Closing the Loop on Deep Brain Stimulation for Treatment-Resistant Depression

Abstract

Major depressive episodes are the largest cause of psychiatric disability, and can often resist treatment with medication and psychotherapy. Advances in the understanding of the neural circuit basis of depression, combined with the success of deep brain stimulation (DBS) in movement disorders, spurred several groups to test DBS for treatment-resistant depression. Multiple brain sites have now been stimulated in open-label and blinded studies. Initial open-label results were dramatic, but follow-on controlled/blinded clinical trials produced inconsistent results, with both successes and failures to meet endpoints. Data from follow-on studies suggest that this is because DBS in these trials was not targeted to achieve physiologic responses. We review these results within a technology-lifecycle framework, in which these early trial "failures" are a natural consequence of over-enthusiasm for an immature technology. That framework predicts that from this "valley of disillusionment," DBS may be nearing a "slope of enlightenment." Specifically, by combining recent mechanistic insights and the maturing technology of brain-computer interfaces (BCI), the next generation of trials will be better able to target pathophysiology. Key to that will be the development of closed-loop systems that semi-autonomously alter stimulation strategies based on a patient's individual phenotype. Such next-generation DBS approaches hold great promise for improving psychiatric care.

Keywords: brain circuits; brain-computer interfaces; deep brain stimulation; depression; electrophysiology; neuro-imaging; neuromodulation.

Figure 1

Conceptual diagram of technology maturation, originally proposed by the Gartner Group as the “hype cycle.” From a technology's inception (A), early positive results propel it to a “peak of inflated expectations” (B), where interest in the technology exceeds its maturity or its creator's understanding. This leads into a “valley of disillusionment” (C), where early adopters become discouraged and may abandon the technology. A slower phase of progress climbs a “slope of enlightenment” (D), where the lack of intense public attention enables more thoughtful development of robust solutions. This leads ultimately to a “plateau of productivity” (E) and readiness for wide adoption. We believe that recent studies have moved psychiatric DBS to the cusp between (C,D).

Figure 2

Schematic illustrations, re-used with permission, of the three DBS targets that have undergone human trials for treatment-resistant depression. (A) Subgenual cingulate gyrus (Cg25). (B) Ventral internal capsule/ventral striatum (VC/VS), also sometimes called anterior limb of internal capsule (ALIC). (C) Supero-lateral medial forebrain bundle (MFB).

Figure 3

Circuit model of MDD proposed by Mayberg (2009), adapted and simplified to emphasize DBS effects. Boxes denote clusters of structures with evidence for tight anatomic inter-connectivity and relevance to a sub-domain of impairment within MDD. Arrows denote long-range interactions believed to exist between these subnetworks. Major DBS targets are highlighted with thickened outlines. The three extant targets all project to a broad prefrontal network, and all are structures believed to generate more “primitive” internal experiences. CC, cingulate cortex; Hpc, hippocampus; MFB, medial forebrain bundle; OFC, orbitofrontal cortex; ParC, parietal cortex; PFC, prefrontal cortex; VTA, ventral tegmental area.