Coma and Disorders of Consciousness [3 ed.] 303150562X, 9783031505621

Citation preview

Coma and Disorders of Consciousness Caroline Schnakers Steven Laureys Editors Third Edition

123

Coma and Disorders of Consciousness

Caroline Schnakers  •  Steven Laureys Editors

Coma and Disorders of Consciousness Third Edition

Editors Caroline Schnakers Research Institute Casa Colina Hospital and Centers for Healthcare Pomona, CA, USA

Steven Laureys Coma Science Group GIGA-Consciousness University of Liège Liège, Belgium CERVO Brain Research Center Laval University Québec, QC, Canada

ISBN 978-3-031-50562-1    ISBN 978-3-031-50563-8 (eBook) https://doi.org/10.1007/978-3-031-50563-8 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2023, 2012, 2018 Springer Nature Switzerland AG This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland Paper in this product is recyclable

To medical teams and families we see every day and who inspire us.

Foreword

It is with great pleasure that I write this foreword for the third edition Coma and Disorders of Consciousness edited by Caroline Schnakers and Steven Laureys. The third edition has been both revised and updated with the inclusion of some new areas of focus as well as new authors. The last edition of this important text on Disorders of Consciousness (DoC) was in 2018. In the last 5 years, we have certainly seen quite a burgeoning of research and expansion of our collective knowledge base regarding DoC that serves as fertile soil for this third edition. As our scientific evidence-base grows across different aspects of assessment and management of DoC, we will be better able to drive care based on science as opposed to historical dogma and/or practitioner experience. Although DoC controversies still persist both neuroscientifically and neurorehabilitatively, this volume by Schnakers and Laureys addresses such a diversity of topics that anyone involved with interfacing at any level with this patient population should have this book on their shelf. Important topics that are covered include updates on neural correlates of consciousness, behavioral assessment including challenges of bedside assessment and neurologic recovery patterns, BCIs and their application to patients with DOC, as well as prognostication and long-term patterns of recovery as well as their medical, ethical, and legal implications. Neuromedical comorbidities across different states of disordered consciousness as well as management of same is an essential chapter relative to implications for decreasing morbidity and mortality in this patient population. The longstanding and controversial topic of sensory stimulation programs is addressed with an emphasis on newer data examining the potential efficacy of music therapy in this context. Chapters dealing specifically with treatment interventions involving pharmacotherapy as well as neuromodulation provide readers with up-to-date evidence-based information as well as theoretical posits of how such interventions may alter recovery of consciousness. Important chapters that are often not included in such texts cover topics on ethics, systems of care as well as caregiver burden/quality of life. Pediatric issues germane to diagnosis, prognosis, and treatment are also covered including the exploration of the gaps in this literature base. A very interesting chapter on near-death experiences examines psychological as well as neurobiological mechanisms of NDE and their vii

viii

Foreword

implications for better understanding of the neural correlates of consciousness. The last chapter, authored by the editors, examines directions for the future driven by ongoing guideline development as well as advances in neuroimaging and EEG-­ based BCIs. This volume will provide readers, whether physicians, other healthcare practitioners or other interested parties, with a critical informational foundation to truly gain an understanding of our current knowledge regarding disorders of consciousness. The information contained herein is quintessential to moving the field forward in improving the diagnosis, prognosis and treatment, both acute and long-term, of persons with DoC. The topics covered are diverse yet essential in garnering an appreciation for the complexities that we as clinicians face with this small albeit challenging patient population. Department of Physical Medicine and Rehabilitation Virginia Commonwealth University Richmond, VA, USA Concussion Care Centre of Virginia, Ltd. Richmond, VA, USA

Nathan D. Zasler

Preface

Consciousness is a word worn smooth by a million tongues. Depending upon the figure of speech chosen it is a state of being, a substance, a process, a place, an epiphenomenon, an emergent aspect of matter, or the only true reality. George Armitage Miller

Fifty years ago, the field of disorders of consciousness was a very limited field of research. Severely brain-injured patients, who are most likely to present impaired consciousness during recovery, often died. In the 1950s, the introduction of artificial breathing changed everything. The life of these patients could be extended even in cases of severe lesions to brain areas supporting the control of vital functions. The clinician started to face patients who were alive but not reactive to their surroundings. In this context, a new field was called to emerge, the disorders of consciousness. In the 1960s, Plum and Posner defined for the first time a clinical entity called the coma. Slightly later, Jennett and Teasdale developed the well-known Glasgow Coma Scale for assessing the progress of comatose patients in intensive care units. The 1980s were characterized by the development of a new kind of treatment, the sensory stimulation programs. Finally, in the late 1990s, the emergence of neuroimaging techniques opened new opportunities to study brain reactivity in patients with severe brain injuries. While the management of patients with severe brain injury remains challenging, the field is rapidly evolving. Just a decade ago, the primary focus was on characterizing and comprehending consciousness processing. Clinicians predominantly collaborated with neuroscientists, helping them in recruitment to enhance theoretical understanding, even if its direct impact on practice was very limited. Currently, the amassed knowledge—continuously expanding—holds tangible potential for patient assessment and treatment. Should this exponential surge in publications and interest persist, substantial shifts in the perception of neurorehabilitation’s role for patients with disorders of consciousness (DOC) are imminent. In light of this, recent endeavors have arisen to foster global networks of collaboration. Guidelines containing research evidence on treatment and appropriate management of DoC have been published and are disseminated. Of course, conducting research within an experimental framework also presents considerable difficulties ix

x

Preface

when dealing with this population. These patients are occasionally hard to enlist and maintain, often prone to fatigue and agitation. Thus, the existing coordination of multidisciplinary resources and knowledge—particularly between clinicians and neuroscientists—has become more crucial than ever before. Such collaboration undoubtedly holds the potential to surmount these complexities and is already now driving substantial enhancements in the care of patients with severe brain injuries. This book is meant both for clinicians and researchers and contains a summary of recent findings on diagnostic/prognostic, assessment techniques (i.e., behavioral scales, multi-modal assessment and brain–computer interface), clinical management and experimental treatment (particularly, neuromodulation) which, we hope, will stimulate ideas for clinicians and future research. In conclusion, we hope to have reached our aim by offering a comprehensive and reader-friendly book to readers both familiar or not with the difficult but intriguing field of disorders of consciousness. We hope you enjoy reading this book. Pomona, CA, USA Liege, Belgium 

Caroline Schnakers Steven Laureys

Contents

1

 Neural Correlates of Consciousness ������������������������������������������������������    1 Benedetta Cecconi, Glenn van der Lande, and Arianna Sala

2

Behavioral Assessment and Diagnosis of Disorders of Consciousness��������������������������������������������������������������������������������������   17 Caroline Schnakers and Katherine O’Brien

3

Brain–Computer Interfaces and Their Place in the Management of Disorders of Consciousness������������������������������   35 Michiel Meys, Aurore Thibaut, and Jitka Annen

4

 Prognosis in Disorders of Consciousness ����������������������������������������������   59 Anna Estraneo, Luigi Trojano, and Flora M. Hammond

5

Neuromedical Comorbidities and Their Management in Patients with DoC��������������������������������������������������������������������������������   77 Rita Formisano, Marta Aloisi, and Francesca Pistoia

6

 Music as Sensory Stimulation and Therapeutic Intervention�������������   99 Fabien Perrin and Wendy L. Magee

7

Pharmacological Treatments������������������������������������������������������������������  115 E. Szymkowicz, N. Alnagger, F. Seyfzadehdarabad, P. Cardone, J. Whyte, and O. Gosseries

8

Emerging Treatment for Patients with Disorders of Consciousness: The Field of Neuromodulation��������������������������������  147 Amber R. Hopkins, Marie M. Vitello, Aurore Thibaut, and Martin M. Monti

9

The Ethics in the Management of Patients with Disorders of Consciousness��������������������������������������������������������������������������������������  209 Michele Farisco

xi

xii

Contents

10 Taking  Care of Patients with Disorders of Consciousness: Caregivers’ Burden and Quality of Life������������������������������������������������  221 Matilde Leonardi, Davide Sattin, Martina Cacciatore, Camilla Ippoliti, Filippo Barbadoro, and Francesca G. Magnani 11 Models  and Systems of Care for Patients with Disorders of Consciousness��������������������������������������������������������������������������������������  243 Yelena Bodien, Chethan Venkatasubba Rao, Jan Lavrijsen, and Joseph T. Giacino 12 Pediatric  DOC: Diagnosis, Prognosis, and Treatment��������������������������  263 Erika Molteni, Beth S. Slomine, and Stacy J. Suskauer 13 Near-Death  Experiences: What Do We Know?������������������������������������  287 Pauline Fritz, Nicolas Lejeune, Helena Cassol, Steven Laureys, Olivia Gosseries, and Charlotte Martial 14 Future  Perspectives of Clinical Coma Science��������������������������������������  313 Steven Laureys and Caroline Schnakers Index������������������������������������������������������������������������������������������������������������������  319

Chapter 1

Neural Correlates of Consciousness Benedetta Cecconi, Glenn van der Lande, and Arianna Sala

Abstract  Significant advances have been made in the behavioral assessment and clinical management of disorders of consciousness (DoC). In addition, functional neuroimaging paradigms are now available to help assess consciousness levels in this challenging patient population. The success of these neuroimaging approaches as diagnostic markers is, however, intrinsically linked to understanding the relationships between consciousness and the brain. In this context, we will review the neural correlates of consciousness through various altered states such anesthesia and sleep and review how these relate to pathologies such as DoC after a severe acquired brain injury.

Introduction Understanding the relationships between consciousness and the brain is a long-­ standing question of neuroscience, psychology, and philosophy [1]. In this chapter, we will provide an overview of the current state of research in this field, from the standpoint of neuroscience. Firstly, we will provide a definition of neural correlates of consciousness (NCC), distinguishing between content- and state-NCC, background conditions, and consequences of consciousness. Secondly, we will describe neurophysiological findings in physiological (sleep), pharmacological (anesthesia), and pathological (disorders of consciousness) states of consciousness. Finally, we will discuss challenges and future perspectives in integrating these findings to produce a definitive account on the NCC.

B. Cecconi · G. van der Lande · A. Sala (*) Coma Science Group, GIGA-Consciousness, University of Liège, Liège, Belgium Centre du Cerveau(2), University Hospital of Liège, Liège, Belgium e-mail: [email protected]; [email protected]; [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 C. Schnakers, S. Laureys (eds.), Coma and Disorders of Consciousness, https://doi.org/10.1007/978-3-031-50563-8_1

1

2

B. Cecconi et al.

The Neural Correlates of Consciousness From a philosophical perspective, consciousness has been defined as the phenomenological experience or the “what it is like” to perceive, feel, or think about a given object [2]. Investigating the variations in neural activity accompanying variations in conscious experience represents a common search strategy to identify the neural correlates of consciousness (NCC) [1]. More precisely, finding the NCC means identifying the “minimum neural mechanisms sufficient for any one specific conscious percept” [1] or identifying “a neural system whose state directly correlates with whether a subject is conscious or not” [2]. Building on these complementary definitions, a distinction has been proposed between the so-called content- and state-NCC [3]. The content-NCC are the neural mechanisms underlying a given “phenomenal distinction” in subjective experience, being specific to a given content of consciousness. The content-NCC are typically studied in experimental paradigms where the visibility of a target is systematically manipulated so that in some trials the subject is aware of the target and in others (s)he is not, while the general state of the subject is maintained unchanged (e.g., constantly awake, attentive); to distill the content-­ NCC, the neural activity when the target stimulus is perceived is compared to the neural activity when the target stimulus is not perceived [1]. The state-NCC are the neural mechanisms underlying global, non-specific states of consciousness (e.g., alert wakefulness, dreaming, anesthesia) that regulate the range of cognitive systems an individual can mobilize, and hence the range and quality of conscious contents an individual can experience while in that state [4]. The state-NCC are typically studied by comparing neural activity in a state of consciousness (e.g., alert wakefulness) with neural activity in a state of unconsciousness (e.g., coma). Alternatively, within-state paradigms can be used, where brain activity across fluctuations in consciousness is contrasted within the same physiological state, e.g., dreaming during Rapid Eye Movement (REM) sleep (denoting consciousness) vs dreamless REM sleep (unconsciousness) [1]. The latter approach has the advantage of removing physiological differences that might occur across different states of consciousness, which are nevertheless unrelated to consciousness, and that might confound the results [5]. A further distinction is made between NCC and their background conditions and consequences. The background conditions are defined as factors that enable consciousness but do not directly contribute to it, as is the case for the neural activating system that contributes to vigilance and attention by widespread modulation of cortical activity, but does not contribute directly to the conscious experience [5]. The consequences of consciousness are, for example, the ability to produce a verbal or otherwise behavioral report of the subject’s own conscious experience. The background conditions and consequences of consciousness require particular attention as they might act as confounders in studies aiming at identifying the NCC.  For example, when studying the NCC using task-based paradigms that require behavioral reporting on the subject’s subjective experience, one should be careful in distinguishing between neural activity directly related to the conscious experience and neural activity related to the task of reporting [1].

1  Neural Correlates of Consciousness

3

In this chapter, we will focus on the state-NCC as identified via investigation of neural activity in different physiological, pharmacological, and pathological states of consciousness. These different states of consciousness are characterized by different levels of arousal (referring to the general state of wakefulness or alertness) and awareness (referring properly to the subjective experience) [6] (see Fig. 1.1). Brain activity is compared across or within different states of consciousness, using different neurophysiological techniques, to identify the patterns of brain activity that co-vary with awareness.

Fig. 1.1  Level of arousal and awareness across physiological, pharmacological, and pathological states of consciousness. In physiological states of consciousness (green), arousal and awareness typically co-vary, with levels of arousal and awareness decreasing from wakefulness, to drowsiness, to light and deep sleep. During dreaming, awareness increases while arousal remains low. In pharmacological states of consciousness (blue), under general anesthesia, low levels of arousal are accompanied by varying levels of awareness, with a presumably considerable proportion of subjects presenting no awareness whatsoever, and with others having dream-like experiences or even awareness of the environment. In pathological states of consciousness and associated syndromes (yellow), arousal and awareness vary widely. Arousal and awareness are absent in patients in coma; after coma, patients regain arousal and may (or may not) regain awareness, from UWS/VS (no awareness) to MCS (fluctuating but reproducible signs of awareness, like, e.g., visual pursuit in MCS- or language-related behaviors such as command following in MCS+) to EMCS (recovery of awareness and communication). Note that while patients might appear as behaviorally unresponsive (i.e., behaviorally presenting as UWS/VS), neurophysiological testing might suggest partially to fully preserved covert awareness, as in the case of MCS*, CMD, or locked-in syndrome. Abbreviations: CMD cognitive motor dissociation, EMCS emerging from minimally conscious state, MCS minimally conscious state, UWS/VS unresponsive wakefulness syndrome/vegetative state

4

B. Cecconi et al.

Neurophysiological Techniques Used in the Study of the NCC Neurophysiological techniques typically used to measure brain activity in the study of the NCC include positron emission tomography (PET), functional magnetic resonance imaging (fMRI), and electroencephalography (EEG). PET is a molecular imaging technique that allows for measurement of different biochemical processes of interest, with high biological specificity and good spatial resolution, in both cortical and subcortical structures. PET is most commonly used in combination with the fluoro-deoxy-glucose (FDG) tracer, to measure glucose metabolism. Glucose metabolism is the main substrate for energy in neurons and astrocytes  and is directly linked to excitatory glutamatergic neurotransmission. Both absolute changes in brain global glucose consumption and relative changes in local glucose consumption can be assessed. Typically, relative decreases in  local glucose metabolism, called hypometabolism, are investigated in the study of NCC. fMRI is a functional imaging technique that allows for measurement of blood oxygenation, as indexed by the blood oxygen-level dependent (BOLD) signal, with good spatial and temporal resolution, in both cortical and subcortical structures. BOLD signal is indirectly coupled to neural activity via the hemodynamic response function. Co-fluctuations of BOLD signal across time are deemed as a proxy for synchronized neural activity, also known as “functional connectivity”. A way of describing the functional connectivity of brain regions is their organization in “brain networks”. Brain networks include, for example, the default mode and the executive control networks, which are of particular relevance in the study of NCC. EEG is an electrophysiological technique that allows for measurement of voltage fluctuations via electrodes placed on the scalp, with extremely good temporal resolution. The typical parallel organization of cortical pyramidal neurons allows the summation of their postsynaptic potentials to reach fluctuations measurable through the scalp. As such, the number of synchronized neurons determines the energy of the measurable voltage on the scalp, while the rhythm of their synchronized firing determines its frequency. Classically, this can be analyzed on the scalp level as a power-spectral density, a curve that shows the energy across different frequency bands, typically delta (1–4 Hz), theta (4-8 Hz), alpha (8–12 Hz), beta (12–30 Hz), and gamma (>30 Hz). EEG has also been used for estimation of functional connectivity, either at electrode level or at the cortical level, using source reconstruction methods that allow for estimation of where the electrical signal in the brain originated. Finally, EEG has been used to investigate the complexity of activity in the brain. One example of this is the Perturbational Complexity Index (PCI), which is a quantification of the complexity of the brain response to transcranial magnetic stimulation (TMS).

1  Neural Correlates of Consciousness

5

Physiological States of Consciousness: NCC in Sleep Sleep is the most natural fluctuation in consciousness. It is widely preserved across the animal kingdom, and while its putatively essential function remains unknown [7], it has been associated with, for instance, plasticity and memory consolidation [8] or synaptic homeostasis [9]. Behaviorally, sleep manifests as periods of quiescence and reduced responses to the environment, which are coupled to profound changes in the state of the brain. As such, sleep is seen as a different state of consciousness compared to wakefulness [10]. The gold standard of tracking the neurophysiology of sleep is through polysomnography, which includes an EEG and other physiological signals such as heart beat, respiration, and muscle tension [11]. During a sleep cycle, the brain goes through several stages at approximately 90 min per cycle [12]. The cycle starts with a transition from wakefulness to non-rapid eye movement (NREM) stage 1, marked by a significant slowing of the EEG with the disappearance of alpha (8–12 Hz) oscillations, and is deemed light sleep in terms of consciousness [11]. Sleep, and thus potentially unconsciousness, is deepened in NREM2, which is characterized by further EEG slowing and the appearance of sleep spindles, a waxing and waning waveform of 12–16  Hz, and K-complexes, large amplitude waves of a short negative and longer positive deflection. The deepest stage of sleep is NREM3 where the EEG is dominated by large-amplitude slow oscillation (28 days) disorders of consciousness (DoC) [5, 6]. Indeed, patients with DoC require specialized care and assessment. At least 3 of the 18 AAN recommendations address behavioral assessment of consciousness while all EAN recommendations address this topic. Accurate behavioral assessment requires expertise on behalf of the clinician as there is no true measure of consciousness but more an assessment of behavioral proxies. The clinician will be observing the patient’s spontaneous behaviors and their reactions to stimuli occurring in their environment and deciphering the patient’s level of consciousness. The clinician’s mindset and biases can greatly impact the result of a behavioral assessment. It also depends on the physical (neuromuscular) and mental capacities (particularly, the arousal level) of the patient at the time of assessment. Missing signs of consciousness is not a rare fact, and diagnostic errors are frequent (i.e., around 40%) [7, 8]. The accuracy of the assessment is, however, crucial clinically, psychologically, legally, and ethically. It influences the way the patient’s care will be oriented. Developing and administering valid and sensitive behavioral scales to detect the presence of signs of consciousness, even subtle, therefore represents a significant challenge.

2  Behavioral Assessment and Diagnosis of Disorders of Consciousness

19

Disorders of Consciousness Coma Patients who survive a severe brain injury can remain unconscious for several weeks, being neither awake nor conscious. They are in a state called coma, defined as “a pathological state marked by severe and prolonged dysfunction of vigilance and consciousness” [9] (Fig.  2.1 and Table  2.1). This state usually results either from a lesion limited to the brainstem (involving the reticular activating system) or from global brain dysfunction (most often caused by bihemispheric, diffuse axonal injury after traumatic brain injury). The distinguishing features of coma are the continuous absence of eye opening (spontaneously or after stimulation) and the absence of oriented or voluntary motor or verbal (including vocalization) responses. There is no evidence of visual fixation or pursuit, even after manual eye-opening. This state must last at least 1 h to be differentiated from a transient state such as syncope, acute confusion, or delirium. Prolonged coma is rare as this condition typically resolves within 2–4 weeks, most often evolving into vegetative state or minimally conscious state [10]. It is worth noting that being in a coma is different from being brain dead. The term brain death requires a bedside demonstration of irreversible cessation of all functions of the brain, including brainstem functions, for

Conscious Non-Conscious Behavioral

Conscious wake

Behaviorally indistinguishable patients

Drowsiness

Light sleep

Ability to produce motor behavior

MCS

Deep sleep

Non-behavioral Coma

General Anesthesia

VS

Non-awake Level of consciousness (wakefulness)

Aware Content of consciousness (awareness) Non-aware

Awake

Fig. 2.1  The conundrum of consciousness. Disorders of consciousness are defined by two main components: the level of consciousness and the content of consciousness. This figure illustrates where different states (i.e., coma, vegetative state—VS, minimally conscious state—MCS but also states related to sleep and anesthesia) are placed on both continuums. It also represents where patients with covert cognition would be placed. (Reprinted from [69] with permission from Annual Reviews, Inc.)

None

None

None

None

None

None

Affective response Response to command Verbalization

Communication

None

Random

None

Visual response

VS Spontaneous Reflexive/ patterned Flexion withdrawal/ posturing Startle

Posturing/ none

Coma None None

Response to pain

Eye opening Movement

Random vocalization/none Unreliable

None

Object localization/ pursuit/fixation Contingent

MCS− Spontaneous Automatic/object manipulation Localization

Unreliable

Intelligible words

Reproducible

Contingent

Object recognition

MCS+ Spontaneous Automatic/object manipulation Localization

Consistent/ reproducible Intelligible words Reliable

Object recognition Contingent

EMCS Spontaneous Functional object use N/A

Unreliable/reliable as detected by neuroimaging or electrophysiology)

Consistent/reproducible (as detected by neuroimaging or electrophysiology) None

Random

Startle

Flexion withdrawal/posturing

CA Spontaneous Reflexive/patterned

Table 2.1  Comparison of the behavioral features of coma, VS, MCS−, MCS+, emergence from MCS (EMCS) and covert awareness (CA)

20 C. Schnakers and K. O’Brien

2  Behavioral Assessment and Diagnosis of Disorders of Consciousness

21

example ruling out all pharmacological or medical confounds. In contrast to brain death, critical functions such as respiration, circulation, and neuroendocrine and homeostatic regulation are present in patients with coma [11].

Vegetative State/Unresponsive Wakefulness Syndrome In 1972, the term vegetative state (VS) was first introduced by Jennett and Plum to describe “an organic body capable of growth and development but devoid of sensation and thought” [9] (Fig. 2.1 and Table 2.1). The term vegetative indicates preserved physiological functions without clear signs of consciousness of either the self or the environment. This clinical entity was identified following the implementation of artificial breathing techniques in intensive care units. Since then, the number of scientific studies performed on VS patients has continuously increased. The term “persistent VS” has been used for patients who have been in this state for more than a month, while “permanent VS” has been used for patients who have been in this state for more than 6 months after a non-traumatic brain injury (e.g., stroke or anoxia) or more than a year after a traumatic brain injury [10]. However, recent evidence suggests that some patients may recover even after a year, and the AAN recommends replacing “permanent VS” with “chronic VS,” with the duration specified [5]. Given the negative connotation of the term “vegetative state,” the European Task Force on Disorders of Consciousness has also proposed the use of a more neutral and descriptive term, such as “Unresponsive Wakefulness Syndrome” (UWS) [12]. Behaviorally, patients in VS/UWS open their eyes spontaneously or in response to stimulation and present with preserved autonomic functions (e.g., breathing, cardiovascular regulation, thermoregulation), but they are not conscious and show only reflexive behaviors [10]. The VS/UWS often results from an injury involving the white matter or bilateral lesions of the thalamus (i.e., intralaminar nuclei) [13, 14]. These patients are able to maintain spontaneous breathing and can open their eyes periodically. The resumption of spontaneous eye opening does not necessarily indicate the return of normal sleep–wake cycles [15]. The dissociation of eye opening and presence of consciousness can be confusing for the patient’s family. The patient may also vocalize and demonstrate facial expressions that can be interpreted as smiling, crying, or grimacing but these behaviors often occur out of context. Even randomly produced single words have been reported in patients diagnosed with VS/ UWS. Providing information related to level of consciousness, as accurately as possible, is essential for helping the family make the best decisions for their loved one and themselves [16]. The confidence in the behavioral findings should be relayed in an understandable fashion, as noted by the Minimal Competency Recommendations for treatment of DoC [17].

22

C. Schnakers and K. O’Brien

Minimally Conscious State The minimally conscious state (MCS) was identified more recently than VS/UWS and coma. It was defined in 2002 by the Aspen Workgroup as being characterized by the presence of inconsistent but clearly discernible behavioral signs of consciousness (Fig. 2.1 and Table 2.1) [18]. Patients evolving from a VS/UWS to an MCS begin to demonstrate behaviors suggestive of awareness of their environment. Visual fixation and tracking are usually the first signs of consciousness but can be difficult to detect and require the use of sensitive diagnostic tools (e.g., the use of a mirror) [19]. Signs of consciousness in MCS patients can be inconsistent due to high vigilance fluctuations, but they must be replicated within a given examination to meet diagnostic criteria. Conducting serial examinations (n = 5) is also crucial before an accurate diagnosis can be made [20]. Mat and coworkers recently published new behavioral signs of consciousness that are not always detected using the current standardized measures, such as: resistance to eye opening, spontaneous blink rate, facial grimace, habituation to auditory startle reflex, leg crossing, as well as oral feeding and swallowing [21]. Auditory localization has also recently been shown to represent MCS rather than initially presented as a reflection of VS/UWS [22]. In recent years, MCS has been further subdivided into two clinical entities, MCS+ and MCS−, based on the presence or absence of language function (e.g., response to command and/or verbalization) [23]. The clinical subcategorization of MCS is supported by metabolic differences in areas that are associated with both consciousness (i.e., lower metabolism in precuneus and thalamus in MCS−) [24] and language (i.e., lower metabolism in the left middle temporal cortex and lower connectivity between left angular gyrus and left prefrontal cortex in MCS−) [25, 26]. Several studies recently showed that, at discharge from acute rehabilitation, patients in an MCS+ recovered consciousness most frequently and had the least disability [24, 27]. Emergence from MCS is defined by the demonstration of reliable and consistent functional communication (which may occur through speech, writing, yes/no signals, or augmentative communication devices), or functional object use (i.e., discrimination and appropriate use of two or more objects). A majority of patients with severe brain injury recover functional communication within 8 weeks post injury [28]. There also have been recent findings showing that consistent response to command is associated with emergence from MCS and should be recognized as such when updating the diagnostic criteria for MCS [29]. When emerged from MCS, most patients evolve toward an acute confusional state characterized by cognitive impairment (98%), disorientation (93%), and agitation (69%) [30]. Most of those who remain in MCS for 12 months remain severely disabled. However, recent studies suggest that signs of recovery after severe brain injury may be observable over longer time periods and approximately one in five MCS patients will eventually continue to live at home or in the community [2]. The duration of MCS nevertheless seems to be a strong predictor of the length of confusional state [30].

2  Behavioral Assessment and Diagnosis of Disorders of Consciousness

23

Covert Awareness For the past decade, clinicians have been encountering a new type of patient showing covert awareness. These patients are unable to display any physical signs of consciousness but can demonstrate high cognitive processing since they are able to respond mentally to active neuroimaging or electrophysiological tests (Fig. 2.1 and Table  2.1) [31, 32]. These patients might be misdiagnosed as being in locked-in syndrome (LIS) [33], which is caused by a localized ventropontine lesion resulting in paralysis of all four limbs without affecting consciousness or cognition. However, recent studies suggest that this clinical entity may be caused by impaired connectivity between the thalamus and primary motor cortex, which interferes with the execution of voluntary motor actions [34]. Two meta-analyses revealed that this clinical entity is uncommon but not rare in patients in a vegetative state (14–17%), but more frequent in those with traumatic brain injury [35, 36]. Future studies will need to be conducted across multiple centers to gather enough data to establish a complete profile of this clinical condition. Moreover, even if this phenomenon has been known for more than a decade, there has been no agreement regarding the taxonomy to use for these patients who are able to respond to active neuroimaging or electrophysiological paradigms. A recent systematic review found 25 different names given to this entity across the literature [37]. The five following names were the ones the most frequently used: covert awareness, cognitive motor dissociation, functional locked-in, and non-behavioral MCS (MCS*). Future studies will have to help in reaching a standard taxonomy, which will be key to achieve a successful clinical translation that is crucially needed. Previous research has demonstrated that some patients with covert cognition have the ability to communicate. Therefore, recent studies have been exploring the potential benefits of brain–computer interfaces (BCI) for severely brain-injured patients. These interfaces utilize neuroimaging or electrophysiological signals to enable communication for patients who are otherwise unable to express themselves [38]. However, BCI paradigms are currently hard to implement. The tasks used in BCI communication are complex and some patients may not be able to respond, even if they are conscious. To implement augmentative communication techniques successfully, future studies will need to gain a better understanding of the residual cognitive functioning of these patients [39].

Bedside Assessment In light of the behavioral fluctuations that commonly occur in this population, evaluations should be repeated over time and measures should be sensitive enough to detect subtle but clinically meaningful changes in neurobehavioral responsiveness [40]. Conventional bedside assessment procedures and neurosurgical rating scales such as the Glasgow Coma Scale [41] (GCS) have limited utility when used to

24

C. Schnakers and K. O’Brien

monitor progress in patients with prolonged disturbance in consciousness. These procedures detect relatively gross changes in behavior and are not designed to distinguish random or reflexive behaviors from those that are volitional. Standardized rating scales have been devised to address these shortcomings, to assess a broad range of neurobehavioral functions, and to rely on fixed administration and scoring procedures. In 2010, the American Congress of Rehabilitation Medicine published the results of the first evidence-based review of neurobehavioral rating scales designed specifically for patients with DOC [42]. Six of the 13 scales that qualified for the review were recommended for use in clinical practice. The Coma Recovery Scale-Revised (CRS-R) received the strongest recommendation (“minor reservations”), based on its performance across a panel of psychometric quality indicators. The CRS-R is also one of the Traumatic Brain Injury (TBI) Common Data Elements (CDE) suggested by the US National Institute of Neurological Disorders and Stroke (NINDS) and the method of choice for monitoring recovery of consciousness in TBI research [43]. The Full Outline of UnResponsiveness score (FOUR score) has greater sensitivity than the GCS for detecting different levels of brainstem function in the acute stage of severe brain injury, but because the FOUR score does not include a systematic assessment of signs of consciousness, it may not capture the transition from VS/ UWS to MCS [44]. Standardized neurobehavioral assessment measures tailored for DOC patients include the Coma Recovery Scale-Revised (CRS-R), the Coma-Near Coma Scale (CNC), the Western Neurosensory Stimulation Profile (WNNSP), the Western Head Injury Matrix (WHIM), the Sensory Modality and Rehabilitation Technique (SMART), and the DoC Scale. Although item content varies across measures, all evaluate behavioral responses to a variety of auditory, visual, motor, and communication prompts. All of these instruments have been shown to have adequate reliability and validity; however, there is considerable variability in their psychometric integrity and clinical utility [42].

The Coma Recovery Scale-Revised Among the mentioned measures, the CRS-R is unique in its direct incorporation of the established diagnostic criteria for VS/UWS and MCS within its administration and scoring process (Table 2.2). The EAN has also strongly recommended its use to assess DoC due to its excellent performance across various psychometric quality indicators [6]. Lastly, the scale has been translated into over 10 languages and is currently utilized worldwide. The CRS-R scale comprises 23 items organized into six subscales that cover auditory, visual, motor, verbal/oromotor, communication, and arousal functions. Scoring follows standardized guidelines and is based on the presence or absence of operationally defined behavioral responses to specific sensory stimuli. Psychometric studies have demonstrated that the CRS-R meets rigorous criteria for measurement

2  Behavioral Assessment and Diagnosis of Disorders of Consciousness Table 2.2  Coma Recovery Scale-Revised

25

Auditory function scale 4—Consistent movement to commanda 3—Reproducible movement to commanda 2—Localization to sound 1—Auditory startle 0—None Visual function scale 5—Object recognitiona 4—Object localization: reachinga 3—Pursuit eye movementsa 2—Fixationa 1—Visual startle 0—None Motor function scale 6—Functional object useb 5—Automatic motor responsea 4—Object manipulationa 3—Localization to noxious stimulationa 2—Flexion withdrawal 1—Abnormal posturing 0—None/flaccid Oromotor/verbal scale 3—Intelligible verbalizationa 2—Vocalization/oral movement 1—Oral reflexive movement 0—None Communication scale 2—Functional: accurateb 1—Non-functional: intentionala 0—None Arousal scale 3—Attentiona 2—Eye opening w/o stimulation 1—Eye opening with stimulation 0—Unarousable a

 Denotes MCS  Denotes emergence from MCS

b

and evaluation tools used in interdisciplinary rehabilitation settings. Trained examiners can administer the CRS-R reliably, and scores tend to remain relatively stable across repeated assessments. The CRS-R has also been adapted for use in the pediatric population. The CRS-Pediatrics (CRS-P) is suitable for children as young as 12 months of age [45].

26

C. Schnakers and K. O’Brien

In subacute settings, the CRS-R has exhibited superior diagnostic accuracy compared to other consciousness scales like the GCS or the FOUR [42]. Recent research findings indicate that higher CRS-R scores upon admission to inpatient rehabilitation facilities can help differentiate patients who will have better outcomes at discharge. This information is valuable for rehabilitation planning and effective communication with patients and their caregivers [46, 47]. A study by Wannez and colleagues demonstrated that focusing on the five most commonly observed behaviors in the CRS-R assessment (i.e., fixation, visual pursuit, reproducible movement to command, automatic motor response, and localization to noxious stimulation) successfully identified 99% of patients in MCS [48]. Based on these findings, the Simplified Evaluation of CONsciousness Disorders (SECONDs) has recently been developed for allowing shorter but sensitive assessments and might be an ideal alternative in subacute settings where time is limited [49, 50].

Individualized Quantitative Behavioral Assessment (IQBA) Clinicians involved in the care of DoC patients often encounter situations in which the patient’s behavioral responses are ambiguous or occur too infrequently to clearly discern their significance. These problems are often due to injury-related sensory, motor, and arousal deficits. For this reason, a technique referred to as Individualized Quantitative Behavioral Assessment (IQBA) was developed by Whyte and colleagues [51]. IQBA is intended to address case-specific questions using individually tailored standardized assessment procedures, operationally defined target responses, and controls for examiner and response bias. Once the target behavior (e.g., command-­following, visual tracking) has been operationalized, the frequency of the behavior is recorded following administration of an appropriate command, a contra-­ command, and during a rest interval. Data are analyzed statistically to determine whether the target behavior occurs significantly more often in one condition relative to other conditions. For example, if the frequency of observed behaviors is similar during the “command” and “contra-command” conditions and significantly higher relative to the “rest” condition, it is likely that the behaviors represent response to verbal input but not command-following (e.g., aphasia). Day and colleagues demonstrated that IQBA can provide consistent responses to commands and detect consciousness earlier than the CRS-R in some patients. Therefore, IQBA approaches could be used in conjunction with the CRS-R to improve diagnostic accuracy [52].

Pain Assessment Detecting whether a patient with DoC is experiencing pain is important for clinicians and families. However, self-reporting is not possible in these patients due to their inability to communicate. Currently, the Nociception Coma Scale-Revised

2  Behavioral Assessment and Diagnosis of Disorders of Consciousness

27

(NCS-R) can be used to evaluate nociception and pain in patients with DoC. The Nociception Coma Scale (NCS) and its revised version (NCS-R) were developed using pre-existing pain scales validated for non-communicative patients with dementia and newborns [53, 54]. The NCS-R includes three subscales that assess motor and verbal functions, as well as facial expression, making it suitable for assessing nociception and pain in patients with DoC. Based on this revised version (score range: 0–9), Chatelle and coworkers defined a cut-off score of above 4 as a potential clinical threshold for detecting pain in DoC patients (sensitivity of 73%, specificity of 97%, and accuracy of 85%) [54]. Using this threshold, Chatelle and coworkers assessed the clinical usefulness of the NCS-R and showed, in DoC patients with potential painful conditions (e.g., due to fractures, decubitus ulcers, or spasticity), decreased total scores under analgesic treatment without a decrease of level of consciousness [55]. In a recent neuroimaging study using an 18-fluoro-­ deoxyglucose PET scan, the same investigators have found a significant correlation between NCS-R total scores, its pain threshold, and brain metabolism in the pain network and more particularly with the anterior cingulate cortex (ACC), which is associated with the emotional processing of pain [56]. These results suggest that the NCS-R is related to pain processing and constitutes an appropriate behavioral tool to assess, monitor, and treat pain in non-communicative patients with DoC.  The NCS and its revised version have been translated into eight languages, including English, French, Italian, Flemish, Danish, Portuguese, Russian, and Thai. They are used in various settings, from intensive care [57] to long-term facilities [58]. A recent systematic review found that both scales are valid and useful instruments for assessing pain in DoC patients based on their psychometric properties [59]. The NCS is the first scale developed for assessing nociceptive pain in patients with severe brain injury. However, Whyte and colleagues have recently developed another measure of nociception specifically for these patients called the Brain Injury Nociception Assessment Measure (BINAM). Preliminary results suggest that the BINAM is a reliable and feasible tool to assess the intensity of nociception, independent of the level of consciousness, which is not the case with the NCS-R. More data are needed to further establish the psychometric properties of the new scale [60].

Language Assessment The current behavioral scales used to evaluate cognition in individuals with DoC lack the ability to identify precise cognitive impairments. Schnakers and colleagues have demonstrated that assessing consciousness is challenging due to the simultaneous presence of cognitive deficits like difficulties in receptive language. This highlights the importance of creating novel tools and scales to accurately assess these impairments in these patients [61]. A new cognitive assessment tool called the Cognitive Assessment by Visual Election (CAVE) has been developed specifically for patients with DoC [24, 62]. The CAVE consists of six subscales, each comprising ten items, which evaluate the

28

C. Schnakers and K. O’Brien

recognition of real objects, numbers, written words, letters, pictures, and colors. During the test, the patient is instructed to focus on a specific target while ignoring any distractions. Since the test requires the ability to maintain visual fixation, it is designed for patients in an MCS−, an MCS+, and emerging from MCS. On average, the administration of the CAVE takes between 10 and 30 min. Preliminary findings have shown a high level of agreement among different raters, indicating strong inter-rater reliability, along with satisfactory internal consistency [62]. Additionally, CAVE scores appear to decrease as the CRS-R total score decreases, establishing a consistent behavioral and cognitive profile for each patient. Lastly, patients with higher CRS-R and CAVE scores exhibited less pronounced cerebral hypometabolism in the language network [25]. Finally, the Brief Evaluation of Receptive Aphasia (BERA) was created with the intention of assessing remaining language skills in a detailed and thorough manner. Its purpose is to differentiate between receptive phonological, semantic, and morphosyntactic abilities by observing the visual fixation on target stimuli amid phonologically or semantically similar distractions. While this scale has been validated for language-specific evaluation among individuals with aphasia, its complete validation in patients with DoC is nevertheless still pending [63].

Diagnostic Accuracy A series of studies have found that around 40% of patients believed to be in VS/ UWS are misdiagnosed and are in fact conscious [8]. Diagnostic accuracy for differentiating MCS from VS/UWS can be impacted by various factors, including biases from the examiner, patient-related factors, and environmental factors. Standardized rating scales, such as the Coma Recovery Scale-Revised (see section above), can help reduce examiner error in diagnosing patients with MCS or VS/ UWS [64]. However, it is important for examiners to follow specific administration and scoring guidelines to ensure diagnostic accuracy. Serial assessments, meaning multiple assessments conducted over a short period of time, have also been recommended by both the AAN and EAN to reduce misdiagnosis [5, 6]. Recent research has suggested that performing at least five assessments within a 2-week period may reduce misdiagnosis compared to a single assessment. Additionally, using relevant stimuli, such as a mirror to detect visual pursuit or familiar objects, during assessments may also improve diagnostic accuracy [65, 66]. Involving the family in the assessment should be considered as it might lead to increased responsiveness (see Chap. 11). Fluctuations in arousal level, fatigue, seizures, acute infections with fever (e.g., urinary tract infection or aspiration pneumonia), pain (e.g., due to decubitus, fractures, heterotopic ossification), (cortical) sensory deficits (leading to visual and hearing impairments), sympathetic storming, and motor impairment (e.g., generalized hypotonus, spasticity, or paralysis) also decrease the probability of observing signs of consciousness (see Chap. 5). Optimizing arousal is crucial for accurate

2  Behavioral Assessment and Diagnosis of Disorders of Consciousness

29

consciousness assessment in patients with disorders of consciousness, such as those in VS/UWS or MCS. Fluctuations in arousal level can make it difficult to assess the patient’s level of consciousness, as they may have brief periods of wakefulness and responsiveness that can be missed during an assessment. The AAN and EAN guidelines emphasize the importance of optimizing arousal in patients with disorders of consciousness, through measures such as reducing sedation, avoiding medications that may depress consciousness, and providing sensory stimulation to increase wakefulness [5, 6]. Recent studies have highlighted the impact of circadian rhythms on arousal and the potential use of body temperature to identify the optimal time to assess patients. For example, patients with disorders of consciousness often have disrupted sleep–wake cycles, which can affect their level of arousal and responsiveness during assessments. By monitoring their body temperature over time, it may be possible to identify their circadian rhythm and determine the best time to conduct assessments when their level of arousal is highest. However, further research is needed to determine the utility of this approach in clinical practice [67, 68]. Finally, the environment in which the patient is evaluated may bias assessment findings. Restricted range of movement stemming from restraints and immobilization techniques, poor positioning, under-suctioning, no adequate rest prior to assessment (care/therapy just before), and excessive ambient noise/heat/light can all decrease or distort voluntary behavioral responses (see Chap. 5).

Conclusion The detection of signs of consciousness can be challenging at the bedside as we are limited to the behavioral responses within the patient’s capacity. The use of sensitive standardized scales is crucial but does not protect against confounds to consciousness or mimics. As misdiagnosis can lead to serious consequences especially in terms of pain treatment and planning of care decisions, neuroimaging could constitute a complementary tool when disentangling the different states of consciousness.

References 1. CDC. Emergency department visits, hospitalizations, and deaths—United States. 2019. https:// www.cdc.gov/traumaticbraininjury/pdf/TBI-­Surveillance-­Report-­FINAL_508.pdf. 2. Hammond FM, Giacino JT, Nakase Richardson R, et al. Disorders of consciousness due to traumatic brain injury: functional status ten years post-injury. J Neurotrauma. 2019;36(7):1136–46. 3. Wilkins TE, Beers SR, Borrasso AJ, et al. Favorable functional recovery in severe traumatic brain injury survivors beyond six months. J Neurotrauma. 2019;36(22):3158–63. 4. Wijdicks EF.  Minimally conscious state vs. persistent vegetative state: the case of Terry (Wallis) vs. the case of Terri (Schiavo). Mayo Clin Proc. 2006;81(9):1155–8.

30

C. Schnakers and K. O’Brien

5. Giacino JT, Katz DI, Schiff ND, et  al. Comprehensive systematic review update summary: disorders of consciousness: report of the guideline development, dissemination, and implementation subcommittee of the American Academy of Neurology; the American Congress of Rehabilitation Medicine; and the National Institute on Disability, Independent Living, and Rehabilitation. Neurology. 2018;91(10):461–70. 6. Kondziella D, Bender A, Diserens K, et al. European Academy of Neurology guideline on the diagnosis of coma and other disorders of consciousness. Eur J Neurol. 2020;27(5):741–56. 7. Schnakers C, Vanhaudenhuyse A, Giacino J, et al. Diagnostic accuracy of the vegetative and minimally conscious state: clinical consensus versus standardized neurobehavioral assessment. BMC Neurol. 2009;9:35. 8. van Erp WS, Lavrijsen JC, Vos PE, et al. The vegetative state: prevalence, misdiagnosis, and treatment limitations. J Am Med Dir Assoc. 2015;16(1):85.e9–85.e14. 9. Posner J, Saper C, Schiff N, Claassen J. Plum and Posner’s diagnosis of stupor and coma. 5th ed. New York: Oxford University Press; 2019. 10. The Multi-Society Task Force on Persistent Vegetative State. Medical aspects of the persistent vegetative state. N Engl J Med. 1994;330(21):1499–508. 11. Wijdicks EF, Varelas PN, Gronseth GS, et al. American Academy of Neurology. Evidence-­ based guideline update: determining brain death in adults: report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology. 2010;74(23):1911–8. 12. Laureys S, Celesia GG, Cohadon F, et al. Unresponsive wakefulness syndrome: a new name for the vegetative state or apallic syndrome. BMC Med. 2010;8:68. 13. Fernández-Espejo D, Junque C, Bernabeu M, et  al. Reductions of thalamic volume and regional shape changes in the vegetative and the minimally conscious states. J Neurotrauma. 2010;27(7):1187–93. 14. Newcombe VF, Williams GB, Scoffings D, et  al. Aetiological differences in neuroanatomy of the vegetative state: insights from diffusion tensor imaging and functional implications. J Neurol Neurosurg Psychiatry. 2010;81(5):552–61. 15. Landsness E, Bruno MA, Noirhomme Q, et  al. Electrophysiological correlates of behavioural changes in vigilance in vegetative state and minimally conscious state. Brain. 2011;134(8):2222–32. 16. Leonardi M, Giovannetti AM, Pagani M, et al. Burden and needs of 487 caregivers of patients in vegetative state and in minimally conscious state: results from a national study. Brain Inj. 2012;26(10):1201–10. 17. Giacino JT, Whyte J, Nakase-Richardson R, et al. Minimum competency recommendations for programs that provide rehabilitation Services for persons with disorders of consciousness: a position statement of the American Congress of Rehabilitation Medicine and the National Institute on Disability, Independent Living and Rehabilitation Research Traumatic Brain Injury Model Systems. Arch Phys Med Rehabil. 2020;101(6):1072–89. 18. Giacino J, Ashwal S, Childs N, et al. The minimally conscious state: definition and diagnostic criteria. Neurology. 2002;58(3):349–53. 19. Overbeek BUH, Eilander HJ, Lavrijsen JCM, et al. Are visual functions diagnostic signs of the minimally conscious state? An integrative review. J Neurol. 2018;265(9):1957–75. 20. Wannez S, Heine L, Thonnard M, et al. The repetition of behavioral assessments in diagnosis of disorders of consciousness. Ann Neurol. 2017;81(6):883–9. 21. Mat B, Sanz LRD, Arzi A, et al. New behavioral signs of consciousness in patients with severe brain injuries. Semin Neurol. 2022;42(3):259–72. 22. Carrière M, Cassol H, Aubinet C, et al. Auditory localization should be considered as a sign of minimally conscious state based on multimodal findings. Brain Commun. 2020;2(2):fcaa195. 23. Thibaut A, Bodien YG, Laureys S, et al. Minimally conscious state “plus”: diagnostic criteria and relation to functional recovery. J Neurol. 2020;267(5):1245–54. 24. Aubinet C, Murphy L, Bahri MA, et al. Brain, behavior, and cognitive interplay in disorders of consciousness: a multiple case study. Front Neurol. 2018;9:665.

2  Behavioral Assessment and Diagnosis of Disorders of Consciousness

31

25. Aubinet C, Larroque SK, Heine L, et al. Clinical subcategorization of minimally conscious state according to resting functional connectivity. Hum Brain Mapp. 2018;39(11):4519–32. 26. Aubinet C, Cassol H, Gosseries O, et al. Brain metabolism but not gray matter volume underlies the presence of language function in the minimally conscious state (MCS): MCS+ versus MCS− neuroimaging differences. Neurorehabil Neural Repair. 2020;34(2):172–84. 27. Giacino JT, Sherer M, Christoforou A, et  al. Behavioral recovery and early decision making in patients with prolonged disturbance in consciousness after traumatic brain injury. J Neurotrauma. 2020;37(2):357–65. 28. Martens G, Bodien Y, Thomas A, et  al. Temporal profile of recovery of communication in patients with disorders of consciousness after severe brain injury. Arch Phys Med Rehabil. 2020. pii: S0003-9993(20)30106-4. 29. Weaver JA, Cogan AM, O’Brien KA, et  al. Determining the hierarchy of Coma Recovery Scale-Revised rating scale categories and alignment with Aspen consensus criteria for patients with brain injury: a Rasch analysis. J Neurotrauma. 2022;39(19–20):1417–28. 30. Bodien YG, Martens G, Ostrow J, et al. Cognitive impairment, clinical symptoms and functional disability in patients emerging from the minimally conscious state. NeuroRehabilitation. 2020;46(1):65–74. 31. Schiff ND.  Cognitive motor dissociation following severe brain injuries. JAMA Neurol. 2015;72(12):1413–5. 32. Owen AM. The search for consciousness. Neuron. 2019;102(3):526–8. 33. Schnakers C, Laureys S. Coma and disorders of consciousness. Springer; 2018. 34. Fernández-Espejo D, Rossit S, Owen AM. A thalamocortical mechanism for the absence of overt motor behavior in covertly aware patients. JAMA Neurol. 2015;72(12):1442–50. 35. Kondziella D, Friberg CK, Frokjaer VG, et al. Preserved consciousness in vegetative and minimal conscious states: systematic review and meta-analysis. J Neurol Neurosurg Psychiatry. 2016;87(5):485–92. 36. Schnakers C, Hirsch M, Noé E, et al. Covert cognition in disorders of consciousness: a meta-­ analysis. Brain Sci. 2020;10(12):930. 37. Schnakers C, Bauer C, Formisano R, et al. What names for covert awareness? A systematic review. Front Hum Neurosci. 2022;16:971315. 38. Gibson RM, Owen AM, Cruse D.  Brain-computer interfaces for patients with disorders of consciousness. Prog Brain Res. 2016;228:241–91. 39. Schnakers C, Giacino JT, Løvstad M, et al. Preserved covert cognition in noncommunicative patients with severe brain injury? Neurorehabil Neural Repair. 2015;29(4):308–17. 40. Giacino JT, Katz DI, Schiff ND, et al. Practice guideline update recommendations summary: disorders of consciousness: report of the guideline development, dissemination, and implementation subcommittee of the American Academy of Neurology; the American Congress of Rehabilitation Medicine; and the National Institute on Disability, Independent Living, and Rehabilitation Research. Neurology. 2018;91(10):450–60. 41. Teasdale G, Jennett B. Assessment of coma and impaired consciousness. Lancet. 1974;2:81–4. 42. Seel RT, Sherer M, Whyte J, et  al. Assessment scales for disorders of consciousness: evidence-­based recommendations for clinical practice and research. Arch Phys Med Rehabil. 2010;91(12):1795–813. 43. Hicks R, Giacino J, Harrison-Felix C, et al. Progress in developing common data elements for traumatic brain injury research: version two—the end of the beginning. J Neurotrauma. 2013;30(22):1852–61. 44. Bruno MA, Ledoux D, Lambermont B, et  al. Comparison of the full outline of unresponsiveness and Glasgow Liege Scale/Glasgow Coma Scale in an intensive care unit population. Neurocrit Care. 2011;15(3):447–53. 45. Slomine BS, Suskauer SJ, Nicholson R, Giacino JT. Preliminary validation of the coma recovery scale for pediatrics in typically developing young children. Brain Inj. 2019;33(13–14):1640–5.

32

C. Schnakers and K. O’Brien

46. Portaccio E, Morrocchesi A, Romoli AM, et al. Score on Coma Recovery Scale-Revised at admission predicts outcome at discharge in intensive rehabilitation after severe brain injury. Brain Inj. 2018;32(6):730–4. 47. Portaccio E, Morrocchesi A, Romoli AM, et al. Improvement on the Coma Recovery Scale-­ Revised during the first four weeks of hospital stay predicts outcome at discharge in intensive rehabilitation after severe brain injury. Arch Phys Med Rehabil. 2018;99(5):914–9. 48. Wannez S, Gosseries O, Azzolini D, et  al. Prevalence of Coma Recovery Scale-Revised signs of consciousness in patients in minimally conscious state. Neuropsychol Rehabil. 2018;28(8):1350–9. 49. Aubinet C, Cassol H, Bodart O, et  al. Simplified evaluation of CONsciousness disorders (SECONDs) in individuals with severe brain injury: a validation study. Ann Phys Rehabil Med. 2021;64(5):101432. 50. Sanz LRD, Aubinet C, Cassol H, et al. SECONDs administration guidelines: a fast tool to assess consciousness in brain-injured patients. J Vis Exp. 2021;168 https://doi.org/10.3791/61968-­v. 51. Whyte J, DiPasquale M, Vaccaro M. Assessment of command-following in minimally conscious brain injured patients. Arch Phys Med Rehabil. 1999;80:1–8. 52. Day KV, DiNapoli MV, Whyte J. Detecting early recovery of consciousness: a comparison of methods. Neuropsychol Rehabil. 2018;28(8):1233–41. 53. Schnakers C, Chatelle C, Vanhaudenhuyse A, et al. The nociception coma scale: a new tool to assess nociception in disorders of consciousness. Pain. 2010;148(2):215–9. 54. Chatelle C, Majerus S, Whyte J, et al. A sensitive scale to assess nociceptive pain in patients with disorders of consciousness. J Neurol Neurosurg Psychiatry. 2012;83(12):1233–7. 55. Chatelle C, De Val MD, Catano A, et  al. Is the nociception coma scale-revised a useful clinical tool for managing pain in patients with disorders of consciousness? Clin J Pain. 2016;32(4):321–6. 56. Bonin EAC, Lejeune N, Thibaut A, et  al. Nociception coma scale revised allows to identify patients with preserved neural basis for pain experience. J Pain. 2019. pii: S1526-5900(19)30063-X. 57. Bernard C, Delmas V, Duflos C, et al. Assessing pain in critically ill brain-injured patients: a psychometric comparison of 3 pain scales and videopupillometry. Pain. 2019;160(11):2535–43. 58. Poulsen I, Balle M, Givard KL. Nociception coma scale-revised: nurses’ experience in clinical practice. Pain Manag Nurs. 2019;20(6):592–8. 59. Vink P, Lucas C, Maaskant JM, et  al. Clinimetric properties of the nociception coma scale (-revised): a systematic review. Eur J Pain. 2017;21(9):1463–74. 60. Whyte J, Poulsen I, Ni P, et al. Development of a measure of nociception for patients with severe brain injury. Clin J Pain. 2020;36(4):281–8. 61. Schnakers C, Bessou H, Rubi-Fessen I, et al. Impact of aphasia on consciousness assessment: a cross-sectional study. Neurorehabil Neural Repair. 2015;29(1):41–7. 62. Murphy L. The cognitive assessment by visual election (CAVE): a pilot study to develop a cognitive assessment tool for people emerging from disorders of consciousness. Neuropsychol Rehabil. 2018;28(8):1275–84. 63. Aubinet C, Chatelle C, Gillet S, et al. The brief evaluation of receptive aphasia test for the detection of language impairment in patients with severe brain injury. Brain Inj. 2021;35(6):705–17. 64. Wade DT. How often is the diagnosis of the permanent vegetative state incorrect? A review of the evidence. Eur J Neurol. 2018;25(4):619–25. 65. Heine L, Tillmann B, Hauet M, et al. Effects of preference and sensory modality on behavioural reaction in patients with disorders of consciousness. Brain Inj. 2017;31(10):1307–11. 66. Sun Y, Wang J, Heine L, et al. Personalized objects can optimize the diagnosis of EMCS in the assessment of functional object use in the CRS-R: a double blind, randomized clinical trial. BMC Neurol. 2018;18(1):38. 67. Bareham CA, Allanson J, Roberts N, et  al. Longitudinal assessments highlight long-term behavioural recovery in disorders of consciousness. Brain Commun. 2019;1(1):fcz017.

2  Behavioral Assessment and Diagnosis of Disorders of Consciousness

33

68. Blume C, Lechinger J, Santhi N, et al. Significance of circadian rhythms in severely brain-­ injured patients: a clue to consciousness? Neurology. 2017;88(20):1933–41. 69. Monti MM. Cognition in the vegetative state. Annu Rev Clin Psychol. 2012;8:431–54.

Chapter 3

Brain–Computer Interfaces and Their Place in the Management of Disorders of Consciousness Michiel Meys, Aurore Thibaut, and Jitka Annen

Abstract  Brain–computer interfaces (BCI) constitute a growing and constantly evolving field of study showing promising applications that span a multitude of potential disciplines. In this chapter, we will introduce BCIs and the roles that different technologies and paradigms play specifically for the management of patients with a disorder of consciousness (DoC). We will provide an overview of the state of the art concerning BCI research in the field of DoC by highlighting some of the most significant works in the current literature. Contrasting the advances in research with current recommendations and applications in clinical practice exposes the severe lack of recognition that BCI usage receives in routine care for patients with a DoC. To conclude, we mention some potentially interesting future perspectives to further develop this domain.

 hat Are BCIs and How Can Patients with a Disorder W of Consciousness Benefit From Them? A brain–computer interface (BCI) is defined as a system allowing for communication between the brain and the external environment, independent from any peripheral neural or muscular pathways [1]. The basic principle is straightforward: volatile modulation of brain activity in response to a specific task (e.g., imagining a movement) is measured by one of many possible data acquisition techniques, processed accordingly with extraction of relevant features, and finally translated to a desired artificial output (Fig. 3.1). This direct link between the central nervous system and the user’s immediate surroundings creates the opportunity to gain insight into their cognition or to uncover intent, enabling users to communicate or use assistive technologies. Biomedical BCI applications often use this principle to bypass damaged M. Meys · A. Thibaut · J. Annen (*) Coma Science Group, GIGA-Consciousness, University of Liège, Liège, Belgium Centre du Cerveau2, University Hospital of Liège, Liège, Belgium e-mail: [email protected]; [email protected]; [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 C. Schnakers, S. Laureys (eds.), Coma and Disorders of Consciousness, https://doi.org/10.1007/978-3-031-50563-8_3

35

36

M. Meys et al.

Fig. 3.1  Schematic overview of a typical BCI setup. A certain paradigm is presented to the patient, from which the response can be measured as brain activity (through direct electrophysiological recordings (e.g., EEG) or as a metabolic proxy (e.g., hemodynamic response, fMRI, fNIRS)) or using other physiological signals (e.g., EMG, EOG, breathing). The resulting signal is processed accordingly, usually consisting of preprocessing to clean the data of artifacts, extraction of relevant stimulus-related features, and translation to a useful output. This control signal depends on the specific intended BCI use, which can range from low-level applications that inform about the user’s health to high-level applications of communication or control using assistive technologies. The output can additionally be reused as feedback directly to the patient or to dynamically adapt the paradigm, which then results in a closed-loop BCI. BCI brai n–computer interface, EEG electroencephalography, fMRI functional magnetic resonance imaging, fNIRS functional near-infrared spectroscopy, EMG electromyography, EOG electrooculography, SMR sensorimotor rhythms, SSVEP steady-state visually evoked potentials

motor pathways and to subsequently circumvent the associated impaired functionalities. Although the principle is simple, the patient populations that BCIs are designed for are quite heterogeneous in terms of brain damage and subsequent needs. There is no one-size-fits-all solution, leading to a plentitude of BCI acquisition, preprocessing, and analysis techniques, ideally tailored to the single patient. The eventual applicability of a BCI system is heavily dependent on the choice of an appropriate data acquisition technique. Numerous possible sensory modalities exist foremost with varying degrees of invasiveness [2]. The most direct way to measure brain signals is by implanting intracortical microelectrodes that pick up extracellular potentials in the immediate vicinity of neural populations of interest [3]. This technique allows for the capture of neural activity at virtually any desired location within the brain with very high precision. Partially invasive BCIs on the other hand rely on electrocorticography (ECoG), during which electrodes are placed on the brain surface to detect electrical activity originating from the cerebral cortex.

3  Brain–Computer Interfaces and Their Place in the Management of Disorders…

37

Both techniques measure in close proximity to the neural populations which generate the activity that they can perceive, which minimizes attenuations and distortion caused by propagation of the signals throughout multiple tissues. The resulting data consequently requires limited additional preprocessing due to the relative lack of noise (e.g., no eyeblink or muscle artifacts to remove, no need to correct for head movements). These recordings are of high quality, benefitting from a good spatial resolution (high precision, albeit never whole-brain) alongside an excellent temporal resolution in the order of milliseconds. The main downside of invasive techniques is undoubtedly the need for surgery, as craniotomy is required even for partially invasive interventions. The risk of complications such as infection or tissue damage, deteriorating effects such as rejection or encapsulation of the electrodes, and limited flexibility in terms of fine-tuning options following implantation dictates that invasive BCIs should only be considered when deemed overwhelmingly beneficial without (further) compromising the user’s health [4]. The aforementioned considerations alongside more general ethical issues associated with implantation of medical devices explain why the vast majority of BCI applications are strictly noninvasive. In this chapter, we specifically focus on BCIs for severely brain-injured patients with disorders of consciousness (DoC). These patients form a heterogeneous group who, following a period of coma, experience no or limited awareness of themselves or the environment. Broadly, the unresponsive wakefulness syndrome (UWS; sometimes also referred to as vegetative state (VS)) [5] and minimally conscious state (MCS) [6] can be distinguished. The former group is fully unaware despite periods of arousal, while the latter shows fluctuating awareness levels over time. MCS patients can be subcategorized in MCS− or MCS+, based on the absence or presence of preserved language processing [7] and more specifically command following, intelligible vocalization, and intentional communication [8]. Patients who recover functional communication or object use are considered emergent from MCS (EMCS) [6]. Although no longer considered to have a DoC by definition, they can greatly benefit from BCI applications to facilitate communication or in the form of assistive technology as they are often still severely disabled. Nevertheless, patients with a DoC are vulnerable as a result of their compromised health conditions, which is why most (if not all) BCI applications in this population exclusively rely on noninvasive data acquisition techniques. One of the earliest and undoubtedly the most famous application of a BCI paradigm in patients with DoC stems from 2006 and concerns the work of Owen and colleagues [9]. They were able to probe awareness in a patient diagnosed as unresponsive at the bedside through detection of reproducible responses to imagining of playing tennis and a spatial navigation task using functional magnetic resonance imaging (fMRI). fMRI scanners are, however, large, stationary, and very expensive, limiting their availability to hospital settings. One can also argue whether examples like these can truly be considered a BCI, since they lack the often-associated real-­ time aspect due to long acquisition times and the eventual analysis that occurs offline after the fact (even if online applications are possible, e.g., [10]). Another technique which relies on brain activity measured as the associated hemodynamic

38

M. Meys et al.

response is functional near-infrared spectroscopy (fNIRS). fNIRS and fMRI both estimate blood flow properties, the former based on changes in light absorption of (de)oxygenated blood (with a high temporal resolution) rather than the magnetic properties leveraged by the latter. Certain fNIRS-based BCI systems allow for direct visualization of the hemodynamic response, which shows whether users are correctly performing the presented task in real time [11]. fNIRS offers a lower spatial resolution compared to fMRI, and the number of applications in practice remains limited since these acquisition systems are not commonly available. Current BCI research focuses predominantly on electroencephalography (EEG), which captures electrical fields at the scalp level originating from the summated postsynaptic potentials of synchronously firing pyramidal neurons in the cortex. EEG provides a direct measure of brain activity equivalent to both discussed invasive techniques and shares their high temporal resolution, with the added value of being more easily applicable and having few contraindications, making it a valuable option for patients with a DoC. A tradeoff is again the significantly worse spatial resolution due to the presence of volume conduction and the overall noisier signal, resulting in the need for extensive preprocessing. fNIRS is an ideal candidate for multimodal use with EEG because of their complementary natures [12]. Another important aspect to consider is the fact that both these methods are limited to the cortex, while fMRI can reveal subcortical neural activity as well, such as in the parahippocampal gyrus for spatial navigation [9, 13]. Importantly, in this clinical population of patients with a DoC, BCIs have the capacity to be used as a method of assessment of (covert) consciousness, which once detected can be exploited to assess higher-level functions like communication. The following section will delve deeper into this aspect and the subsequent implications it has for diagnostic taxonomy.

 CIs Help to Refine the Taxonomy of Disorders B of Consciousness Behavioral assessment is currently still the gold standard for diagnosis of patients with a DoC in clinical practice, by evaluating auditory, visual, (oro-)motor, and communication abilities. Without the use of standardized scales such as the Coma Recovery Scale-Revised (CRS-R) [14], misdiagnosis percentages have been observed as high as 41% [15]. Regardless of sensitive behavioral scales, factors such as deafness, blindness, motor deficits, and fluctuations in arousal may hamper the patient’s ability to physically respond and thus mask signs of consciousness. It has recently become strikingly apparent that the underlying conscious state of patients with a DoC is not always unambiguously reflected by their behavioral profile at the bedside. This can go beyond inconsistencies caused by fluctuations in awareness, which should be accounted for through repetition of behavioral assessments [16]. Overt awareness, as quantified using methods such as the CRS-R or its

3  Brain–Computer Interfaces and Their Place in the Management of Disorders…

39

faster alternative SECONDs [17], is in certain cases an underestimation of a patient’s residual cognitive capabilities. Neurophysiological or neuroradiological evaluation can bypass the behavioral impairments that often lie at the root of this problem and can therefore provide a more accurate diagnosis. The emergence of BCI usage in DoC research has led to the conceptualization of covert awareness [18, 19], the presence of non-behavioral signs of consciousness as revealed by neuroimaging or electrophysiological paradigms. Notably, covert awareness as a concept is descriptive rather than diagnostic. Actual diagnostic terms have been used almost interchangeably throughout the literature. A recent systematic review showed that a consensus for an unambiguous nomenclature to define these clinical entities has yet to be reached [20]. The authors pointed out that prior use of the proposed taxonomies could be considered contradictory between certain studies. Moreover, one specific instance of covert awareness can often not be univocally represented by a single term since there is a considerable degree of overlap between definitions, and there does not seem to be a clear hierarchical structure either. Among the different terms proposed for covert awareness are MCS star or non-behavioral MCS (MCS*), cognitive motor dissociation (CMD), higher-order cortex motor dissociation (HMD), functional locked-in syndrome (LIS), and certain subcategories of the cortically mediated state (CMS) (Fig. 3.2). MCS* was introduced for patients who would be behaviorally diagnosed as UWS, while their residual underlying brain activity is more in line with MCS [21]. It reflects a broad categorization also including those with preserved brain activity during the resting state [19]. CMD usually describes the subsample of patients that exhibits covert awareness in the form of covert command following in response to active paradigms specifically [22]. HMD on the other hand describes patients who show no behavioral signs of language comprehension while nevertheless exhibiting brain responses to certain passive paradigms (e.g., sound or language) [23]. MCS* similarly encompasses CMD and HMD when considering unresponsive patients according to the aforementioned definitions. Yet another related categorization has been introduced based on behavior and neuroimaging. It considers MCS patients showing relative preservation of non-communicating behavior or brain activity in resting-state, passive, or active paradigms to be in a cortically mediated state or CMS, and patients showing communication at the bedside or using neuroimaging to be in a conscious state (CS) [24]. Finally, functional locked-in syndrome denoted the dissociation between patients’ motor dysfunction and their preserved higher cognitive functions as shown by the ability to communicate using functional imaging techniques only [7]. The use of this term has been criticized over the years due to the apparent association with LIS, which is by definition not a DoC and concerns only patients with a very distinct neuropathology [21]. As highlighted in a recent gap analysis paper on CMD conducted by the Coma Science Working Group, there is an urgent need to refine the terminology of these different states [25]. The increase in BCI and neuroimaging-supported research over the last decade has allowed us to obtain an idea of the overall occurrence of covert awareness in DoC, as several synthesizing works have since then shed light on the significant

40

M. Meys et al.

Fig. 3.2  Overview of the classical DoC diagnoses as determined through behavioral evaluation, their respective potential BCI uses, and the resulting refined diagnoses. Behavioral diagnoses with the CRS-R range from states closest to coma without awareness despite arousal (UWS), over minimal consciousness characterized by either preserved intentional behaviors (MCS−) or additional language-related capabilities (MCS+), to emergence from a DoC diagnosed through functional object use or functional communication (EMCS). UWS and MCS− patients can be categorized as either CMD or HMD depending on whether they show responses to active (e.g., command following) or passive (e.g., sound or language perception) paradigms, respectively. UWS patients who show preserved brain activity in line with MCS using any assessment or BCI technique, including resting-state electrophysiological or neuroimaging-based evaluations, are considered MCS*. Both MCS subcategorizations furthermore fall under CMS in case non-communicating behavior is determined by functional neuroimaging. For MCS+ patients, as well as EMCS patients for whom only object use is present, it might be possible to establish functional communication by means of active BCI paradigms. Such patients are subsequently considered to have functional LIS. The most high-level BCI uses are reserved for EMCS patients as they are able to benefit from assistive technologies (e.g., wheelchair control, Internet access) in case communication was already restored. All instances in which functional communication is present (including MCS+ and EMCS patients) comprise the CS. UWS unresponsive wakefulness syndrome, MCS minimally conscious state, EMCS emergence from MCS, CMD cognitive motor dissociation, HMD higher-order cortex motor dissociation, MCS* non-behavioral MCS, LIS locked-in syndrome, CMS cortically mediated state, CS conscious state

prevalence of this phenomenon. Using resting-state positron emission tomography (PET), Thibaut et al. recently demonstrated that a large percentage of UWS patients had residual brain metabolism compatible with the diagnosis of MCS, as much as 67% of the sample [19]. Perhaps more importantly, this study highlighted the prognostic implications of covert awareness, as MCS* patients presented a better outcome after 1 year. These results are in line with other studies [26]. BCI paradigms beyond resting-state examinations can be categorized into passive and active paradigms. Passive paradigms differ from resting-state alternatives in the sense that external stimuli are applied in order to elicit brain responses. These do not necessarily require active engagement but rather decode cognitive states from the subject’s

3  Brain–Computer Interfaces and Their Place in the Management of Disorders…

41

cerebral signals in a reactive manner. Responses are often indicative of neural processing related to sensory information which, according to several definitions of the phenomenon, is not always sufficient to infer the presence of consciousness. Nevertheless, they do provide interesting insights that could still be used to differentiate between clinical entities of DoC. Active paradigms encompass techniques that match closest to those observed in a typical BCI setup. These can often stem from a passive equivalent by instructing the subject to perform a specific task related to the stimulus (e.g., asking to count the occurrence of the subject’s own name (SON) rather than simply presenting it). Active paradigms are based on willful modulation of neural activity in response to a command and involve activation of higher-order cognitive processes, which is indicative of consciousness. A review of the literature showed that 15% of patients with a clinical diagnosis of UWS could willfully modulate their brain activity to follow commands as measured with EEG and fMRI [27]. More recent studies detected CMD in as high as a quarter of unresponsive patients, which was also correlated with better outcome [28]. These findings are undoubtedly most striking for completely unresponsive patients; however, an even higher proportion of respondents to both active and passive paradigms was found in behavioral MCS, as was later reinforced by a subsequent meta-­analysis [29]. Hence it seems that a considerable proportion of patients could benefit from both passive and active BCI technologies. In the next section we will review several BCI approaches developed and used in the field of DoC specifically.

 he Applications of BCIs in Research for Patients T with a Disorder of Consciousness Following the very first use of a BCI paradigm in the field of DoC, research over the subsequent years has been continuously investigating different techniques for their link to consciousness in addition to potential diagnostic and prognostic capabilities. The result is a substantial body of evidence comprising various study setups and findings, which has helped shape our current understanding of impaired consciousness. This section provides an overview of some of the most well-known and important works in this regard. BCI applications in practice can come in many forms regardless of whether they are to be presented in an active or passive manner. Each of these main paradigms will be illustrated, roughly ordered according to their prevalence within BCI research for patients with DoC.

42

M. Meys et al.

P3 Most of the research involving BCIs in the field of DoC is based on the P3, an event-­ related potential (ERP) component which can be observed in response to an oddball paradigm. It manifests as a positive infliction following an unexpected deviant stimulus in a sequence of regular counterparts with a higher probability of occurrence. The P3 consists of an early and late subcomponent, referred to as P3a and P3b, respectively [30]. The frontal P3a is thought to reflect exogenous attention and is elicited by stimuli processed in a bottom-up fashion, which can be task independent. The parietal P3b on the other hand suggests top-down cognitive processing of task-­ relevant features through endogenous attention. The BCI paradigms discussed next mostly rely on this latter component due to its relation to conscious processing [31] and will therefore be implied when referencing P3 unless explicitly stated otherwise. P3 responses in DoC patients differ from healthy subjects both in terms of increased latency and reduced amplitude [32, 33]. This same observation can be made when comparing UWS and MCS; however, this is not always sufficient to discriminate between the two [34]. Although all manifestations of the P3 can be ascribed to one and the same principle, the elicitation of this ERP component can be achieved by stimuli targeting different senses. Each of these has its respective advantages and disadvantages and is therefore fit to be employed in different types of applications. One of the major applications of the visual P3 is the spelling device. By presenting the subject with a grid of letters and symbols from which selections are alternatingly illuminated, it is possible to determine the character being focused on through detection of a P3 response. The efficacy of such visual speller design was shown in healthy subjects as well as LIS patients, reaching accuracies up to 90% and 70%, respectively [35]. This specific application is, however, already quite a high-level example, as it requires visual fixation, and its goal to reach communication is therefore primarily reserved for those in more advanced stages of recovery. The auditory P3 on the other hand is less prone to physical limitations beyond deafness. Most auditory oddball paradigms consist of only two stimuli (e.g., a standard and deviant one), although extension to more classes is certainly feasible. However, in a study implementing a four-choice paradigm, only one MCS patient out of 13 showed command following without functional communication, highlighting the need for sufficiently simplified tests [36]. The P3 component can furthermore be studied by means of the “local-global” protocol, which embeds two levels of auditory regularity, both within and across trials [37]. The presence of a global effect has been proposed as a signature of consciousness; however, this was found to be foremost true on an individual patient basis [38, 39]. The auditory P3 fails to reach the discriminatory power of other EEG-derived measures such as power or functional connectivity at the group level [40, 41]. Another commonly applied auditory oddball paradigm is that of the SON. Aside from the SON being deviant among unrelated names, the use of such a salient

3  Brain–Computer Interfaces and Their Place in the Management of Disorders…

43

stimulus should elicit a larger response. This paradigm can be presented as either an active or passive task by asking the subject to count their own name or by giving no further instructions instead. Multiple studies showed that the SON under active conditions evoked stronger responses in DoC patients [42, 43], although passive implementations can still lead to significant results by adopting an appropriate experimental design [44]. It should be noted that such paradigms are attention mediated and thus require cooperation from the subject. It is therefore crucial that patients are actively aroused when awareness seems lost. P3 responses can furthermore be elicited through (vibro)tactile stimulation. Instructing patients who showed no behavioral command following (8 UWS, 4 MCS−) to count the occurrence of vibrations administered at either the left or right hand established covert command following in one MCS− patient, as was confirmed by preserved glucose uptake in the language network using PET [45]. Repetition of BCI assessments using this paradigm is important to detect command following in unresponsive patients, which can even result in the establishment of binary communication in certain cases [46]. When both vibrotactile and auditory P3 paradigms are applied to patients with a DoC, performances were found to be independent of one another, exemplifying the usefulness of multimodal BCI assessments [47]. The combination of stimuli targeting different senses leads to the notion of hybrid BCIs, as will be discussed further.

Motor Imagery The seminal work of Owen and colleagues involving imagining playing tennis and spatial navigation is a prime example of motor imagery using fMRI [9]. Imagining certain tasks will activate associated brain regions, which can subsequently be visualized by functional neuroimaging. This exact paradigm was later used in a large cohort of 54 patients with a DoC, five of which could willfully modulate their brain activity to follow commands (3 MCS, 2 UWS) [13]. One MCS patient could even achieve binary communication as a result by associating playing tennis with “yes” and imagining navigating one’s house with “no.” The same concept is used in EEG as well and can be characterized by sensorimotor rhythms (SMR): oscillatory electrophysiological brain activity in the beta frequency range (13–35  Hz) that is associated with movement. Reduced SMRs, or event-related desynchronizations, are observed when a person prepares for or executes a motion, and more importantly, also when imagining doing so. The inverse effect occurs after the movement during relaxation as event-related synchronization [48]. When applied as a task involving the imagining of squeezing their hand or moving their toes to a sample of exclusively UWS patients, 19% responded and thus seemingly exhibited covert awareness [49]. It became apparent afterwards that such results should be interpreted carefully, since their conclusions were later refuted following reanalysis [50]. A slightly larger percentage of covert command following could be observed in MCS patients, namely 22% [51]. This study did not specify

44

M. Meys et al.

whether these patients were diagnosed as MCS− or MCS+, which would have revealed the prevalence of type II errors in this paradigm based on the number of negative respondents in the latter cohort. When applied to an acute cohort of 16 severely brain-injured patients in an intensive care setting, this exact paradigm using both EEG and fMRI showed that the former had a lower sensitivity (33.3% vs. 42.9%) but a higher specificity (100% vs. 50%) compared to the latter for detecting behavioral signs of language [23]. CMD was identified in four patients (3 UWS, 1 MCS−), while passive listening paradigms additionally revealed two instances of HMD (both MCS−). The following year, a study investigating patients with a prolonged DoC used a motor imagery task with four commands, consisting of “tennis,” “opening/closing hand,” “spatial navigation,” and “swimming” [52]. Evidence of the capacity to follow commands was found in 21 out of 28 patients based on EEG (3 UWS, 11 MCS, 7 EMCS), nine of whom also demonstrated similar evidence using fMRI (1 UWS, 5 MCS, 3 EMCS).

SSVEP Steady-state visually evoked potentials (SSVEP) are neural responses to periodic visual stimuli. Presenting a flickering stimulus will induce measurable rhythmic EEG patterns in the occipital brain region at that same frequency [53]. An SSVEP-­ based speller and P3-based alternative were used in a cohort of seven LIS patients to assess and compare performance [54]. The SSVEP variant resulted in more instances of high accuracy (at least 70%), that being all seven users rather than just three with P3, along with a lower mental workload and higher overall satisfaction. Such BCIs are characterized by fast response times and having a low susceptibility to noise. They furthermore require no training and can subsequently be used by many subjects. One potential drawback, however, is the paradigm’s usual reliance on shifts in gaze to express attention, thus requiring voluntary eye control. As a solution, Lesenfants and colleagues implemented a gaze-independent approach by presenting both stimulation frequencies in an overlapping grid pattern [55]. Two out of six LIS patients showed response to command by achieving offline accuracies above chance level, one out of four even being able to communicate online. Extension to more than two target classes is straightforward and leads to overall improved results, as illustrated in another cohort of five LIS patients [56]. Evoked responses can similarly be elicited using auditory and somatosensory stimuli (resulting in steady-state auditory (SSAEP) and somatosensory (SSSEP) evoked potentials, respectively), but are less commonly employed in practice [57]. One study investigated a combined SSSEP and P3 EEG paradigm with vibrotactile stimulation in a sample of 14 patients with a DoC, all of whom responded to the former but none to the latter [58]. Interestingly, a subsample of eight patients did show evidence of bottom-up attention (P3a response), who were the only ones to exhibit command following either behaviorally or through alternative fMRI paradigms. These findings suggest the relevance of P3a beyond unconscious processing,

3  Brain–Computer Interfaces and Their Place in the Management of Disorders…

45

while highlighting the non-specificity of SSSEPs and the problem of false negatives for the classical P3 paradigm. Auditory steady-state responses were more recently investigated for their diagnostic capacity. Passive listening to modulated tones in the low gamma frequency range (±40 Hz) correlated with behavioral scores and could consistently differentiate between UWS and MCS [59, 60].

Hybrid BCIs Rather than relying on one single signal, it is also possible to integrate multiple physiological measures into a hybrid BCI. Such a multimodal signal can consist of brain activity provoked by different paradigms (e.g., motor imagery and oddball task) or combined with a non-brain signal (e.g., ocular activity or heart rate). The different components of a hybrid BCI can be structured simultaneously to reinforce one another or sequentially to facilitate complementary actions (e.g., focusing on an item and the subsequent selection thereof) [61]. One of the most widely employed hybrid paradigms is a combined visual P3 and SSVEP, which can be realized by presenting images of target and non-target stimuli flickering at different frequencies. Command following could be revealed in approximately one third of patients with a DoC in response to familiar and unfamiliar faces [62], and similarly by assessing arithmetic abilities through number processing and mental calculation tasks [63]. The same principle was used more recently in a large cohort with the intention of detecting covert awareness [64]. CMD was determined for patients with BCI accuracies above chance level, which was apparent in 40% of UWS and 48% of MCS cases. The use of an asynchronous hybrid BCI, which gives the user more control by dynamically readjusting the window of opportunity to respond, enabled three out of seven MCS patients to achieve online binary communication while also improving behavioral scores [65]. What should be noted is that the hybrid BCIs in every single one of the aforementioned studies outperformed the separate paradigms, highlighting the added value of this seemingly more complex multimodal approach. Alongside the observation that even healthy people can outright fail to achieve proficiency with a certain paradigm (BCI illiteracy, in 15–30% of users [66]), some techniques might also not be applicable in patients with a DoC as a result of specific impairments (e.g., blindness or deafness). The use of hybrid BCIs can combat both issues by targeting multiple senses. Wang and colleagues first illustrated the improved discriminatory power of an audiovisual P3 BCI during detection of awareness, resulting in command following and number recognition for five out of seven patients with a DoC (1 UWS, 4 MCS+) [67]. Recent years have seen a substantial increase in audiovisual hybrid BCI implementations, including but not limited to: tools to supplement the CRS-R for assessment of communication abilities [68], gaze-independent auxiliary detection of awareness at the bedside [69], improved object recognition [70], and evaluation of sound localization [71]. Moreover, many studies seem to consider multiple control signals by default, which is expedited by

46

M. Meys et al.

the constant evolution of the field and the more advanced technologies that become available as a result.

Body–Computer Interfaces (Biofeedback Machines) As an alternative to BCIs using brain activity, it is furthermore possible to infer communication or command following from other types of physiological activity, presented here as body–computer interfaces. With electromyography (EMG, recording of muscle activity), it is possible to detect subliminal muscular responses as part of an active BCI paradigm. Bekinschtein and colleagues used EMG to reveal command following in one out of eight UWS and in both MCS patients enrolled in their study (MCS− and MCS+), suggesting its use in awareness detection [72]. A later study in a larger cohort showed similar results for UWS, with only one out of ten being able to respond [73]. However, the issue of false negatives associated with EMG was made apparent as well, since none of the eight MCS− and merely three out of 20 MCS+ patients had significant discernible responses to target commands. Finally, Lesenfants et al. proposed a novel methodology that evaluated responses on a single-trial basis to overcome the undesired influence of fluctuations in arousal and awareness in a total of 45 brain-injured patients [74]. This implementation illustrated command following in all LIS (n = 2), EMCS (n = 3), and MCS+ (n = 14) patients, with two out of eight MCS− patients showing an EMG response as well. One downside of EMG in this context is that the sensor is usually applied to a specific muscle, which is not necessarily a muscle that a patient still has some degree of control over. Muscular responses detected with EMG might furthermore be unreliable due to spastic paresis, a motor disorder extremely frequent in patients with a DoC [75]. As an alternative, it has been shown that LIS patients, for instance, could control a speller to write text through voluntary control of breath and sniffing [76]. This paradigm later proved unsuccessful when applied to UWS patients; however, it did enable one out of 14 MCS patients to follow commands without any further motor control [77]. Interestingly, this specific patient was one out of three included MCS− cases while none of the MCS+ patients could perform the paradigm, illustrating a 100% false negative rate in this group. A more recent study investigated the potential of olfactory function as a biomarker for consciousness and concluded that the sniff response could reliably discriminate between UWS and MCS at the group level [78]. As for clinical implications, the presence of this sniff response was found to be indicative of full recovery of consciousness at the single-­ patient level and associated with survival rates in the long term. The body–computer interface techniques mentioned up to this point for the most part still require residual voluntary control of the sensory modality in question to a certain extent. Also, they might be influenced by spontaneous movements or eyeblinks. Recently, it has been shown that volatile and non-intentional actions can be distinguished based on brain activity preceding the action [79]. By gaining a deeper understanding of these biomarkers of volition, false positives and false negatives

3  Brain–Computer Interfaces and Their Place in the Management of Disorders…

47

might be avoided. Besides, further encouragement of the use of paradigms that are not only motor-independent but that rely on completely involuntary processes is warranted. Pupillometry, for instance, can probe awareness by measuring subtle changes in pupil diameter associated with cognitively demanding mental tasks. The effectiveness of this paradigm has been proven by establishing binary communication in LIS patients, going as far as revealing command following in an MCS patient [80]. Salivary pH has been successfully used for this same purpose as well, during which an LIS patient had to imagine the taste of either a lemon or milk [81]. Aside from the handful of non-brain activity-based BCI instances mentioned here, there is an apparent lack of further application beyond one-time implementations for research purposes. The nature of these techniques gives room to substantially more degrees of freedom and the subsequent increased need for standardization. As they are currently nowhere near being part of routine clinical practice, they are rarely mentioned by international guidelines, if at all. The following section will therefore go over clinical recommendations regarding the more usual BCI realizations instead, as these become increasingly common in the management of patients with a DoC.

 ecommendations for BCI Use According to Current R Guidelines for Clinical Management of Patients with a Disorder of Consciousness Despite the limited but steadily increasing amount of BCI research involving patients with a DoC, its role as part of routine clinical practice is still not nearly as established as behavioral or resting-state evaluations. In an effort to further promote the integration of BCIs in this field while also illustrating the present state of the art, we provide an overview of current recommendations as published in several recent clinical guidelines. Most of these concern neurophysiological techniques in the broader sense, which relate to BCI paradigms either directly or indirectly and can therefore be extrapolated to fit the narrative of this chapter, given that they might facilitate the detection of covert awareness. The guidelines in question consist of synthesizing works drafted by the European Academy of Neurology [82], the American Academy of Neurology [83], and the UK Royal College of Physicians [84] regarding the use of resting-state, passive, and active paradigms to diagnose patients with a DoC. Regarding neurophysiological examination in general, the consensus is predominantly positive toward the insights it provides into DoC as well as the management of affected patients. EU guidelines advocate for multimodal evaluations that integrate the current standardized clinical methods with EEG-based techniques and functional neuroimaging, where all approaches hold an equal weight in categorizing states of consciousness. The importance of avoiding misdiagnosis and uncovering covert awareness is hereby especially highlighted. They suggest resting-state fMRI

48

M. Meys et al.

and PET to complement behavioral evaluation and strongly recommend standard clinical EEG to rule out confounding factors that could affect consciousness (e.g., non-convulsive status epilepticus), albeit primarily through qualitative visual inspection. US guidelines recommend incorporation of functional imaging or electrophysiological studies in case of confounders for behavioral evaluation (e.g., brain injury-­ related sequelae such as severe hypertonus) or persistent ambiguity despite serial behavioral evaluations, in which auxiliary assessment may lead to an alternate diagnosis. They do, however, asterisk this by stating that there is insufficient evidence to conclusively support or refute such techniques as clinically useful adjuncts to current established methods of awareness detection (e.g., behavioral evaluation). Functional neuroimaging is furthermore not widely available and may not be clinically feasible in a significant proportion of patients. UK guidelines go one step further and explicitly state that advanced neuroimaging techniques and electrophysiology, as opposed to visual analysis of EEG or structural imaging by means of computed tomography or MRI, might not be considered as part of routine clinical evaluation for patients with prolonged DoC. Alongside the arguments made by the US guidelines, this conclusion is based on the current lack of interpretability as well as ethical considerations due to the lack of access and uncertainty of their prognostic implications. Despite this reluctance to acknowledge the clinical significance of any functional imaging technique, the UK guidelines do recognize the potentially greater clinical applicability of task-free examinations in nonresearch settings. The use of PET is specifically mentioned since prior research investigating metabolic brain activity has resulted in accurate outcome prediction [85]. EU guidelines limit their recommendation of passive fMRI paradigms to research protocols because of limited effects and considerable heterogeneity. They do, however, encourage the use of salient stimuli and/or familiar activities to increase sensitivity in both active and passive paradigms when examining patients with a DoC.  Passive EEG paradigms, including cognitive evoked potentials (e.g., P3), might be considered as part of multimodal assessment. The value that they exhibit for differentiating UWS from MCS patients is, however, accompanied by low sensitivity even in healthy controls due to the need for attention, calling for the use of advanced statistical and analysis techniques. US guidelines do not recommend or refute passive fMRI paradigms for diagnostic purposes based on a single study matching their inclusion criteria which indicated limited effectiveness [86]. They do recognize the probable prognostic utility of both passive fMRI (activation of auditory association cortex) and EEG (presence of P3) paradigms using the SON presented by a familiar voice, since these were associated with increased chances of recovering consciousness and favorable outcomes [87, 88]. Passive fMRI paradigms fall under advanced neurophysiological examinations according to UK guidelines and are therefore inherently not considered as part of routine clinical evaluation of patients with a DoC. EEG sensory-evoked potentials are only deemed useful when investigating the integrity of associated pathways in case no visual or auditory startle is discernible, rather than for the purpose of detecting signs of awareness. Cognitive evoked potentials (including P3) on the other hand

3  Brain–Computer Interfaces and Their Place in the Management of Disorders…

49

should be able to distinguish between levels of DoC [89] but are held back by their poor predictive value compared to standard EEG visual inspection and reactivity [90]. EU guidelines suggest that active fMRI paradigms should be considered as part of multimodal assessment in patients without command following at the bedside. A similar recommendation is given for active paradigms based on either standard or high-density EEG since these equally allow for identification of patients who present CMD. It follows that both active fMRI and EEG paradigms have a high specificity but very low sensitivity for detection of covert awareness. The absence of command following should therefore not necessarily imply the absence of consciousness. Consequently, they call for further refinement of the framework in which these techniques will be used for future research and clinical implementations. US guidelines do not recommend active fMRI paradigms when executed in the form of a word-counting task based on a study suggesting its inability to distinguish UWS from MCS [91], while no conclusive advice is given for motor imagery due to a lack of evidence [92]. Active EEG paradigms were not considered. They are, however, cautiously optimistic about the diagnostic value of EMG to detect command following as it could differentiate between UWS and MCS in multiple instances, evidently after adjusting for involuntary movements [73, 74]. UK guidelines for the most part disregard sophisticated neurophysiological techniques, which by default include active BCI paradigms, mainly due to the lack of availability. They reinforce this judgment by referring to the false negative findings of fMRI motor imaging tasks that arise even in healthy controls, the clinical significance of which has not been sufficiently established. To conclude, it seems there are important differences between the guidelines and attitudes toward neuroimaging-based assessment (and therefore toward BCIs as well) of patients with a DoC. EU and US guidelines are generally positive toward the possibility of supplementing behavioral evaluations with resting-state neuroimaging assessments, especially in case of physical limitations. Passive paradigms are not recommended or refuted by either EU or US guidelines. According to the EU guidelines, active paradigms could be a helpful tool in patients without behavioral command following, while the US guidelines are positive toward EMG to assess covert command following. UK guidelines are quite skeptical toward any application of neuroimaging in patients with a DoC. It is apparent that important steps in the direction of improving clinical care of DoC patients have already been made; however, there exist several important avenues for future research.

 uture Research and Clinical Directions to Encourage F Development and Implementation of BCIs for Patients with a Disorder of Consciousness Although individual studies have shown impressive results and hold great promise for clinical implementation, there is a large heterogeneity in experimental setup and subsequent success rates which cannot be overlooked. The lack of standardization in the field is likely the main reason for the conservative attitude toward clinical

50

M. Meys et al.

integration of BCI-based assessment. This leads to a substantial gap between scientific advancements and clinical availability and applicability. Standardization of data acquisition and analysis should be invested in to compile convincing evidence for the day-to-day usage of these technologies (e.g., by ensuring that single studies are not overfitting the data). Only then can the clinical usage of state-of-the-art techniques be promoted. Recent efforts to define common data elements (e.g., through the use of dedicated case report forms describing all information that needs to be collected and reported) for neuroimaging in patients with a DoC are a good step in that direction [93]. On the data analysis side, standardized and ready-to-apply pipelines should be made widely available to facilitate clinical implementation in non-­ expert centers [94]. Despite the scarce everyday use of these technologies, it is of utmost importance to be prepared for the ethical considerations that assessment using BCI technologies will undoubtedly raise. In other populations, several concerns regarding personhood, stigma, autonomy, privacy, research ethics, safety, responsibility, and justice have been identified [95], once more exemplifying the need for proper recommendations and regulations. BCI-based assessment in patients with a DoC specifically introduces an additional major concern, namely the question of the presence of awareness or the lack thereof, especially in cases where overt awareness is lacking [96]. This in turn leads to another important issue: can we trust the machine? While BCIs can certainly contribute to the clinical care and acceptance of the patient’s current cognitive state by family members, several potential negative effects become apparent as well. Underestimating the level of consciousness as determined by a BCI (false negative results) would induce false despair, while overestimation (false positive results) would evoke false hope and unrealistic expectations for patients’ caregivers and loved ones [97]. Several approaches exist to reduce these false positive and negative results as well as overfitting. First, it is important to define and use proper benchmark populations to test assessment and BCI systems initially [98]. The choice of healthy volunteers as the control group could be suboptimal, as they did not suffer severe brain injury and are therefore not immediately comparable. The inclusion of LIS patients might be the better solution but is more challenging to implement in practice. In reality, not all studies include a control group at all, which is problematic as the false positive and negative rates in a conscious population then remain unknown. Second, the use of proper statistics is very important as well. This was illustrated by Goldfine and colleagues, who showed that using an appropriate methodological approach produced results that could no longer be deemed significant, effectively refuting the apparent observed responses [50]. The choice of a well-suited statistical test can lead to unbiased estimations of significance and provide a robust interpretation of results, irrespective of the applied validation schemes [99]. However, statistical procedures that are too strict might also be harmful by increasing the type II error rates and potentially underestimating the patient’s level of consciousness, which is the foremost reason to perform assessments and BCI sessions in patients with DoC in the first place.

3  Brain–Computer Interfaces and Their Place in the Management of Disorders…

51

Aside from these technical validations and approaches to avoid false results, clinical safeguards can be put in place to obtain the most accurate findings. Resonating with other literature reviewed above, it is important to note that some patients perform well at one BCI assessment while failing to do so with another (e.g., [47]). Ideally, multiple BCIs are tested for the same patient to identify a technology that aligns with their cognitive and physical ability to avoid false negatives. Likewise, the arousal fluctuations frequently observed in DoC (e.g., with EEG [100]) lead to a behavioral underestimation of consciousness if patients are not assessed at least five times within 10 days [16]. Following this literature, it would be best if BCIs were tested multiple times before accepting negative results. However, because of the longer preparation times among other reasons, BCI systems are usually tested only once. The use of closed-loop BCI systems that can track patients’ arousal levels and solely assess at suitable moments could help to overcome this current limitation, similar to closed-loop systems that trigger a treatment based on the patient’s level of vigilance [101]. To minimize chances of obtaining false positive BCI results, one option would be to cross-check these with other neuroimaging modalities to establish whether the neural substrates required for the specific BCI are indeed still intact [45]. The importance of BCIs in leading to the discovery of covert awareness cannot be understated. How this should be implemented in clinical practice, however, is currently still unclear. As active assessments might be more prone to false negative results, they could underestimate the degree to which a patient is conscious. Hence, active paradigms might be best preserved for communication or control applications. Passive paradigms on the other hand would be better suited for assessment. In patients performing well at passive paradigms, active tasks could be tried next to evaluate potential further diagnostic improvements. One open question in the field remains how the clinical management of patients with an improved diagnosis based on neuroimaging should change. One could argue that treatment should be made readily available for these patients; however, it is unclear if and how these expensive options would translate to increased welfare and quality of life [102]. In comatose patients, the added value of BCI tools for clinical treatment is more straightforward. BCI applications are becoming more frequently tested in acute settings, showing that up to 30% of the patients are covertly aware already in the intensive care unit, including patients in a coma [103]. Although a negative test result is not a vote in favor of ceasing life-sustaining treatment in these patients, it should influence pain management and medical decision making for those who are found to be covertly aware. Despite the current state of the art not yet being translated to the clinic, some future perspectives can already be discussed. Over recent years, new treatment options have been identified for patients with a DoC [104]. These can potentially improve the behavioral diagnosis and underlying physiology of a selection of patients, such as an UWS patient who regained command following after transcranial direct current stimulation (tDCS) and showed activation with fMRI mental imagery [105], or strictly lead to physiological improvements [106]. In the latter group, it remains to be investigated whether the lack of behavioral improvement is

52

M. Meys et al.

a result of the physiological changes being unrelated to consciousness, or due to physical limitations of the patient. The use of additional techniques to prime the brain to a better attentional state and to ensure that patients are optimally arousable can in turn reveal more signs of consciousness. For example, tDCS has shown its potential to effectively modulate cortical excitability in patients with a DoC and could therefore allow for easier detection of changes in brain states as leveraged by the BCI [107]. Presenting preferred music on the other hand showed its beneficial effect by increasing responses to the SON paradigm [108]. The plentitude of potential extensions, improvements, and more advanced or elegant assessment and BCI tools hold promise for bringing these techniques to the patient’s bedside in the near future.

Conclusion BCIs to detect covert awareness have been developed and tested in research settings with varying success. They should be simple and easy to use, something that is sometimes overlooked during the development of demanding BCI systems with more complicated interactions. Besides the technical advancements, repeated BCI assessments are indicated to reduce false negative results. Appropriate statistical approaches, not too strict or too liberal, should be adopted to minimize false negative and false positive findings. Standardization and replication of approaches is needed to increase confidence in these techniques and to better assess their clinical usability. Once these conditions are met, the application of assessment and BCI tools might evolve to be widely recommended by clinical guidelines worldwide. Such advancement would facilitate clinical translation and BCI use might become more standard practice. However, it is important that the ethical aspects of BCI implementation in clinics (e.g., how clinical management should change after assessment) are mapped and addressed.

References 1. Wolpaw JR, Birbaumer N, McFarland DJ, Pfurtscheller G, Vaughan TM.  Brain–computer interfaces for communication and control. Clin Neurophysiol. 2002;113:767–91. 2. Martini ML, Oermann EK, Opie NL, Panov F, Oxley T, Yaeger K.  Sensor modalities for brain-computer interface technology: a comprehensive literature review. Clin Neurosurg. 2020;86:E108–17. 3. Wang Y, Yang X, Zhang X, Wang Y, Pei W. Implantable intracortical microelectrodes: reviewing the present with a focus on the future. Microsyst Nanoeng. 2023;9:1–17. 4. Zhao ZP, Nie C, Jiang CT, Cao SH, Tian KX, Yu S, Gu JW. Modulating brain activity with invasive brain–computer interface: a narrative review. Brain Sci. 2023;13:134. 5. Laureys S, Celesia GG, Cohadon F, et al. Unresponsive wakefulness syndrome: a new name for the vegetative state or apallic syndrome. BMC Med. 2010;8:1–4.

3  Brain–Computer Interfaces and Their Place in the Management of Disorders…

53

6. Giacino JT, Ashwal S, Childs N, et  al. The minimally conscious state. Neurology. 2002;58:349–53. 7. Bruno MA, Vanhaudenhuyse A, Thibaut A, Moonen G, Laureys S. From unresponsive wakefulness to minimally conscious PLUS and functional locked-in syndromes: recent advances in our understanding of disorders of consciousness. J Neurol. 2011;258:1373–84. 8. Thibaut A, Bodien YG, Laureys S, Giacino JT. Minimally conscious state “plus”: diagnostic criteria and relation to functional recovery. J Neurol. 2020;267:1245–54. 9. Owen AM, Coleman MR, Boly M, Davis MH, Laureys S, Pickard JD. Detecting awareness in the vegetative state. Science (1979). 2006;313:1402. 10. Sorger B, Goebel R.  Real-time fMRI for brain-computer interfacing. Handb Clin Neurol. 2020;168:289–302. 11. Benitez-Andonegui A, Burden R, Benning R, Möckel R, Lührs M, Sorger B. An augmented-­ reality fNIRS-based brain-computer interface: a proof-of-concept study. Front Neurosci. 2020;14:502570. 12. Li R, Yang D, Fang F, et al. Concurrent fNIRS and EEG for brain function investigation: a systematic, methodology-focused review. Sensors. 2022;22:5865. 13. Monti MM, Vanhaudenhuyse A, Coleman MR, Boly M, Pickard JD, Tshibanda L, Owen AM, Laureys S. Willful modulation of brain activity in disorders of consciousness. N Engl J Med. 2010;362:579–89. 14. Giacino JT, Kalmar K, Whyte J. The JFK Coma Recovery Scale-Revised: measurement characteristics and diagnostic utility. Arch Phys Med Rehabil. 2004;85:2020–9. 15. Schnakers C, Vanhaudenhuyse A, Giacino J, Ventura M, Boly M, Majerus S, Moonen G, Laureys S. Diagnostic accuracy of the vegetative and minimally conscious state: clinical consensus versus standardized neurobehavioral assessment. BMC Neurol. 2009;9:1–5. 16. Wannez S, Heine L, Thonnard M, Gosseries O, Laureys S.  The repetition of behavioral assessments in diagnosis of disorders of consciousness. Ann Neurol. 2017;81:883–9. 17. Aubinet C, Cassol H, Bodart O, et al. Simplified evaluation of CONsciousness disorders (SECONDs) in individuals with severe brain injury: a validation study. Ann Phys Rehabil Med. 2021;64:101432. 18. Casali AG, Gosseries O, Rosanova M, et  al. A theoretically based index of consciousness independent of sensory processing and behavior. Sci Transl Med. 2013;5:198ra105. https:// doi.org/10.1126/scitranslmed.3006294. 19. Thibaut A, Panda R, Annen J, et al. Preservation of brain activity in unresponsive patients identifies MCS star. Ann Neurol. 2021;90:89–100. 20. Schnakers C, Bauer C, Formisano R, et al. What names for covert awareness? A systematic review. Front Hum Neurosci. 2022;16:971315. 21. Gosseries O, Zasler ND, Laureys S. Recent advances in disorders of consciousness: focus on the diagnosis. Brain Inj. 2014;28:1141–50. https://doi.org/10.3109/02699052.2014.920522. 22. Schiff ND.  Cognitive motor dissociation following severe brain injuries. JAMA Neurol. 2015;72:1413–5. 23. Edlow BL, Chatelle C, Spencer CA, et al. Early detection of consciousness in patients with acute severe traumatic brain injury. Brain. 2017;140:2399–414. 24. Naccache L. Minimally conscious state or cortically mediated state? Brain. 2018;141:949–60. 25. Claassen J, Kondziella D, Alkhachroum A, et al. Cognitive motor dissociation: gap analysis and future directions. Neurocrit Care. 2023;2023:1–18. 26. Pan J, Xie Q, Qin P, et al. Prognosis for patients with cognitive motor dissociation identified by brain-computer interface. Brain. 2020;143:1177–1189. 27. Kondziella D, Friberg CK, Frokjaer VG, Fabricius M, Møller K.  Preserved consciousness in vegetative and minimal conscious states: systematic review and meta-analysis. J Neurol Neurosurg Psychiatry. 2016;87:485–92. 28. Egbebike J, Shen Q, Doyle K, et  al. Cognitive-motor dissociation and time to functional recovery in patients with acute brain injury in the USA: a prospective observational cohort study. Lancet Neurol. 2022;21:704–13.

54

M. Meys et al.

29. Schnakers C, Hirsch M, Noé E, et al. Covert cognition in disorders of consciousness: a meta-­ analysis. Brain Sci. 2020;10:930. 30. Comerchero MD, Polich J.  P3a and P3b from typical auditory and visual stimuli. Clin Neurophysiol. 1999;110:24–30. 31. Dehaene S, Changeux JP. Experimental and theoretical approaches to conscious processing. Neuron. 2011;70:200–27. 32. Perrin F, Schnakers C, Schabus M, et  al. Brain response to one’s own name in vegetative state, minimally conscious state, and locked-in syndrome. Arch Neurol. 2006;63:562–9. 33. Schnakers C, Perrin F, Schabus M, et  al. Voluntary brain processing in disorders of consciousness. Neurology. 2008;71:1614–20. 34. Real RGL, Veser S, Erlbeck H, Risetti M, Vogel D, Müller F, Kotchoubey B, Mattia D, Kübler A.  Information processing in patients in vegetative and minimally conscious states. Clin Neurophysiol. 2016;127:1395–402. 35. Ortner R, Aloise F, Prückl R, Schettini F, Putz V, Scharinger J, Opisso E, Costa U, Guger C. Accuracy of a P300 speller for people with motor impairments: a comparison. Clin EEG Neurosci. 2011;42:214–8. https://doi.org/10.1177/155005941104200405. 36. Lulé D, Noirhomme Q, Kleih SC, et al. Probing command following in patients with disorders of consciousness using a brain–computer interface. Clin Neurophysiol. 2013;124:101–6. 37. Bekinschtein TA, Dehaene S, Rohaut B, Tadel F, Cohen L, Naccache L. Neural signature of the conscious processing of auditory regularities. Proc Natl Acad Sci U S A. 2009;106:1672–7. 38. Faugeras F, Rohaut B, Weiss N, et al. Probing consciousness with event-related potentials in the vegetative state. Neurology. 2011;77:264–8. 39. Faugeras F, Rohaut B, Weiss N, et al. Event related potentials elicited by violations of auditory regularities in patients with impaired consciousness. Neuropsychologia. 2012;50:403–18. 40. Sitt JD, King JR, El Karoui I, Rohaut B, Faugeras F, Gramfort A, Cohen L, Sigman M, Dehaene S, Naccache L.  Large scale screening of neural signatures of consciousness in patients in a vegetative or minimally conscious state. Brain. 2014;137:2258–70. 41. Engemann DA, Raimondo F, King JR, et al. Robust EEG-based cross-site and cross-protocol classification of states of consciousness. Brain. 2018;141:3179–92. 42. Schnakers C, Giacino JT, Løvstad M, Habbal D, Boly M, Di H, Majerus S, Laureys S.  Preserved covert cognition in noncommunicative patients with severe brain injury? Neurorehabil Neural Repair. 2015;29:308–17. 43. Hauger SL, Schnakers C, Andersson S, Becker F, Moberget T, Giacino JT, Schanke AK, Løvstad M.  Neurophysiological indicators of residual cognitive capacity in the minimally conscious state. Behav Neurol. 2015;2015:1. https://doi.org/10.1155/2015/145913. 44. Kempny AM, James L, Yelden K, Duport S, Farmer SF, Diane Playford E, Leff AP. Patients with a severe prolonged disorder of consciousness can show classical EEG responses to their own name compared with others’ names. Neuroimage Clin. 2018;19:311–9. 45. Annen J, Blandiaux S, Lejeune N, Bahri MA, Thibaut A, Cho W, Guger C, Chatelle C, Laureys S. BCI performance and brain metabolism profile in severely brain-injured patients without response to command at bedside. Front Neurosci. 2018;12:364432. 46. Guger C, Spataro R, Pellas F, et al. Assessing command-following and communication with vibro-tactile P300 brain-computer interface tools in patients with unresponsive wakefulness syndrome. Front Neurosci. 2018;12:359448. 47. Annen J, Mertel I, Xu R, Chatelle C, Lesenfants D, Ortner R, Bonin EAC, Guger C, Laureys S, Müller F.  Auditory and somatosensory P3 are complementary for the assessment of patients with disorders of consciousness. Brain Sci. 2020;10:748. 48. Pfurtscheller G, Lopes Da Silva FH. Event-related EEG/MEG synchronization and desynchronization: basic principles. Clin Neurophysiol. 1999;110:1842–57. 49. Cruse D, Chennu S, Chatelle C, Bekinschtein TA, Fernández-Espejo D, Pickard JD, Laureys S, Owen AM. Bedside detection of awareness in the vegetative state: a cohort study. Lancet. 2011;378:2088–94.

3  Brain–Computer Interfaces and Their Place in the Management of Disorders…

55

50. Goldfine AM, Bardin JC, Noirhomme Q, Fins JJ, Schiff ND, Victor JD.  Reanalysis of “Bedside detection of awareness in the vegetative state: a cohort study”. Lancet. 2013;381:289–91. 51. Cruse D, Chennu S, Chatelle C, Fernández-Espejo D, Bekinschtein TA, Pickard JD, Laureys S, Owen AM. Relationship between etiology and covert cognition in the minimally conscious state. Neurology. 2012;78:816–22. 52. Curley WH, Forgacs PB, Voss HU, Conte MM, Schiff ND. Characterization of EEG signals revealing covert cognition in the injured brain. Brain. 2018;141:1404–21. 53. Vialatte FB, Maurice M, Dauwels J, Cichocki A.  Steady-state visually evoked potentials: focus on essential paradigms and future perspectives. Prog Neurobiol. 2010;90:418–38. 54. Combaz A, Chatelle C, Robben A, Vanhoof G, Goeleven A, Thijs V, Van Hulle MM, Laureys S. A comparison of two spelling brain-computer interfaces based on visual P3 and SSVEP in locked-in syndrome. PLoS One. 2013;8:e73691. 55. Lesenfants D, Habbal D, Lugo Z, et al. An independent SSVEP-based brain–computer interface in locked-in syndrome. J Neural Eng. 2014;11:035002. 56. Hwang HJ, Han CH, Lim JH, Kim YW, Choi SI, An KO, Lee JH, Cha HS, Hyun Kim S, Im CH. Clinical feasibility of brain-computer interface based on steady-state visual evoked potential in patients with locked-in syndrome: case studies. Psychophysiology. 2017;54:444–51. 57. Ahn S, Kim K, Jun SC.  Steady-state somatosensory evoked potential for brain-computer interface-present and future. Front Hum Neurosci. 2016;9:179038. 58. Gibson RM, Chennu S, Fernández-Espejo D, Naci L, Owen AM, Cruse D. Somatosensory attention identifies both overt and covert awareness in disorders of consciousness. Ann Neurol. 2016;80:412–23. 59. Górska U, Binder M. Low- and medium-rate auditory steady-state responses in patients with prolonged disorders of consciousness correlate with Coma Recovery Scale-Revised score. Int J Psychophysiol. 2019;144:56–62. 60. Binder M, Górska U, Pipinis E, Voicikas A, Griskova-Bulanova I.  Auditory steady-state response to chirp-modulated tones: a pilot study in patients with disorders of consciousness. Neuroimage Clin. 2020;27:102261. 61. Pfurtscheller G, Allison BZ, Bauernfeind G, Brunner C, Solis Escalante T, Scherer R, Zander TO, Mueller-Putz G, Neuper C, Birbaumer N. The hybrid BCI. Front Neurosci. 2010;4:1283. 62. Pan J, Xie Q, He Y, Wang F, Di H, Laureys S, Yu R, Li Y. Detecting awareness in patients with disorders of consciousness using a hybrid brain–computer interface. J Neural Eng. 2014;11:056007. 63. Li Y, Pan J, He Y, Wang F, Laureys S, Xie Q, Yu R. Detecting number processing and mental calculation in patients with disorders of consciousness using a hybrid brain-computer interface system. BMC Neurol. 2015;15 https://doi.org/10.1186/S12883-­015-­0521-­Z. 64. Pan J, Xie Q, Qin P, et al. Prognosis for patients with cognitive motor dissociation identified by brain-computer interface. Brain. 2020;143:1177–89. 65. Huang J, Qiu L, Lin Q, et al. Hybrid asynchronous brain–computer interface for yes/no communication in patients with disorders of consciousness. J Neural Eng. 2021;18:056001. 66. Vidaurre C, Blankertz B. Towards a cure for BCI illiteracy. Brain Topogr. 2010;23:194–8. 67. Wang F, He Y, Pan J, Xie Q, Yu R, Zhang R, Li Y. A novel audiovisual brain-computer interface and its application in awareness detection. Sci Rep. 2015;5:1–12. 68. Wang F, He Y, Qu J, et al. Enhancing clinical communication assessments using an audiovisual BCI for patients with disorders of consciousness. J Neural Eng. 2017;14:046024. 69. Xie Q, Pan J, Chen Y, He Y, Ni X, Zhang J, Wang F, Li Y, Yu R. A gaze-independent audiovisual brain-computer interface for detecting awareness of patients with disorders of consciousness. BMC Neurol. 2018;18:1–12. 70. Wang F, He Y, Qu J, Cao Y, Liu Y, Li F, Yu Z, Yu R, Li Y. A brain-computer interface based on three-dimensional stereo stimuli for assisting clinical object recognition assessment in patients with disorders of consciousness. IEEE Trans Neural Syst Rehabil Eng. 2019;27:507–13.

56

M. Meys et al.

71. Xiao J, He Y, Yu T, et al. Toward assessment of sound localization in disorders of consciousness using a hybrid audiovisual brain-computer interface. IEEE Trans Neural Syst Rehabil Eng. 2022;30:1422–32. 72. Bekinschtein TA, Coleman MR, Niklison J, Pickard JD, Manes FF. Can electromyography objectively detect voluntary movement in disorders of consciousness? J Neurol Neurosurg Psychiatry. 2008;79:826–8. 73. Habbal D, Gosseries O, Noirhomme Q, Renaux J, Lesenfants D, Bekinschtein TA, Majerus S, Laureys S, Schnakers C. Volitional electromyographic responses in disorders of consciousness. Brain Inj. 2014;28:1171–9. https://doi.org/10.3109/02699052.2014.920519. 74. Lesenfants D, Habbal D, Chatelle C, Schnakers C, Laureys S, Noirhomme Q.  Electromyographic decoding of response to command in disorders of consciousness. Neurology. 2016;87:2099–107. 75. Thibaut A, Chatelle C, Wannez S, Deltombe T, Stender J, Schnakers C, Laureys S, Gosseries O.  Spasticity in disorders of consciousness: a behavioral study. Eur J Phys Rehabil Med. 2015;51:389. 76. Plotkin A, Sela L, Weissbrod A, Kahana R, Haviv L, Yeshurun Y, Soroker N, Sobel N. Sniffing enables communication and environmental control for the severely disabled. Proc Natl Acad Sci U S A. 2010;107:14413–8. 77. Charland-Verville V, Lesenfants D, Sela L, Noirhomme Q, Ziegler E, Chatelle C, Plotkin A, Sobel N, Laureys S. Detection of response to command using voluntary control of breathing in disorders of consciousness. Front Hum Neurosci. 2014;8:123178. 78. Arzi A, Rozenkrantz L, Gorodisky L, et al. Olfactory sniffing signals consciousness in unresponsive patients with brain injuries. Nature. 2020;581:428–33. 79. Derchi CC, Mikulan E, Mazza A, Casarotto S, Comanducci A, Fecchio M, Navarro J, Devalle G, Massimini M, Sinigaglia C.  Distinguishing intentional from nonintentional actions through EEG and kinematic markers. Sci Rep. 2023;13:1–10. 80. Stoll J, Chatelle C, Carter O, Koch C, Laureys S, Einhäuser W. Pupil responses allow communication in locked-in syndrome patients. Curr Biol. 2013;23:R647–8. 81. Wilhelm B, Jordan M, Birbaumer N. Communication in locked-in syndrome: effects of imagery on salivary pH. Neurology. 2006;67:534–5. 82. Kondziella D, Bender A, Diserens K, et  al. European academy of neurology guideline on the diagnosis of coma and other disorders of consciousness. Eur J Neurol. 2020;27:741–56. 83. Giacino JT, Katz DI, Schiff ND, et al. Practice guideline update recommendations summary: disorders of consciousness. Neurology. 2018;91:450–60. 84. Royal College of Physicians. Prolonged disorders of consciousness following sudden onset brain injury: national clinical guidelines. London: Royal College of Physicians; 2020. 85. Stender J, Gosseries O, Bruno MA, et al. Diagnostic precision of PET imaging and functional MRI in disorders of consciousness: a clinical validation study. Lancet. 2014;384:514–22. 86. Kotchoubey B, Yu T, Mueller F, Vogel D, Veser S, Lang S.  True or false? Activations of language-related areas in patients with disorders of consciousness. Curr Pharm Des. 2013;999:27–8. 87. Wang F, Di H, Hu X, Jing S, Thibaut A, Di Perri C, Huang W, Nie Y, Schnakers C, Laureys S. Cerebral response to subject’s own name showed high prognostic value in traumatic vegetative state. BMC Med. 2015;13:1–13. 88. Cavinato M, Freo U, Ori C, Zorzi M, Tonin P, Piccione F, Merico A. Post-acute P300 predicts recovery of consciousness from traumatic vegetative state. Brain Inj. 2009;23:973–80. https://doi.org/10.3109/02699050903373493. 89. Li R, Song WQ, Du JB, Huo S, Shan GX. Connecting the P300 to the diagnosis and prognosis of unconscious patients. Neural Regen Res. 2015;10:473–80. 90. Kotchoubey B, Pavlov YG. A systematic review and meta-analysis of the relationship between brain data and the outcome in disorders of consciousness. Front Neurol. 2018;9:356959.

3  Brain–Computer Interfaces and Their Place in the Management of Disorders…

57

91. Monti MM, Rosenberg M, Finoia P, Kamau E, Pickard JD, Owen AM. Thalamo-frontal connectivity mediates top-down cognitive functions in disorders of consciousness. Neurology. 2015;84:167–73. 92. Forgacs PB, Conte MM, Fridman EA, Voss PhD HU, Victor JD, Schiff ND. Preservation of electroencephalographic organization in patients with impaired consciousness and imaging-­ based evidence of command-following. Ann Neurol. 2014;76:869–79. 93. Edlow BL, Boerwinkle V, Annen J, et al. Common data elements for disorders of consciousness: recommendations from the working group on neuroimaging, vol. 39. Neurocrit Care; 2023. p. 753. 94. Sala A, Schindele A, Beliy N, et  al. An automated FDG-PET pipeline for the analysis of glucose brain metabolism in disorders of consciousness. J Cereb Blood Flow Metab. 2022;42:108–273. 95. Burwell S, Sample M, Racine E.  Ethical aspects of brain computer interfaces: a scoping review. BMC Med Ethics. 2017;18:1–11. 96. Luauté J, Morlet D, Mattout J. BCI in patients with disorders of consciousness: clinical perspectives. Ann Phys Rehabil Med. 2015;58:29–34. 97. Jox RJ, Bernat JL, Laureys S, Racine E. Disorders of consciousness: responding to requests for novel diagnostic and therapeutic interventions. Lancet Neurol. 2012;11:732–8. 98. Demertzi A, Sitt JD, Sarasso S, Pinxten W.  Measuring states of pathological (un)consciousness: research dimensions, clinical applications, and ethics. Neurosci Conscious. 2017;2017:1–13. 99. Noirhomme Q, Lesenfants D, Gomez F, Soddu A, Schrouff J, Garraux G, Luxen A, Phillips C, Laureys S.  Biased binomial assessment of cross-validated estimation of classification accuracies illustrated in diagnosis predictions. Neuroimage Clin. 2014;4:687–94. 100. Piarulli A, Bergamasco M, Thibaut A, Cologan V, Gosseries O, Laureys S.  EEG ultradian rhythmicity differences in disorders of consciousness during wakefulness. J Neurol. 2016;263:1746–60. 101. Martens G, Ibáñez-Soria D, Barra A, et al. A novel closed-loop EEG-tDCS approach to promote responsiveness of patients in minimally conscious state: a study protocol. Behav Brain Res. 2021;409:113311. 102. Peterson A, Aas S, Wasserman D. What justifies the allocation of health care resources to patients with disorders of consciousness? AJOB Neurosci. 2021;12:127–39. https://doi.org/1 0.1080/21507740.2021.1896594. 103. Ferré F, Heine L, Naboulsi E, et al. Self-processing in coma, unresponsive wakefulness syndrome and minimally conscious state. Front Hum Neurosci. 2023;17:1145253. 104. Thibaut A, Schiff N, Giacino J, Laureys S, Gosseries O. Therapeutic interventions in patients with prolonged disorders of consciousness. Lancet Neurol. 2019;18:600–14. 105. Thibaut A, Chatelle C, Vanhaudenhuyse A, Martens G, Cassol H, Martial C, Carrière M, Barra A, Laureys S, Gosseries O. Transcranial direct current stimulation unveils covert consciousness. Brain Stimul. 2018;11:642–4. 106. Edlow BL, Sanz LRD, Polizzotto L, et al. Therapies to restore consciousness in patients with severe brain injuries: a gap analysis and future directions. Neurocrit Care. 2021;35:68–85. 107. Bai Y, Xia X, Kang J, Yang Y, He J, Li X. TDCS modulates cortical excitability in patients with disorders of consciousness. Neuroimage Clin. 2017;15:702–9. 108. Castro M, Tillmann B, Luauté J, Corneyllie A, Dailler F, André-Obadia N, Perrin F. Boosting cognition with music in patients with disorders of consciousness. Neurorehabil Neural Repair. 2015;29:734–42.

Chapter 4

Prognosis in Disorders of Consciousness Anna Estraneo, Luigi Trojano, and Flora M. Hammond

Abstract  Clinical evolution and prognostic markers in disorders of consciousness (DoC) have not been fully established yet. Several scientific efforts have highlighted that clinical evolution is determined by several factors closely interacting with each other: patient age (likely influencing the physiological process of recovery, e.g., brain plasticity), etiology of brain damage (traumatic vs. non-traumatic), and diagnosis (vegetative state or minimally conscious state, likely related to the severity of brain damage). Time course of clinical evolution has not been fully defined, but to date several studies and international guidelines have suggested that some form of neural plasticity can develop even beyond the time limits originally identified by the Multi-Society Task Force on Vegetative State, and support late recovery of responsiveness and consciousness. Empirical evidence is accumulating about the possible prognostic value of further factors beyond patient age, etiology, and diagnosis. In particular, clinical evaluation by standardized tools and several neurophysiological indices, easy to collect at the bedside in the rehabilitative phase, are able to assess the functional integrity of neuronal populations, and to provide some cues for predicting outcomes. In selected patients advanced neurophysiological and neuroimaging methods can add useful additional diagnostic and prognostic information, particularly in patients in whom “covert” cognition is possibly present (even with the lack of behaviorally evident responses). Nonetheless, diagnosis and prognosis can be made difficult by the presence of clinical complications that also impact care strategies. To fully comprehend long-term evolution of DoC, prospective longitudinal systematic investigations of outcome in large groups of patients are needed. Identification of solid and reliable prognostic markers will help clinicians to update A. Estraneo (*) IRCCS Fondazione Don Carlo Gnocchi ONLUS, Florence, Italy e-mail: [email protected] L. Trojano Department of Psychology, University of Campania “Luigi Vanvitelli”, Caserta, Italy e-mail: [email protected] F. M. Hammond Indiana University School of Medicine, Indianapolis, IN, USA e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 C. Schnakers, S. Laureys (eds.), Coma and Disorders of Consciousness, https://doi.org/10.1007/978-3-031-50563-8_4

59

60

A. Estraneo et al.

current positions on medical, ethical, and legal issues connected with the management and care of patients with DoC.

Introduction After severe acquired brain injuries, survivors may remain in a clinical condition of prolonged disorder of consciousness (pDoC), which can persist chronically, even for a lifetime [1], or evolve toward recovery of consciousness. The most severe form of pDoC is vegetative state (VS) (also called unresponsive wakefulness syndrome— UWS) [2, 3], in which patients regain wakefulness, but show no evidence of conscious behaviors [2, 3]. An intermediate state within the continuum from VS/UWS to full awareness is defined as minimally conscious state (MCS) [4], characterized by minimal, inconsistent, but clearly discernible intentional responses to salient multisensorial stimuli. Based on the complexity of patients’ behaviors, MCS can be subcategorized into “MCS−” (with low-level intentional behavior, such as visual pursuit or localization of noxious stimulation) and “MCS+” (with high-level behavioral interactions, e.g., command following) [5]. Notwithstanding the lack of conclusive epidemiological data, it is commonly thought that incidence and prevalence of patients with prolonged and chronic DoC is progressively increasing, paradoxically, thanks to improvements in medical intervention techniques in the acute and post-acute stages [6]. Consequently, long-term management of these patients has growing clinical, economic, and ethical impact [7]. In this context, comprehending the evolution of these severe clinical conditions and identifying reliable prognostic markers would allow clinicians and patients’ families to make appropriate decisions concerning treatment and care [8]. This chapter is aimed at providing a brief overview on clinical evolution and prognostic markers of individuals in VS/UWS or MCS following traumatic or non-­ traumatic (including vascular and anoxic etiology) acquired brain injury. This effort will help clinicians to optimize use of the always scarce human and economic resources of national health systems.

Evolution Since the first definition of VS [9], several studies have addressed clinical evolution of pDoC. However, to date, the long-term clinical outcome of patients with pDoC is not well defined, because most patients quickly leave the medical system to be transferred home or to be admitted into chronic facilities and be cared for by their own families or non-specialized caregivers [10]. In several cases, natural clinical evolution is halted by early withdrawal of life-sustaining therapy (eWLST), even with the lack of solid, clear prognostic information. A recent longitudinal study on a large cohort of cardiac arrest survivors showed that in about 20% of patients who

4  Prognosis in Disorders of Consciousness

61

underwent eWLST, a good outcome could be predicted based on available prognostic information [11]. This report raised critical questions about the need for solid predictive factors, considering that eWLST, or the discontinuation of intensive long-­ term management, is often applied solely based on the patient’s family preference, or driven by an unsubstantiated belief of the treating clinical team that the prognosis is poor (resulting in the so-called fulfilling prophecy). Moreover, available outcome studies are not conclusive, because of heterogeneity in the study populations (patients with different etiologies and time from injury) and in duration of follow-up. Longitudinal studies (e.g., the Multi-Society Task Force on VS) [3] published before diagnostic criteria for MCS [4] did not consider MCS as a possible outcome, whereas even more recent longitudinal studies do not consider possible evolution to a condition of “covert cognition/awareness” [12]. Yet, this last condition could be present in up to 30% of patients clinically classified as in VS/UWS and can be identified only by means of advanced tools, such as functional brain-imaging techniques, detecting specific cortical responses to active paradigms, a finding highly suggestive of residual conscious processing [13]. The presence of covert cognition seems to have important prognostic implications, as it could herald further positive clinical evolution, but its identification poses a critical challenge, as advanced neuroimaging or neurophysiological tools cannot be applied in all patients due to technical problems (e.g., agitation, presence of mechanical ventilation) [14]. In addition, although these tools have been recommended by American and European Academies of Neurology guidelines on pDoC [1, 15], they are not widely available in routine care, due to logistical problems or lack of expertise. Notwithstanding these methodological limitations, however, the available longitudinal studies are quite consistent in pointing to patient’s age at onset, clinical diagnosis (VS/UWS vs. MCS), and etiology as the main determinants of clinical evolution, in terms of recovery of consciousness and survival. Accordingly, younger people and patients in MCS and with traumatic etiology are considered to have higher survival rates and more frequent recovery of full consciousness compared to elderly individuals in VS/UWS and with non-traumatic etiology [16–22]. As regards clinical diagnosis, patients in MCS are thought to have a higher probability of clinical improvement, a better outcome in terms of functional independence in basic daily activities [16–18, 21–24], and a longer life expectancy than VS/ UWS patients [20, 21]. This finding could be ascribed to lower severity of brain damage and to better response to neuromodulation treatment in MCS patients compared to VS/UWS patients [25–27]. In considering clinical diagnosis (VS/UWS vs. MCS or higher) as a prognostic factor, it is important to consider the high rate of inaccuracies in documented clinical diagnosis and the implications of such errors for prognostic accuracy. Rates of misdiagnosis (i.e., believing someone is in VS/ UWS when she/he is in MCS) of up to 40% have been reported [28]. This applies also to diagnosis of covert cognition/awareness [13]. Complicating the use of clinical diagnosis further, responsiveness may fluctuate, and its evaluation may be influenced by medical and physical factors. The consequences of misdiagnosis may be profound when leading to inappropriate and irreversible care decisions, such as eWLST [11] and denial of specialized rehabilitation and post-acute care.

62

A. Estraneo et al.

Most of the available studies on pDoC reported a better evolution in patients with traumatic etiology compared to those of non-traumatic origin (i.e., anoxic and vascular etiology) [16–22]. This finding might suggest that the severity of diffuse anoxic brain damage, and the detrimental conditions usually associated with vascular etiology (e.g., older age and premorbid coexisting diseases such as diabetes, hypertension, or dyslipidemia) could hamper survival and recovery of consciousness in patients with pDoC. Anoxic etiology is considered to have worse outcome, and this could lead to eWLST, particularly in patients with other predictors of poor outcome. Also, for this reason, few data are available about care pathways for survivors with post-anoxic pDoC. Recently, a meta-analysis on anoxic pDoC showed a long-term pooled rate of mortality of 26%, whereas the rate of recovery of full consciousness was low (17%) [29]. Although the time course of long-term clinical evolution has not been fully elucidated yet, some longitudinal studies showed that mortality rate is higher in the first year after brain injury than in the second year in both diagnostic groups (i.e., VS/ UWS and MCS) [20]. This finding could be ascribed, in part, to high clinical instability and high risk of developing severe medical complications in the first year post injury [30] and sometimes to eWLST [11]. The recovery of consciousness is most frequent in the first months after acute insult, likely because of enhanced adaptive and restorative neural plasticity in the initial period after brain injury [31]. According to a systematic review on functional neuroimaging and neurophysiological assessment, brain changes during recovery of consciousness seem to be particularly evident in the prefrontal cortex, basal forebrain, anterior cingulate cortex, and parietal cortex [32]. These brain areas are the main structures showing neural plasticity mechanisms and can be targets for neuromodulation [27]. The recovery of highly complex behaviors such as command following in the evolution from MCS− to MCS+ seems to be associated with an increase in brain metabolism in regions involved in language processing, such as the left fusiform gyrus, angular gyrus, and temporal cortex [33]. These findings would suggest that improvements in language processing are associated with restoration of brain connectivity in specific areas during MCS evolution. It is very important to recall that, probably due to the continuous improvements in therapeutic strategies [27], the originally proposed time limits for regaining consciousness in traumatic (12 months) and non-traumatic brain injury (3–6 months) [3] are no longer considered valid. Since the Multi-Society Task Force (MSTF) pivotal epidemiological study of VS/UWS [3], case reports [34–36] and large cohort studies [18, 21, 37–39] have documented late recovery of varying levels of consciousness beyond the above classical time limits, even in vascular and anoxic etiology. For instance, Estraneo et al. found that 20% of patients meeting MSTF criteria for “permanent” VS/UWS eventually improved to MCS [38]. Based on this kind of evidence, in 2018 the American Academy of Neurology (AAN) and American Congress of Rehabilitation Medicine (ACRM) suggested avoiding the term “permanent,” closely tied to the irreversibility of the clinical state [1]. In addition, the AAN and ACRM guidelines proposed using the term “prolonged” for DoC conditions from 28 days after onset and the term “chronic” when the elapsed time since onset is longer than the time limits proposed by the MSTF [1].

4  Prognosis in Disorders of Consciousness

63

It is important to underline that recovery of consciousness does not parallel the functional cognitive/motor improvement, as patients with different clinical diagnosis (i.e., patients in VS/UWS, MCS±, or even with full consciousness) can be affected by similar disability levels, especially if recovery of consciousness is late [40]. Patients who regain some level of consciousness at a late time (after MSTF proposed limits), however, often achieve poor functional improvements, and may remain in very severely disabled clinical conditions, with a considerable weight of secondary pathologies due to long-lasting immobility [40], such as abnormal posturing due to contractures or neurogenic heterotopic ossifications [41]. However, recent large traumatic brain injury cohort studies have demonstrated that clinical and neuropsychological conditions may progress for many years after traumatic brain injury. For example, among individuals admitted for traumatic brain injury rehabilitation who were not following commands, the majority achieved independence in functional activities with significant improvements continuing throughout at least the first decade after traumatic brain injury [42, 43]. Taken together, these findings would suggest that some kind of brain plasticity cannot be excluded even a long time after a severe brain injury. It is advisable to perform long-term monitoring of patients with pDoC, particularly of patients in MCS, to fully comprehend clinical evolution of these severe clinical conditions. For this purpose, cooperative systematic studies in large groups of prolonged DOC patients with a long-term follow-up seem to be strongly needed. Additionally, occurrence of severe motor and cognitive deficits in “late-­ recovered” patients highlights the need for an early appropriate level of rehabilitation, from the acute phase in the intensive care unit [23]. This could help prevent secondary medical complications, facilitate recovery of responsiveness, and minimize motor disabilities (e.g., contractures, joint limitations) due to long immobility, which negatively influence functional independence and quality of life of patients and of their families [44].

Prognostic Issues Starting in the acute phase in the intensive care unit, clinicians are routinely called to identify patients with high probability of improving, to plan appropriate care pathways in terms of intensity and duration of treatment. Clinicians involved in the follow-up phase must provide the patient’s family with reliable prognostic information to guide decision-making on long-term management. For this reason, in recent years a growing number of studies have focused on identification of prognostic markers for DoC (for details see Table 4.1). However, identifying valid prognostic markers for clinical evolution of patients with pDOC is still very challenging, and prognostic procedures can vary from country to country [45]. For instance, very informative neurophysiological data are used more frequently for prognostication in European countries than in the United States. Beyond non-traumatic etiology, older

64

A. Estraneo et al.

Table 4.1  Details of the cited studies on prognostic factors in disorders of consciousness First author [ref] TBI Kowalski [46]

Anoxic Estraneo [48]

Patients (number)

Time post-onset

Follow-up length (years or mos)

Not Rehabilitation Rehabilitation following admission discharge commands by rehabilitation admission

VS/UWS (43)

All etiologies Eilander VS/UWS or [24] MCS (145)

Marker

Outcome

Absence of intraventricular hemorrhage and intracranial mass effect Younger age, male sex, and absence of intraventricular hemorrhage, intracranial mass effect, and subcortical contusion

Recovery of consciousness

Functional Improvement (FIM)

1–6 mos

2-y post-onset CRS-R total Recovery of score ≥6, responsiveness presence of SEPs

6

>6

3 UWS (21)

>3

Yamamoto et al. 2010 [102]b

21 VS/ UWS, 5 MCS

Prospective (OL)/none

>3

Yamamoto et al. 2005 [101]b

21 VS/ UWS

Prospective (OL)/none

Yamamoto et al. 2003 [100]b

Cm-pf (unilat)

Ant int thalamus and paralaminar regions (bilat) Ant int thalamus and paralaminar regions (bilat)

Cuneiform (2 VS/ UWS) and Cm-pf (19 VS/UWS, 5 MCS) Cuneiform (2) and Cm-pf (19)

Cuneiform (2) and Cm-pf (19)

CRS-R; EEG

Daily; 40–110 Hz

30 min every 2–3 h; 25 Hz

CRS-R; EEG

Daily; 100 Hz

Of 21 treated VS/UWS pts, 8 emerged within 1 year but remained disabled; 16 out of 107 pts had pre-stimulation EEG desynchronization, Vth wave of the ABR, and high-amplitude pain-related P250 Improved consciousness in all (1/1) MCS pts; increased ACC and frontal activation EEG

Of 21 VS/UWS pts, 8 recovered consciousness within 1 year, survival rate higher for responders (6 years) than others (3 years), four pts (three responders) alive after 10 years; responders showed better pre-stimulation EEG desynchronization, Vth wave of the ABR, and pain-related P250 Of 21 VS/UWS pts, 8 recovered consciousness within 1 year, four-fifths of MCS pts were no longer bedridden

(continued)

Stable improvement in all (2/2) VS/UWS and all (1/1) MCS pts in 12+ months; increased EEG desynchronization, theta, and gamma power RDR, C/NC Two-tenths of MCS pts recovered consciousness and functional independence, one-tenth of MCS pts and one-fourth of VS/ UWS pts recovered consciousness

IS, GOS; EEG

IS; EEG, ABR, SEP

IS; EEG, ABR, SEP

30 min every 2–3 h daily; 25–50 Hz

30 min every 2–3 h daily; 25–50 Hz

30 min every 2–3 h daily; 25 Hz

8  Emerging Treatment for Patients with Disorders of Consciousness: The Field… 155

Case report (OL)/none

Prospective (OL)/none

Case report (OL)/none

Crossover RCT (DB)/ sham

Adams et al. 2017 [107]

Raguz et al. 2021 [108]c

Arnts et al. 2022 [109]

Dang et al. 2023 [110]

96

6–12

9 MCS

2–14

252

84

Cm-pf (bilat)

Central thalamus (unilat)

Ant int thalamus and paralaminar regions (bilat) Cm-pf (unilat)

EEG, LFP

Outcomes CRS-R; PET

Single-session crossover then continuous; 100 Hz

CRS-R; EEG

30 min every 2 h CRS-R; daily; 50–130 Hz MEG

30 min every 2 h RDR, C/ daily; 25 Hz NC; MRI

Daily for 3 years; CRS-R; 100 Hz EEG

Stimulation 3 months crossover then continuous; 30 Hz Intralaminar nuclei 70–250 Hz and TRN (bilat)

TSI (months) Target >12 Ant int thalamus and pallidum (bilat)

1 MCS

2 VS/ UWS, 3 MCS

1 MCS

1 MCS

Sample (treated) 1 VS/ UWS, 4 MCS

LFP local field potential, GP globus pallidus, TRN thalamic reticular nucleus a,b,c  Publications that share or might share patients

Case report (OL)/none

Wojtecki et al. 2014 [106]

Design/ Author(s) control Lemaire et al. Crossover 2018 [105] RCT (OL)/ sham

Table 8.1 (continued)

Modulation of beta and theta oscillations within central thalamus and increased thalamo-cortical coherence in theta when presented with emotionally valent stimuli No change in consciousness; improved spindle activity and recovery of REM sleep stages over 3 years Notable behavioral improvement in all (2/2) VS/UWS and all (3/3) MCS pts; increased volume in limbic, paralimbic, and subcortical areas 7 years later No lasting behavioral improvement; larger volume of activation during 50 Hz and increased functional connectivity and neural variability across frequency bands and areas Behavioral improvement in three of nine MCS pts; increased functional connectivity across the cortex, and especially between/within frontal, central, and parietal areas

Results Improved consciousness in all (1/1) VS/UWS pts and a quarter of MCS pts; increased medial cortex metabolism in responders

156 A. R. Hopkins et al.

8  Emerging Treatment for Patients with Disorders of Consciousness: The Field…

157

studies provided observational outcomes, and subsequent attempts were small-­ cohort, open-label trials using different behavioral scales to assess outcome, making them difficult to compare. Moreover, the diagnosis of VS/UWS or MCS is debatable in these early trials [95]. While designs, diagnoses, inclusion criterion, and outcome variables have advanced, the benefits of DBS for DOC remain difficult to fully assess, since no parallel sham-controlled trials have been conducted to date because of ethical and inclusion constraints. Even so, most DBS trials report behavioral and/or neurophysiological improvements for at least some DOC patients who receive DBS. Etiology does not seem to predict DBS success [95], despite reasons to expect that TBI patients with more focal/less widespread injuries might benefit most [92, 96]. However, MCS patients do generally show greater improvement than VS/UWS patients. Changes in brain activity relevant for consciousness observed during DBS intervention include increased CBF and metabolism throughout the cortex, increased front-parietal functional connectivity, as well as increased high-frequency activity and desynchronization observed in the EEG, further supporting the idea that this technique can ameliorate the specific biomarkers of impairment in DOC. Transcranial Focused Ultrasound Stimulation (tFUS) Low-intensity tFUS is a novel non-invasive technique that involves the delivery of sound waves that transiently excite or inhibit targeted tissue (depending on stimulation parameters and constituting cell types) with millimeter-level precision and without affecting intervening tissue (see Fig. 8.3) [111]. The focal point of the transducer (point of maximal intensity of the ultrasound) can target any area of the brain Fig. 8.3  tFUS schematic. A transducer (gray puck) is placed on the side of the subject’s head that delivers an ultrasound beam (transparent purple) with focal point (dark purple point) or point of maximal intensity

158

A. R. Hopkins et al.

without affecting intervening tissue, providing researchers and clinicians with the unprecedented ability to stimulate deep, subcortical structures non-invasively. While the underlying mechanisms of tFUS are not yet clear, one dominant perspective is that the sound emitted by tFUS causes tissue to compress and/or tense, which leads to changes in mechanosensitive membrane proteins and resting membrane potential [2, 111–113]. Nonetheless, the neurostimulatory effects of tFUS have been demonstrated and replicated with in-vitro tissue cultures, non-human animals, and more recently, humans [114, 115]. tFUS appears very safe, with no reported adverse events or side effects with human patients or healthy volunteers to date [116]. To date, four studies have reported using tFUS for DOC rehabilitation, including two “first-in-man” case studies demonstrating its application in this patient cohorts, and two small sample, perspective, open-label trials, one in acute and one in chronic patients (see Table 8.2). All trials to date have targeted the thalamus using a uniform set of stimulation parameters with rationale, at least in part, rooted in the Mesocircuit—that central stimulation of the thalamus and the thalamo-cortical system might promote the recovery of consciousness from DOC. Collectively, the results from these studies have yielded important information about how this technique could help DOC patients. First, the safety of this technique is confirmed, with no observed alteration of routine clinical measures during exposure (e.g., heartbeat, pulse oximetry) or adverse events tied to the stimulation. Second, most patients did show short-term improvements in responsiveness, as assessed with standard neurobehavioral tools, including acute and chronic patients. Finally, in both acute and chronic patients, concurrent neuroimaging shows that the change in thalamo-cortical connectivity observed during stimulation is linked to the degree of behavioral amelioration observed in the week following stimulation, thus providing a potential mechanistic link between the intervention and its behavioral effects [116, 117]. It should be stressed, however, that although exciting these initial feasibility studies did not include a sham control, suggesting that caution in interpreting their results is warranted. tFUS could thus be a promising rehabilitation technique for DOC. Lacking side effects and being non-invasive, it is an ideal candidate for robust, controlled clinical trials which can be deployed at the bedside. The flexibility the technique offers in targeting deep and/or cortical structures also opens the door for novel treatment strategies, such as concurrent multi-target treatment or longer interventions. However, given the uncertainty concerning the mechanisms of tFUS, studies should further clarify how the technique is affecting the behavior and brain activity of patients by mapping stimulation parameters to observed changes. These studies would be valuable and informative for planning future clinical trials.

Prospective (OL)/none

Prospective (OL)/none

Prospective (OL)/none

Cain et al. 2021 [116]a

Cain et al. 2022 [117]a

Cain et al. 2023 [119]a

14–66

TSI (months) 0.5

Two VS/UWS, eight MCSVS/ UWS

14–240

Two coma, three