Cart before the horse: The need to establish rTMS efficacy before positioning it

In a recent editorial published in the Australian and New Zealand Journal of Psychiatry, Malhi et al. (2021) drastically discuss the positioning of repetitive transcranial magnetic stimulation (rTMS) in the guidelines for the treatment and management of depression. One of the critical points raised by the authors is that rTMS has been promoted as a treatment for depression because of its novel mechanism of action rather than its efficacy. Malhi et al. (2021) conclude that it is not possible to meaningfully position rTMS in the sequence care for the routine treatment of depression because of the lack of robust data regarding its use beyond the acute phase of the illness. We totally agree with this conclusion and we would like to present some complementary arguments to support this position. Indeed, we strongly believe that caution is warranted in accepting rTMS as an established treatment for depression and treatment-resistant depression (TRD).

The evidence base for rTMS clinical efficacy in depression remains under debate. While the Food and Drug Administration (FDA) approved its use as a treatment for depression in 2008 (guidance revised in 2011), the National Institute for Health and Care Excellence (NICE) adopted more nuanced positions arguing that ‘the evidence on its efficacy in the short-term is adequate, although the clinical response is variable’. In several countries such as France, the use of rTMS in clinical practice is still not recognized by health authorities.

In this context, we have recently showed evidence of a spurious excess of statistically significant results in the rTMS literature for neuropsychiatric disorders, including depression (Amad et al., 2019a). To test for an excess number of statistically significant results in the literature on rTMS therapeutic efficacy, we reviewed a wide range of meta-analyses and we characterized the power of studies included in these meta-analyses. Based on power calculations, we computed the expected number of ‘positive’ datasets for a medium effect size (standardized mean difference = 0.30) and compared it with the number of observed ‘positive’ datasets. We found evidence for an excess of significant findings overall and in the majority of meta-analyses, including the most important meta-analysis on depression (Brunoni et al., 2017). We also showed that most studies in the rTMS literature were underpowered. Excess significance may result from a multitude of factors such as lack of adequate pre-registration, large flexibility in the analytical plan, poor statistical methods and reporting and/or publication bias. For example, selective outcome reporting generally encourages false-positive findings and often disturbs the balance of findings in favor of ‘positive’ ones, giving the impression of some very large benefits.

Specifically regarding depression and TRD, the supposed efficacy of the conventional rTMS protocol (high-frequency rTMS delivered to the left dorsolateral prefrontal cortex, that is, HF-rTMS) is based on several meta-analyses (e.g. Brunoni et al., 2017) that included about 50 small sample size randomized controlled trials (RCTs; median number of subjects per study = 32.5, [interquartile range, 24–57.3]). Interestingly, three of the four largest studies failed to show the superiority of active versus sham HF-rTMS in patients with depression or TRD when basic methodological prerequisites are respected, for example, when the primary outcome and the intention-to-treat analyses are considered (for details see Amad et al., 2019b).

In conclusion, we think that the current state of knowledge does not support the recommendation of rTMS as a standard treatment for depression. We are convinced that improperly promoting rTMS for depression carries the risk of potentially disproportionate hope in patients and dilemmas for clinicians. We also want to insist that rTMS should under no circumstances be considered as a substitute for electroconvulsive therapy (ECT). Because of the ‘physical’ nature of the treatment, rTMS has initially been thought as a nonconvulsive alternative for ECT. However, the efficacy of ECT is superior to that of rTMS for depression (Mutz et al., 2019), and ECT is the preferred treatment for severe, melancholic or psychotic depression, especially in urgent indications. That’s why we suggest that efforts should focus on the understanding of the mechanism of action of ECT and reducing its adverse effects. Until future RCTs, including a placebo arm, in large samples definitely validate the efficacy of rTMS in depression, we think that rTMS should only been seen as an option in research protocols or when a good tolerance profile is needed in specific complex neuropsychiatric situations.