Commentary

Transcranial magnetic stimulation (TMS) is a safe non-invasive method to focally stimulate the brain. A hand-held coil applied to a selected scalp region delivers intense but ultra-brief magnetic pulses that in turn induce secondary electrical activation of neurons within the immediately underlying cortex. Trains of repeated “fast” (>1 Hz) TMS have an activating effect whereas slow (<1 Hz) TMS is inhibitory; thousands of such stimuli can be administered daily within a 30 min session for several weeks. Most trials of TMS for depression have focused on activating the left dorsolateral prefrontal cortex (DLPFC) while a few have used TMS to downregulate the right DLPFC and some have combined both approaches, all in an attempt to normalise prefrontal function.

We are now well into the second decade of research on TMS as a treatment for depression with meta-analyses of randomised sham controlled double blind (patients and raters) trials being published on an almost annual basis since 2001. The two reviews by Schutter, and Lam and colleagues, provide the most recent meta-analyses and both include data from the largest (301 patients) multicentred randomised controlled trial to date that was industry sponsored and involved 4 weeks of intensive week-daily TMS.1 Schutter has analysed 30 trials (1164 patients) reported between 1997 and 2008 whereas Lam and colleagues concentrated on 24 studies (overlap of 14 studies) involving 1092 patients designated treatment resistant; the former used continuous depression rating scale scores for the primary outcome and the latter used predefined dichotomous measures of response status.

Despite methodological differences between the two reviews, the results are similar and indeed do not substantially differ from most previous meta-analyses. Namely, real TMS is statistically significantly superior to sham TMS with small to medium effect sizes that are similar to that previously reported for some antidepressant drugs. To place their findings in a clinical context, Lam and colleagues estimated a number needed to treat (NNT) of 6 to obtain a response to TMS. Another recent meta-analysis, including many of the same studies but only those with ⩾10 subjects per group (22 studies, 1107 patients), reported an NNT of 4 (95% CI 3 to 6).2 Additionally, these recent reviews all agree that repeated TMS is a medically safe and well tolerated procedure with few adverse effects.

However, some caution is warranted before TMS could be recommended as a routine treatment for depression. Optimal treatment parameters for TMS have still not yet been systematically characterised and most of the studies included in these meta-analyses are of short duration with only a few having any appreciable systematic follow-up.

Interestingly, Schutter found that neither treatment resistance nor stimulus intensity contributed to the antidepresssant effect of TMS, while Lam and colleagues noted that 4 weeks of TMS, as in the above large trial,1 did not appear to be better than 2 weeks of treatment. This raises the possibility that the modest benefits found with TMS under rigorous trial conditions are non-specific and may be due to placebo effects. This may be compounded by the obviously physical nature of TMS leading to well recognised problems with employing a suitable “sham” coil, which itself may cause some brain stimulation, and how this may affect the integrity of blinding of both patients and raters. These particular methodological issues need to be fully resolved to improve the quality of any future trials.

TMS is labour intensive and time consuming, and compared with the other well established brain stimulation method, electroconvulsive therapy, is unlikely to be cost effective, mainly due to its relatively low effectiveness and subsequent incurred costs.3 Although TMS has not been approved by the National Institute for Health and Clinical Excellence (NICE) in the UK for depression, it has recently been approved by the Food and Drug Administration (FDA) in the USA for patients who have failed to respond to a single antidepressant drug. It remains to be seen whether these recent meta-analyses will change the opinion of NICE.