The definitions of treatment-resistant bipolar disorder (TRBD) have varied across studies. Additionally, its management is clinically challenging. An updated synthesis and appraisal of the available evidence is needed.
A systematic search of major electronic databases from inception up to May 25th, 2020, was conducted to identify randomized controlled trials (RCTs) of pharmacological and non-pharmacological interventions for the management of TRBD. When sufficient evidence was available, a meta-analysis was conducted.
Seventeen studies (n = 928 patients) were included in the qualitative synthesis. Fourteen studies (n = 803) assessed treatment-resistant acute bipolar depression (TRBD-De), including five neuromodulatory and nine pharmacological trials. Rapid- vs. standard up-titration of clozapine showed promising efficacy for TRBD mania, without significant adverse events. Electroconvulsive therapy (ECT) was confirmed to be similarly effective for TRBD-De as for treatment-resistant unipolar depression: odd ratio, OR = 0.919 (95%C.I. = 0.44-1.917), I = 13.98, p = .822. TRBD-De patients exposed to ketamine at day one post-infusion had high odds of response: OR = 10.682 (95%C.I. = 2.142-53.272), I = 0, p = <.005. The pooled drop-out rate in the ketamine trials was 21.2%. Additional evidence is warranted to confirm the potential efficacy of pramipexole or stimulants for TRBD-De.
Publication/measurement bias; exploratory nature of the meta-analyses for interventions that included participants solely with TRBD-De.
Overall, a few interventions are available for TRBD, including pramipexole, ECT, and clozapine, among others. Larger and better-designed trials for TRBD are warranted and should be based on more uniform operational definitions. PROSPERO registration number: CRD42018114567.

Copyright © 2020 Elsevier B.V. All rights reserved.

Author