Noninvasive Brain Stimulation for Depression - The Devil Is in the Dosing.

In this issue of the Journal, Brunoni and colleagues report on transcranial direct-current stimulation (tDCS) in the treatment of major depression.1 This technique, which delivers weak electrical direct current to the scalp to modulate brain function, is one of a growing number of noninvasive brain stimulation interventions that change brain function and offer an opportunity to study brain–behavior relationships in health and disease. These tools have the potential to translate knowledge of the neural circuitry of the brain and patterns of electrical activity within those circuits (known as neural oscillatory dynamics) into treatments for psychiatric and neurologic disorders. Other related technologies that alter electrical brain activity include transcranial magnetic stimulation, transcranial alternatingcurrent stimulation, electroconvulsive therapy, deep-brain stimulation, and focused ultrasound. Of these, only transcranial magnetic stimulation and electroconvulsive therapy are currently cleared by the Food and Drug Administration for the treatment of depression, and deep-brain stimulation has a humanitarian device exemption for the treatment of obsessive–compulsive disorder. In contrast to the invasive nature of deepbrain stimulation, the noninvasive nature of transcranial magnetic stimulation, transcranial alternating-current stimulation, electroconvulsive therapy, and focused ultrasound makes them appealing for studying healthy volunteers with the goal of illuminating basic brain functions and testing hypotheses regarding the neural basis of psychiatric disorders. Unlike functional imaging that passively measures brain function, noninvasive brain stimulation actively changes neural function and can establish likely causal relationships between electrical patterns and mental dysfunction. Understanding these patterns is essential to the design and implementation of noninvasive electrical therapies. The antidepressant efficacy of tDCS has been an active area of study, and a recent meta-analysis has provided support for this use.2 A previous trial conducted by Brunoni and colleagues showed that tDCS had similar efficacy to sertraline in the treatment of depression, but that trial was not powered to test whether tDCS was noninferior to the medication.3 The present trial provides such data. As in the previous trial, the present trial used a double-dummy design, in which patients received the antidepressant escitalopram or placebo and tDCS or sham tDCS in order to mask the group assignment. The tDCS group and the escitalopram group both had efficacy that was greater than that observed in the group that received sham tDCS and placebo. However, a test of the noninferiority of tDCS to escitalopram failed, so the investigators could not draw a firm conclusion that tDCS was as effective as the medication. The trial has a number of limitations; most importantly, the patients in the medication group became aware of their assigned therapy presumably as a result of the side effects of the medication, which may have inflated efficacy and invalidated the noninferiority comparison. Patients in the tDCS group were unaware of the treatment they received; thus, the reported rates of response (41% in the tDCS group vs. 22% in the placebo group) and remission (24% vs. 13%) in depression can be taken as estimates of the antidepressant efficacy of tDCS. Furthermore, the use in this trial of transcranial magnetic stimulation as a probe of inhibitory tone in the brain points to this tone as an interesting potential predictor of response to tDCS (details are provided in the Supplementary Appendix of the article by Brunoni et al., available at NEJM.org). Another limitation, one shared by most other trials of noninvasive brain stimulation, is the lack of a measure of target engagement. The experimental-therapeutics approach that has been adopted by the National Institute of Mental Health is meant to ensure that when clinical trials fail to meet efficacy end points, they are at least informative regarding the reasons for failure. For example, treatment with noninvasive brain stimulation may be ineffective because the interven-