In that study, there was
no statistically significant difference between active and
sham treatment in the primary outcome (change in MADRS
scores). However, among the secondary outcomes, the
HDRS response rates for active rTMS and placebo were
20.6% and 11.6% (P < 0.05), respectively, while remission
rates were 7.1% and 6.2% (not significant). These results
were very similar to the overall results from this systematic
review, suggesting that 4 weeks of treatment may not be
much better than 2.
W La Revue canadienne de psychiatrie, vol 53, no 9, septembre 2008
 Repetitive Transcranial Magnetic Stimulation for Treatment-Resistant Depression: A Systematic Review and Metaanalysis
Another important question, given that rTMS is
labour-intensive and time-consuming (because patients must
attend treatments at least several times each week), is whether
the AD effects are sustained after initial treatment. Unfortu-
nately, only 8 studies included some systematic follow-up
after the acute course of treatment, and most involved only
short-term follow-up of 1 to 3 weeks. The results from the 4
studies available for quantitative synthesis showed that AD
effects persisted for 1 to 2 weeks after discontinuation of
rTMS. Two studies had longer follow-up periods; one study
in which rTMs was started with 20 mg of escitalopram
showed maintenance of rTMS response, compared with
sham, through 5 weeks, but no difference at 9 weeks;33 the
other showed no difference between active and sham at 6
weeks.36 Other studies also included longer follow-up peri-
ods, but either followed responders only, or followed up after
unblinded rTMS treatment. Given these limitations, it is not
surprising that the results were inconsistent, with some studies
showing relapse while others showed enduring therapeutic
effects over several weeks and months of follow-up. No stud-
ies of maintenance rTMS treatment for TRD were identified.
To summarize the clinical findings of this systematic review,
rTMS with short treatment duration (1 to 4 weeks) has clear
AD effects in TRD and is well-tolerated, but the overall
response and remission rates are low and it is unclear whether
the effects are sustained. These results are similar to previous
systematic reviews of mixed populations of depressed
p a tien ts th a t in c lu d e d b o th tr e a tmen t- n a iv e a n d
treatment-resistant cases.2–8 Further studies in TRD are
needed to determine: whether longer duration rTMS treat-
ment (for example, 4 weeks or longer) results in higher
response and remission rates, compared with sham treat-
ments; whether the acute AD effects of rTMS are sustained
over time, even without ongoing use of rTMS; whether rTMS
speeds onset, and more importantly, results in greater overall
response, when used in combination with starting an AD; and
whether less frequent maintenance rTMS (for example, once
weekly or less often) can sustain the acute effects. Some of
these important questions must be answered before rTMS can
be considered a first-line, monotherapy treatment for TRD or
less refractory cases of depression.
Funding and Support
Dr Lam is on speaker and (or) advisory boards for, or has received
research grants from: ANS, Inc, AstraZeneca, Biovail, Canadian
Institutes of Health Research, Canadian Network for Mood and
Anxiety Treatments, Eli Lilly, GlaxoSmithKline, GreatWest Life,
Janssen, Litebook Company, Ltd, Lundbeck, Sanofi-Aventis,
Servier, VGH and UBC Hospital Foundation, and Wyeth.
Dr Chan is on speaker boards for: AstraZeneca, Eli Lilly, Janssen,
Lundbeck, and Organon. Dr Yatham is on speaker and (or)
advisory boards for, or has received research grants from:
AstraAeneca, Bristol Myers Squibb, Canadian Institutes of Health
Research, Canadian Network for Mood and Anxiety Treatments,
The Canadian Journal of Psychiatry, Vol 53, No 9, September 2008 W
Eli Lilly, GlaxoSmithKline, Janssen, Michael Smith Foundation
for Health Research, Pfizer, Servier, and Stanley Foundation.
References
1. Eitan R, Lerer B. Nonpharmacological, somatic treatments of depression:
electroconvulsive therapy and novel brain stimulation modalities. Dialogues
Clin Neurosci. 2006;8(2):241–258.
2. McNamara B, Ray JL, Arthurs OJ, et al. Transcranial magnetic stimulation for
depression and other psychiatric disorders. Psychol Med.
2001;31(7):1141–1146.
3. Holtzheimer PE, III, Russo J, Avery DH. A meta-analysis of repetitive
transcranial magnetic stimulation in the treatment of depression.
Psychopharmacol Bull. 2001;35(4):149–169.
4. Burt T, Lisanby SH, Sackeim HA. Neuropsychiatric applications of transcranial
magnetic stimulation: a meta analysis. Int J Neuropsychopharmacol.
2002;5(1):73–103.
5. Kozel FA, George MS. Meta-analysis of left prefrontal repetitive transcranial
magnetic stimulation (rTMS) to treat depression. J Psychiatr Pract.
2002;8(5):270 –275.
6. Martin JL, Barbanoj MJ, Schlaepfer TE, et al. Repetitive transcranial magnetic
stimulation for the treatment of depression. Systematic review and
meta-analysis. Br J Psychiatry. 2003;182:480 – 491.
7. Couturier JL. Efficacy of rapid-rate repetitive transcranial magnetic stimulation
in the treatment of depression: a systematic review and meta-analysis.
J Psychiatry Neurosci. 2005;30(2):83–90.
8. Schutter DJ. efficacy of high-frequency transcranial magnetic stimulation over
the left dorsolateral prefrontal cortex in double-blind sham-controlled designs: a
meta-analysis. Psychol Med. 2008 Apr 30; [Epub ahead of print].
9. Berlim MT, Turecki G. Definition, assessment, and staging of
treatment-resistant refractory major depression: a review of current concepts
and methods. Can J Psychiatry. 2007;52(1):46 –54.
10. Souery D, Papakostas GI, Trivedi MH. Treatment-resistant depression. J Clin
Psychiatry. 2006;67(Suppl 6):16–22.
11. Dunner DL, Rush AJ, Russell JM, et al. Prospective, long-term, multicenter
study of the naturalistic outcomes of patients with treatment-resistant
depression. J Clin Psychiatry. 2006;67(5):688–695.
12. Kennedy SH, Lam RW. Enhancing outcomes in the management of treatment
resistant depression: a focus on atypical antipsychotics. Bipolar Disord.
2003;5(Suppl 2):36–47.
13. Greden JF. The burden of disease for treatment-resistant depression. J Clin
Psychiatry. 2001;62(Suppl 16):26–31.
14. Nierenberg AA, Katz J, Fava M. A critical overview of the pharmacologic
management of treatment-resistant depression. Psychiatr Clin North Am.
2007;30(1):13–29.
15. Lam RW, Kennedy SH. Evidence-based strategies for achieving and sustaining
full remission in depression: focus on metaanalyses. Can J Psychiatry.
2004;49(3 Suppl 1):17S-26S.
16. McPherson S, Cairns P, Carlyle J, et al. The effectiveness of psychological
treatments for treatment-resistant depression: a systematic review. Acta
Psychiatr Scand. 2005;111(5):331–340.
17. Anonymous. Transcranial magnetic stimulation: potential new treatment for
resistant depression. J Clin Psychiatry. 2007;68:315–330.
18. Higgins J, Green S, editors. Cochrane Handbook for Systematic Reviews of
Interventions 4.2. Chichester (GB): Cochrane Library, Wiley; 2007.
19. Cochrane Collaboration. Review Manager (RevMan) Version 4.2 for Windows.
Copenhagen (DK): The Nordic Cochrane Centre; 2003.
20. Hunter JE, Schmidt FL. Fixed effects vs random effects meta-analysis models:
Implications for cumulative research knowledge. Int J Select Assess.
2000;8:275–292.
21. Higgins JP, Thompson SG, Deeks JJ, et al. Measuring inconsistency in
meta-analyses. BMJ. 2003;327(7414):557–560.
22. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat
Med. 2002;21(11):1539–1558.
23. Fitzgerald PB, Huntsman S, Gunewardene R, et al. A randomized trial of
low-frequency right-prefrontal-cortex transcranial magnetic stimulation as
augmentation in treatment-resistant major depression. Int J
Neuropsychopharmacol. 2006;9(6):655–666; [Epub 2006 Sep 7].
24. Turnier-Shea Y, Bruno R, Pridmore S. Daily and spaced treatment with
transcranial magnetic stimulation in major depression: a pilot study. Aust N Z J
Psychiatry. 2006;40(9):759–763.
25. Fitzgerald PB, Hoy K, McQueen S, et al. Priming stimulation enhances the
effectiveness of low-frequency right prefrontal cortex transcranial magnetic
stimulation in major depression. J Clin Psychopharmacol. 2008;28(1):52–58.
26. Loo C, Mitchell P, Sachdev P, et al. Double-blind controlled investigation of
transcranial magnetic stimulation for the treatment of resistant major
depression. Am J Psychiatry. 1999;156(6):946–948.
27. Bortolomasi M, Minelli A, Fuggetta G, et al. Long-lasting effects of high
frequency repetitive transcranial magnetic stimulation in major depressed
patients. Psychiatry Res. 2007;150(2):181–186; [Epub 2007 Feb 14].
629
Review Paper
28. Moser DJ, Jorge RE, Manes F, et al. Improved executive functioning following
repetitive transcranial magnetic stimulation. Neurology. 2002;58(8):1288–1290.
29. Avery DH, Holtzheimer PE III, Fawaz W, et al. Transcranial magnetic
stimulation reduces pain in patients with major depression: a sham-controlled
study. J Nerv Ment Dis. 2007;195(5):378–381.
30. Pascual-Leone A, Rubio B, Pallardo F, et al. Rapid-rate transcranial magnetic
stimulation of left dorsolateral prefrontal cortex in drug-resistant depression.
Lancet. 1996;348(9022):233–237.
31. Miniussi C, Bonato C, Bignotti S, et al. Repetitive transcranial magnetic
stimulation (rTMS) at high and low frequency: an efficacious therapy for major
drug-resistant depression? Clin Neurophysiol. 2005;116(5):1062–1071.
32. Garcia-Toro M, Mayol A, Arnillas H, et al. Modest adjunctive benefit with
transcranial magnetic stimulation in medication-resistant depression. J Affect
Disord. 2001;64(2–3):271–275.
33. Bretlau LG, Lunde M, Lindberg L, et al. Repetitive transcranial magnetic
stimulation (rTMS) in combination with escitalopram in patients with
treatment-resistant major depression: a double-blind, randomised,
sham-controlled trial. Pharmacopsychiatry. 2008;41(2):41–47.
34. Garcia-Toro M, Pascual-Leone A, Romera M, et al. Prefrontal repetitive
transcranial magnetic stimulation as add on treatment in depression. J Neurol
Neurosurg Psychiatry. 2001;71(4):546–548.
35. Rossini D, Lucca A, Zanardi R, et al. Transcranial magnetic stimulation in
treatment-resistant depressed patients: a double-blind, placebo-controlled trial.
Psychiatry Res. 2005;137(1–2):1–10; [Epub 2005 Oct 12].
36. Mogg A, Pluck G, Eranti SV, et al. A randomized controlled trial with 4-month
follow-up of adjunctive repetitive transcranial magnetic stimulation of the left
prefrontal cortex for depression. Psychol Med. 2008;38(3):323–333. [ Epub 2007
Oct 15].
37. Stern WM, Tormos JM, Press DZ, et al. effects of high and low frequency
repetitive transcranial magnetic stimulation to the dorsolateral prefrontal cortex:
a double-blind, randomized, placebo-controlled trial.
J Neuropsychiatry Clin Neurosci. 2007;19(2):179–186.
38. Herrmann LL, Ebmeier KP. Factors modifying the efficacy of transcranial
magnetic stimulation in the treatment of depression: a review. J Clin Psychiatry.
2006;67(12):1870–1876.
39. O’Reardon JP, Solvason HB, Janicak PG, et al. Efficacy and safety of
transcranial magnetic stimulation in the acute treatment of major depression:
a multisite randomized controlled trial. Biol Psychiatry. 2007;62(11):1208–1216;
[Epub 2007 Jun 14].
40. Anderson IM, Delvai NA, Ashim B, et al. Adjunctive fast repetitive transcranial
magnetic stimulation in depression. Br J Psychiatry. 2007;190:533–534.
41. Avery DH, Holtzheimer PE III, Fawaz W, et al. A controlled study of repetitive
transcranial magnetic stimulation in medication-resistant major depression. Biol
Psychiatry. 2006;59(2):187–194; [Epub 2005 Sep 1].
42. Berman RM, Narasimhan M, Sanacora G, et al. A randomized clinical trial of
repetitive transcranial magnetic stimulation in the treatment of major depression.
Biol Psychiatry. 2000;47(4):332–337.
43. Boutros NN, Gueorguieva R, Hoffman RE, et al. Lack of a therapeutic effect of a
2-week sub-threshold transcranial magnetic stimulation course for
treatment-resistant depression. Psychiatry Res. 2002;113(3):245–254.
44. Fitzgerald PB, Brown TL, Marston NA, et al. Transcranial magnetic stimulation
in the treatment of depression: a double-blind, placebo-controlled trial. Arch Gen
Psychiatry. 2003;60(10):1002–1008.
45. Fitzgerald PB, Benitez J, de Castella A, et al. A randomized, controlled trial of
sequential bilateral repetitive transcranial magnetic stimulation for
treatment-resistant depression. Am J Psychiatry. 2006;163(1):88–94.
630
46. Garcia-Toro M, Salva J, Daumal J, et al. High (20-Hz) and low (1-Hz)
frequency transcranial magnetic stimulation as adjuvant treatment in
medication-resistant depression. Psychiatry Res. 2006;146(1):53–57; [ Epub
2005 Dec 13].
47. Kauffmann CD, Cheema MA, Miller BE. Slow right prefrontal transcranial
magnetic stimulation as a treatment for medication-resistant depression:
a double-blind, placebo-controlled study. Depress Anxiety. 2004;19:59–62.
48. Loo CK, Mitchell PB, Croker VM, et al. Double-blind controlled investigation
of bilateral prefrontal transcranial magnetic stimulation for the treatment of
resistant major depression. Psychol Med. 2003;33(1):33–40.
49. Loo CK, Mitchell PB, McFarquhar TF, et al. A sham-controlled trial of the
efficacy and safety of twice-daily rTMS in major depression. Psychol Med.
2007;37(3):341–349. [Epub 2006 Dec 19].
50. Manes F, Jorge R, Morcuende M, et al. A controlled study of repetitive
transcranial magnetic stimulation as a treatment of depression in the elderly. Int
Psychogeriatr. 2001;13:225–231.
51. Moller AL, Hjaltason O, Ivarsson O, et al. The effects of repetitive transcranial
magnetic stimulation on depressive symptoms and the P(300) event-related
potential. Nord J Psychiatry. 2006;60:282–285.
52. Mosimann UP, Schmitt W, Greenberg BD, et al. Repetitive transcranial
magnetic stimulation: a putative add-on treatment for major depression in
elderly patients. Psychiatry Res. 2004;126(2):123–133.
53. Padberg F, Zwanzger P, Thoma H, et al. Repetitive transcranial magnetic
stimulation (rTMS) in pharmacotherapy-refractory major depression:
comparative study of fast, slow and sham rTMS. Psychiatry Res.
1999;88(3):163–171.
54. Padberg F, Zwanzger P, Keck M, et al. Repetitive transcranial magnetic
stimulation (rTMS) in major depression. Relation between efficacy and
stimulation intensity. Neuropsychopharmacology. 2002;27(4):638–645.
55. Su TP, Huang CC, Wei IH. Add-on rTMS for medication-resistant depression:
a randomized, double-blind, sham-controlled trial in Chinese patients. J Clin
Psychiatry. 2005;66(7):930–937.
Manuscript received September 2007, revised, and accepted January
2008.
1
Professor and Head, Division of Clinical Neuroscience, Department of
Psychiatry, University of British Columbia, Vancouver, British Columbia;
Medical Director, Mood Disorders Centre, UBC Hospital, Vancouver,
British Columbia.
2
Clinical Associate Professor, Department of Psychiatry, University of
British Columbia, Vancouver, British Columbia; Head, ECT Program,
Vancouver General Hospital, Vancouver, British Columbia.
3
Clinical Assistant Professor, Department of Psychiatry, University of
British Columbia, Vancouver, British Columbia; Physician Operations
Leader, Geriatric Psychiatry, Mount Saint Joseph Hospital, Vancouver,
British Columbia.
4
Professor and Associate Head, Research and International Affairs,
Department of Psychiatry, University of British Columbia, Vancouver,
British Columbia; Medical Director, Mood Disorders Clinical Research
Unit, UBC Hospital, Vancouver, British Columbia.
Address for correspondence: Dr RW Lam, Department of Psychiatry,
University of British Columbia, 2255 Wesbrook Mall, Vancouver, BC
V6T 2A1; r.lam@ubc.ca
W La Revue canadienne de psychiatrie, vol 53, no 9, septembre 2008
 Repetitive Transcranial Magnetic Stimulation for Treatment-Resistant Depression: A Systematic Review and Metaanalysis

Résumé : La stimulation magnétique transcranienne répétitive pour la dépression

réfractaire au traitement : un examen systématique et une méta-analyse
Objectif : Les examens systématiques indiquent que la stimulation magnétique transcranienne
répétitive (SMTr) est supérieure aux conditions de contrôle simulées chez les patients souffrant du
trouble dépressif majeur, mais la pertinence clinique n’est pas manifeste. Personne n’a examiné
spécifiquement les résultats chez les patients souffrant de dépression réfractaire au traitement
(DRT).
Méthode : Un examen systématique a été mené en identifiant les essais randomisés contrôlés
publiés sur la SMTr active, comparativement à une condition de contrôle simulée chez les patients
souffrant de DRT définie (c’est-à-dire, au moins un essai échoué d’antidépresseur). Le principal
résultat était la réponse clinique telle qu’elle était déterminée par les notes globales, ou 50 % ou la
plus grande amélioration à une échelle de cotation. Les autres résultats comprenaient la rémission et
les différences moyennes normalisées des scores finals. Une méta-analyse a été menée pour les
différences de risque absolu, à l’aide des modèles d’effets aléatoires. Des analyses de sensibilité et
de sous-groupes ont aussi été menées afin d’explorer l’hétérogénéité et la robustesse des résultats.
Résultats : Un total de 24 études (n = 1 092 patients) satisfaisaient aux critères de synthèse
quantitative. La SMTr active était significativement supérieure aux conditions simulées pour
produire une réponse clinique, avec une différence de risque de 17 % et un nombre de sujets
nécessaire de 6. Les taux regroupés de réponse et de rémission étaient de 25 % et de 17 %, et de
9 % et de 6 % pour la SMTr active et les conditions simulées, respectivement. Les analyses de
sensibilité et de sous-groupes n’ont pas affecté significativement ces résultats. Les abandons et les
retraits attribuables aux effets indésirables étaient très faibles.
Conclusions : Pour les patients souffrant de DRT, la SMTr semble procurer des avantages
significatifs dans les études de traitement à court terme. Cependant, les taux relativement faibles de
réponse et de rémission, les courtes durées du traitement, et le manque relatif d’études de suivi
systématique suggèrent que d’autres recherches sont nécessaires avant de pouvoir considérer que la
SMTr est une monothérapie de première ligne pour la DRT.

The Canadian Journal of Psychiatry, Vol 53, No 9, September 2008 W 631