Mechanisms Underlying Disorders of Consciousness: Bridging Gaps to Move Toward an Integrated Translational Science

Abstract

Aim: In order to successfully detect, classify, prognosticate, and develop targeted therapies for patients with disorders of consciousness (DOC), it is crucial to improve our mechanistic understanding of how severe brain injuries result in these disorders.

Methods: To address this need, the Curing Coma Campaign convened a Mechanisms Sub-Group of the Coma Science Work Group (CSWG), aiming to identify the most pressing knowledge gaps and the most promising approaches to bridge them.

Results: We identified a key conceptual gap in the need to differentiate the neural mechanisms of consciousness per se, from those underpinning connectedness to the environment and behavioral responsiveness. Further, we characterised three fundamental gaps in DOC research: (1) a lack of mechanistic integration between structural brain damage and abnormal brain function in DOC; (2) a lack of translational bridges between micro- and macro-scale neural phenomena; and (3) an incomplete exploration of possible synergies between data-driven and theory-driven approaches.

Conclusion: In this white paper, we discuss research priorities that would enable us to begin to close these knowledge gaps. We propose that a fundamental step towards this goal will be to combine translational, multi-scale, and multimodal data, with new biomarkers, theory-driven approaches, and computational models, to produce an integrated account of neural mechanisms in DOC. Importantly, we envision that reciprocal interaction between domains will establish a "virtuous cycle," leading towards a critical vantage point of integrated knowledge that will enable the advancement of the scientific understanding of DOC and consequently, an improvement of clinical practice.

Keywords: Brain injury; Coma; Consciousness; Electroencephalography; Magnetic resonance imaging; Mechanism; Neuroimaging.

Fig. 1

Overview of white paper recommendations. In this article, we have subdivided the gaps that exist in the field of disorders of consciousness (DOC) research into subdisciplines while stressing their mutual interdependence. The term “subdiscipline” is used for each branch of knowledge that makes up the study of DOC. Specifically, we suggest that efforts should be made to integrate structural and functional correlates, micro- and macroscale phenomena, and data- and theory-driven perspectives. Within each discipline (e.g., structural correlates), specific gaps should be identified and novel methods should be selected to answer these gaps and to reach an improved state of the science. Throughout this process, iterative integration with other disciplines is desired (bottom; note disciplines “A” and “B” can be replaced by any given subdiscipline of DOC research). Collectively, improved integration between these subfields of DOC is likely to provide the best avenue toward the clinical goals of DOC science: improved diagnosis, prognosis, and treatment (center circle). Circular arrows represent iterative processes, whereas two-headed arrows represent bidirectionality, e.g., improved diagnosis is likely to allow for more fine-tuned structural and functional correlates of DOC and vice versa

Fig. 2

Putative relationships between consciousness, environmental connectedness, and responsiveness (C-EC-R). Illustrative examples are shown pertaining to sleep (top ellipse), general anesthesia (bottom ellipse), and disorders of consciousness (middle ellipse). Note that this is not an exhaustive mapping of all possible states of altered consciousness; likewise, this framework does not directly address the question of quantifying residual cognitive function, as this can only be properly assessed in responsive patients. Also note that the relative size of the colored circles is not intended to reflect relative prevalence

Fig. 3

Connectivity in the human brain. a Functional connectivity can be quantified from functional neuroimaging, for example, as the Pearson correlation between regional blood-oxygen-level-dependent time series from functional magnetic resonance imaging (MRI). b Structural connectivity can be quantified from structural imaging, for example, as the number of streamlines between regions, estimated by using diffusion MRI. c Network analysis can provide information about individual nodes (e.g., identification of high-degree nodes, or “hubs”) as well as mesoscopic properties (e.g., modular organization) and macroscale (e.g., average length of shortest path between nodes)

Fig. 4

Overview of whole-brain computational modeling to integrate multimodal and multiscale data. a Whole-brain models combine a model of local regional activity with information about connectivity between regions. Additional information can be provided about neuroanatomy (from structural magnetic resonance imaging [MRI]), brain function (from functional MRI), and neurobiology (e.g., receptor density distribution obtained from in vivo positron-emission tomography [PET]). b Models can be systematically perturbed at different spatial and temporal scales, intervening at the level of individual regions or their connections

Fig. 5

Conceptual overview of levels of analysis to be considered in disorders of consciousness (DOC) research across the microscopic-to-macroscopic spectrum. Gradients indicate the capability of a technique to make measurements relevant to the level indicated above, thus highlighting gaps and possible translational interfaces. Human neuroimaging has produced macroscopic network biomarkers and certain regions/layers whose disruption is associated with DOC. For inquiries at more microscopic scale, animal models are indispensable, in which experimental manipulations (DREADD, optogenetics, lesion approaches, etc.) allow for direct mechanistic investigations, which can produce insights that can in turn be tested in humans in vivo (e.g., by using pharmacological approaches). The wider usage of high-field neuroimaging in both humans and animals will produce particularly relevant integrations of these levels, which will also serve to produce the type of data required to enable the generation of truly mechanistic computational approaches (e.g., whole-brain modeling). Altogether, these levels of analyses and models are complementary and synergistic for the discovery of the biological mechanisms of DOC