Clinical efficacy and safety of repetitive transcranial magnetic stimulation in acute bipolar depression

Though bipolar disorder is characterized by episodes of mania/hypomania, depressive episodes pose the most burden for patients suffering from the disorder. Regrettably, few proven treatments exist for bipolar depression, and many patients either do not respond to, or have difficulty tolerating these treatments. Hence, novel, safe and effective treatments are urgently needed. The neuromodulatory approaches, such as repetitive transcranial magnetic stimulation (rTMS), have been demonstrated to be efficacious in randomized double-blind sham-controlled trials (RCTs) in treating depressive episodes in patients with major depressive disorder. However, it is unclear whether the antidepressant efficacy of rTMS extends to bipolar depression. Many RCTs of rTMS in major depression have included patients with bipolar depression. Therefore, our objective was to systematically review the rTMS literature to identify bipolar patients included in randomized trials in order to synthesize the data on clinical efficacy and safety of rTMS in bipolar depression. We registered the literature review protocol with PROSPERO (CRD#42015017089), which involved considering systematic reviews of rTMS in major depression and searching English-language publications in MEDLINE, EMBASE, and CENTRAL until July 11, 2015. We included randomized, double-blind, sham-controlled trials of rTMS involving ≥5 sessions that randomized patients with bipolar depression to both active and sham rTMS arms. We excluded RCTs that did not include patients with bipolar disorder, and those for which rates of clinical response were not reported or could not be obtained in correspondence with the investigators. We synthesized the data using Comprehensive Meta-Analyses Version 2.0 (Biostat, Englewood, NJ, USA). We analyzed intention to treat data with random effects models. Efficacy was investigated by risk difference (RD) and the number needed to treat (NNT). Supporting materials, including detailed methods, tables and figures are available by contacting the authors (alexander.mcgirr@alumni.ubc.ca). In total, we retained 19 RCTs in our meta-analysis1-19, totaling 181 patients with bipolar disorder (type I, N=40; type II, N=20; unspecified, N=121). The RCTs employed different stimulation targets: the left dorsolateral prefrontal cortex (DLPFC)1-6, 9-11, 13, 16, 17, the right DLPFC8, 14, 15, 18, or bilateral DLPFC7, 12, 17, 19. The majority of studies delivered high-frequency stimulation (HFS)1, 3-6, 9-13, 16, 18, while some delivered low-frequency stimulation (LFS)3, 8, 9, 15, 18, sequential LFS and HFS7, 17, 19, or theta burst stimulation (TBS)2, 14, 17. Significantly more patients receiving active rTMS achieved clinical response at study end compared to patients receiving sham rTMS (47/106, 44.3%, vs. 19/75, 25.3%; RD=0.18, 95% CI: 0.06-0.30, p<0.01). This represents a NNT of 6 (95% CI: 4-15). The fail-safe N was 29, suggesting that 29 missing or null studies are required to render this finding not statistically significant. Examination of the funnel plot revealed an asymmetrical distribution, with substantial loading at RD=0. Despite considerable methodological heterogeneity, there was no statistical evidence of heterogeneity (Q=19.99, df=22, I2=0.00, p=0.58; Egger's intercept =-0.36, t(21)=0.42, p=0.67). The optimal stimulation target and parameters are important considerations in rTMS due to differing physiological effects. We observed a trend towards differential target efficacy (Q=5.72, df=2, p=0.057). Indeed, RCTs targeting the right DLPFC demonstrated superior efficacy, with 9/15 (60.0%) of active rTMS patients achieving clinical response compared to 1/15 (6.6%) of sham rTMS patients. This represents a RD of 0.48 (95% CI: 0.17-0.78, p<0.001) and a NNT of 3 (95% CI: 2-6). RCTs targeting the left DLPFC also separated from placebo, with 33/68 (48.5%) of patients receiving active rTMS achieving clinical response compared to 15/50 (30.0%) of sham-treated patients (RD=0.16, 95% CI: 0.00-0.31, p<0.05), for a NNT of 7 (95% CI: 4-112). We did not observe separation between active and sham rTMS in RCTs employing bilateral stimulation (5/23, 21.73% vs. 3/14, 21.42%, p=0.68). We did not observe differential efficacy based on stimulation parameters. The issue of treatment-emergent affective switches in managing bipolar depression is important and controversial, and extends to neuromodulatory treatments. We observed a very low rate of treatment-emergent affective switches, and we did not observe an increased risk associated with active rTMS (1/106, 0.9% vs. 1/75, 1.3%, p=0.97). Though preliminary in nature, our analyses suggest that rTMS may be a safe and efficacious treatment option for acute bipolar depression. The degree of efficacy appears, on the surface, to be comparable to that observed among patients with major depressive disorder. Indeed, an overall NNT of 6 for clinical response is comparable to the NNTs reported in meta-analyses of rTMS in that disorder. Protocols targeting the right DLPFC with inhibitory LFS or TBS may be particularly efficacious; however, this is based on a small number of trials, influenced by a low sham-response rate, and requires additional investigation. Unfortunately, two RCTs dedicated to bipolar depression, with a total of 25 patients, could not be included, as clinical response and/or treatment protocols were unavailable in published form or through correspondence with investigators. Other biases include methodological heterogeneity between RCTs, and the limited number of RCTs and patients. Moreover, obtaining bipolar depressed data relied in large part on correspondence with investigators, and any bias related to successfully accessing this data remains. This is, to our knowledge, the first meta-analysis of RCTs of rTMS in the treatment of acute bipolar depression. We capitalized on the inclusion of patients with bipolar disorder in sham-controlled RCTs of rTMS in the treatment of major depression to identify 181 patients. Our analyses suggest that rTMS may be efficacious and safe in the treatment of acute bipolar depression and does not appear to be associated with treatment-emergent affective switches. Further, large sham-controlled RCTs are needed in bipolar depression to confirm the efficacy of rTMS. Alexander McGirr1, Sneha Karmani2, Rashmi Arsappa2, Marcelo T. Berlim3,4, Jagadisha Thirthalli2, Kesavan Muralidharan2, Lakshmi N. Yatham1,5 1Department of Psychiatry, University of British Columbia, Vancouver, BC, Canada; 2Department of Psychiatry, National Institute of Mental Health and Neurosciences, Bangalore, India; 3Neuromodulation Research Clinic, Douglas Mental Health University Institute and McGill University, Montréal, Québec, Canada; 4Depressive Disorders Program, Douglas Mental Health University Institute and McGill University, Montréal, Québec, Canada; 5Mood Disorders Centre of Excellence, University of British Columbia, Vancouver, BC, Canada The first two authors contributed equally to this work.