Pre-operative predictors of response to deep brain stimulation in depression: A systematic review and meta-analysis

Deep Brain Stimulation (DBS) for treatment-resistant depression (TRD) results in about 50 % response and 30 % remission. Response prediction may optimize patient selection, which is currently based on expert opinion. Here, we provide a comprehensive and systematic overview of the available data and explore if expert opinion is supported quantitatively. PubMed, Cochrane, Embase and PsycInfo were searched and studies reporting on DBS for TRD were included if they 1) reported on pre-operative predictors of response or 2) compared responders vs. non-responders. Meta-analysis was performed on outcomes with k ≥ 4 studies reporting data separately for responder/non-responder groups. From 7766 screened references, we included 22 studies on 294 unique patients. Seven studies were included for meta-analysis. No significant responder/non-responder differences were found in meta-analysis, but a trend toward shorter duration of current episode in responders (-0.97 years, p = 0.076, k = 6, n = 105) was found. Neuroimaging studies showed relative hyperactivity of the SCC area in responders (k = 2, n = 41), without volumetric differences (k = 2, n = 85). Multiple other structural, metabolic and (neuro)psychological predictors were found in small samples and have not yet been replicated. We conclude that there is no quantitative support for any of the response predictors based on expert opinion, although depressive chronicity and prior treatment responsiveness may be of interest for future research. For clinical practice, this suggests that patients should not be excluded based on the reviewed characteristics. More than a dozen predictors across structural, metabolic and (neuro)psychological outcomes are presented, awaiting replication and prospective validation in future trials.