Coma and Disorders of Consciousness

Caroline Schnakers • Steven Laureys

Editors

Coma and Disorders

of Consciousness
Editors
Caroline Schnakers, Ph.D. Steven Laureys, M.D., Ph.D.
Coma Science Group Coma Science Group
Cyclotron Research Center Cyclotron Research Center
University of Liège, Liège University of Liège, Liège
Belgium Belgium

ISBN 978-1-4471-2439-9 ISBN 978-1-4471-2440-5 (eBook)

DOI 10.1007/978-1-4471-2440-5
Springer Dordrecht Heidelberg New York London

Library of Congress Control Number: 2012940279

© Springer-Verlag London 2012

Coma and disorders of consciousness (ISBN 978-1-000) was previously published in French by Springer
as Coma et états de conscience altérée by Caroline Schnakers and Steven Laureys, in 2011.
Whilst we have made considerable efforts to contact all holders of copyright material contained in this
book, we may have failed to locate some of them. Should holders wish to contact the Publisher, we will
be happy to come to some arrangement with them.
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the
material is concerned, specifically the rights of translation, 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. Exempted from this legal reservation are brief excerpts in connection
with reviews or scholarly analysis or material supplied specifically for the purpose of being entered and
executed on a computer system, for exclusive use by the purchaser of the work. Duplication of this publi-
cation or parts thereof is permitted only under the provisions of the Copyright Law of the Publisher’s
location, in its current version, and permission for use must always be obtained from Springer. Permissions
for use may be obtained through RightsLink at the Copyright Clearance Center. Violations are liable to
prosecution under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publica-
tion 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.
While the advice and information in this book are believed to be true and accurate at the date of publica-
tion, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors
or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the
material contained herein.

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

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

Consciousness is synonymous with human existence. Rene Descarte’s bold procla-

mation, “Cogito ergo sum” (I think therefore I am), elegantly captures this premise.
The clear inference here is that self-doubt about one’s own existence establishes
proof of one’s existence. We exist because we know we exist. But how does one
come to know that someone else is aware of his existence? Without direct access to
knowledge of the self, it is impossible to prove (or disprove) the awareness of
another. In normal consciousness, this problem is obviated by the multitude of
behavioral expressions of conscious awareness manifested by living beings nearly
every waking moment. Words, gestures and actions, the “stuff” of consciousness,
provide compelling evidence of the inner life of another.
A small but significant percentage (5–10%) of those who sustain severe acquired
brain injury experience prolonged disturbance in consciousness. Most will eventu-
ally recover at least basic capacity for self and environmental awareness, but this
may not occur for many months and, in some cases, years. During this period of
severely altered consciousness, one’s existence may be stripped of the usual harbin-
gers of an active inner life. Sensory, motor, language, perceptual and drive systems
may all be compromised in the aftermath of severe brain injury. Consequently, the
repertoire of behaviors available to signal retention of conscious awareness may be
dramatically narrowed or lost altogether. This predicament presents one of human-
kind’s greatest existential dilemmas – is consciousness lost, or simply no longer
apparent? This question is at the heart of, Coma and Disorders of Consciousness
edited by Caroline Schnakers and Steven Laureys of the Coma Science Group based
at the University of Liège. The Coma Science Group is one of the most prolific and
accomplished research teams currently engaged in the study of patients with disor-
ders of consciousness (DoC). Laureys and Schnakers, specifically, are responsible
for many seminal papers in this rapidly advancing field. Together, they have assem-
bled an outstanding list of authors, most associated with the Coma Science Group,
and have compiled a comprehensive volume that aptly depicts the state of the sci-
ence in assessment and treatment of patients with DoC. The first half of the book is
devoted to technological advances in diagnostic and prognostic assessment, appro-
priately reflecting the most important developments over the last 5 years. Novel

vii
viii Foreword I

applications of functional neuroimaging (i.e., fMRI) and electrophysiologic

(i.e., EEG, ERP) methods of detecting conscious awareness in behaviourally non-
responsive patients are addressed, as are imaging techniques that index brain metab-
olites (i.e., spectroscopy), white matter tracts (i.e., diffusion tensor imaging) and
patterns of neural activation (i.e., transcranial magnetic stimulation), which may be
of prognostic significance. The second half of the book focuses largely on treatment
interventions ranging from sensory stimulation to neuromodulation strategies (i.e.,
pharmacotherapy, deep brain stimulation, brain computer interface). Two important
but understudied areas, sleep and oral feeding, are also covered. The book con-
cludes with a discussion of the ethical conundrums associated with clinical manage-
ment of patients with DoC and a forward look at this pivotal area of discovery,
located at the crossroads of philosophy and science. This book continues the search
for the “seat of consciousness” begun more than two millennia ago – a quest that,
by virtue of the human condition, we are compelled to undertake. For this reason,
no reader will turn away from this book disappointed.

Harvard Medical School, Boston, MA, USA Joseph T. Giacino

Foreword II

The definition of “the persistent vegetative state” by Jennett and Plum in 1972 con-
stituted an important step in the understanding and the monitoring of patients pre-
senting a disorder of consciousness after a period of coma. However, for years, the
scientific work on this topic remained insignificant. The merit of the Coma Science
Group is to have been interested in these patients with a new perspective, at a clini-
cal but also scientific and ethical level. These researchers are using tools of modern
neuroscience, particularly, functional neuroimaging and electrophysiology, in order
to better understand the brain functioning of patients recovering from coma. They
showed, for instance, that patients diagnosed as being in a “vegetative” state could
present conscious activity unsuspected at a clinical level. Their approach did not
remain only theoretical, as they were also interested in clinical assessment and,
particularly, in the risk of diagnostic errors, in the care and in the complex ethical
problems associated with these vulnerable patients. This team, in parallel with
British researchers, significantly contributed to the progress of knowledge in this
field. This book coordinated by Steven Laureys and Caroline Schnakers presents a
complete review of the main clinical and scientific advances on the disorders of
consciousness, including a synthesis of their own work. It also tackles questions
about the future, such as pharmacological treatment or deep brain stimulation.
Ethical questions are also discussed. I am sure that this book, coordinated by two of
the best experts in the field, will meet the success it deserves and will quickly
become a classic for all clinicians and researchers working with these patients.

Université de Versailles-Saint Quentin, Philippe Azouvi

Garches, France

ix
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 research
domain. 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 surround-
ings. In this context, a new field was called to emerge. 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 character-
ized by the development of a new kind of treatment, the sensory stimulation pro-
grams. In the late 1990s, the emergence of neuroimaging techniques opened new
opportunities to study brain reactivity in patients with disorders of consciousness.
However, in spite of the medical advances and the increasing number of severely
brain-injured patients, the assessment and treatment of patients recovering from
coma represents a very difficult and delicate task even today. The detection of signs
of consciousness is complicated by the frequent motor and cognitive limitations of
these patients. Treatment options are nearly absent, leaving the clinician often with
a situation of palliative rather than restorative care. Even in an experimental setting,
the study of patients in a coma or related disorders of consciousness is a real chal-
lenge. These patients are easily exhausted, limiting the assessment window, and
spontaneous motor reactions have to be controlled for. The development of a
research environment adapted to the scientific investigation of these patients is time
consuming and requests important clinical and scientific expertise. For over 10 years
now, the Coma Science Group (http://www.coma.ulg.ac.be) has been working on
the scientific exploration of disorders of consciousness, with both scientific and

xi
xii Preface

clinical research agendas. This research team, bridging various medical (neurology,
neurosurgery, intensive care, anesthesia, physical medicine, otorhinolaryngology)
and paramedical disciplines (psychology, speech therapy, physical therapy) as well
as engineering and biological disciplines, has been a major player in the develop-
ment of new assessment, communication, and treatment techniques for disorders of
consciousness, and this at both the behavioral and the neuroimaging level. We
decided to write this book in order to offer clinicians as well as researchers an over-
view of the most recent advancements in this domain.
By focusing on both clinicians and researchers as potential readers of this book,
we decided to include well-established findings about diagnostic/prognostic crite-
ria, ethical issues, assessment techniques (i.e., behavioral scales, electrophysiologi-
cal explorations, and structural/functional neuroimaging), and treatment procedures,
but also techniques under development (i.e., the use of classifiers, brain-computer
interfaces, transcranial magnetic stimulation, or deep brain stimulation) which, we
hope, will stimulate ideas for 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.

University of Liège, Liège, Belgium Caroline Schnakers

Steven Laureys
Acknowledgment

Our work is supported by the National Funds for Scientific Research (FRS), James
S. McDonnell Foundation, Mind Science Foundation Texas, US National Institutes
of Health, European Commission (Mindbridge, DISCOS, DECODER & COST),
French Speaking Community Concerted Research Action, Belgian Federal Public
Service of Health, Fonds Leon Fredericq, Public Utility Foundation “Université
Européenne du Travail,” “Fondazione Europea di Ricerca Biomedica,” and the
University and University Hospital of Liège.

xiii
Contents

1 Behavioral Assessment and Diagnosis of Disorders

of Consciousness ..................................................................................... 1
Caroline Schnakers and Steve Majerus
2 Prognosis of Patients with Altered State of Consciousness ................ 11
Marie-Aurélie Bruno, Didier Ledoux, Audrey Vanhaudenhuyse,
Olivia Gosseries, Aurore Thibaut, and Steven Laureys
3 Functional Imaging and Impaired Consciousness .............................. 25
Audrey Vanhaudenhuyse, Melanie Boly, Marie-Aurélie Bruno,
Olivia Gosseries, Athena Demertzi, Murielle Kirsch, Jean-Flory Luaba
Tshibanda, Marie Thonnard, Vanessa Charland-Verville,
Camille Chatelle, Aurore Thibaut, Steven Laureys, and Andrea Soddu
4 Multivariate Pattern Recognition Analysis: Brain Decoding ............ 35
Jessica Schrouff and Christophe L.M. Phillips
5 Spectroscopy and Diffusion Tensor Imaging in Disorders
of Consciousness ..................................................................................... 45
Jean-Flory Luaba Tshibanda, Athena Demertzi, and Andrea Soddu
6 Electrophysiology and Disorders of Consciousness ............................ 55
Quentin Noirhomme and Rémy Lehembre
7 Brain-Computer Interface: A Communication Aid? ......................... 67
Camille Chatelle, Zulay Lugo, Quentin Noirhomme, Bettina Sorger,
and Dorothee Lulé
8 Using Transcranial Magnetic Stimulation to Measure Cerebral
Connectivity in Patients with Disorders of Consciousness ................. 79
Melanie Boly, Olivia Gosseries, Marie-Aurélie Bruno,
Mario Rosanova, and Marcello Massimini
9 Sleep in Disorders of Consciousness..................................................... 85
Victor Cologan and Manuel Schabus

xv
xvi Contents

10 Sensory Stimulation Program ............................................................... 97

Haibo Di and Caroline Schnakers
11 Feasibility of Oral Feeding in Patients with Disorders
of Consciousness ..................................................................................... 105
Audrey Maudoux, Ingrid Breuskin, Olivia Gosseries,
Caroline Schnakers, and Audrey Vanhaudenhuyse
12 Pharmacological Treatments ................................................................ 121
Olivia Gosseries, Marie Thonnard, and Steven Laureys
13 Deep Brain Stimulation ......................................................................... 139
Marie Thonnard
14 The Ethics of Managing Disorders of Consciousness ......................... 147
Athena Demertzi, Olivia Gosseries, Marie-Aurélie Bruno, Caroline
Schnakers, Audrey Vanhaudenhuyse, Camille Chatelle,
Vanessa Charland-Verville, Aurore Thibaut, Marie Thonnard,
and Steven Laureys
15 Future Perspectives of Clinical Coma Science .................................... 155
Steven Laureys and Caroline Schnakers
Index ................................................................................................................ 163
Contributors

Philippe Azouvi, M.D., Ph.D. Department of Neurological

Rehabilitation, AP-HP, Hôpital Raymond-Poincaré;
Université de Versailles-Saint Quentin, Garches, France
Melanie Boly, M.D., Ph.D. Coma Science Group,
Cyclotron Research Center, University of Liège, Liège, Belgium
Ingrid Breuskin, M.D., Ph.D. Oto-Rhino-Laryngology Head and Neck
Surgery Department, University Hospital of Liège, Liège, Belgium
Marie-Aurélie Bruno, Ph.D. Coma Science Group,
Cyclotron Research Center, University of Liège, Liège, Belgium
Vanessa Charland-Verville, M.Sc. Coma Science Group,
Cyclotron Research Center, University of Liège, Liège, Belgium
Camille Chatelle, M.Sc. Coma Science Group, Cyclotron Research Center,
University of Liège, Liège, Belgium
Victor Cologan, M.Sc. Coma Science Group, Cyclotron Research Center,
University of Liège, Liège, Belgium
Athena Demertzi, M.Sc. Coma Science Group,
Cyclotron Research Center, University of Liège, Liège, Belgium
Haibo Di, Ph.D. Physiology Department, International Vegetative
State and Consciousness Science Institute, Hangzhou Normal
University, Hangzhou, China
Joseph T. Giacino, Ph.D. Department of Physical Medicine
and Rehabilitation Medicine, Spaulding Rehabilitation
Hospital/Harvard Medical School, Boston, MA, USA

xvii
xviii Contributors

Olivia Gosseries, M.Sc. Coma Science Group, Cyclotron Research Center,

University of Liège, Liège, Belgium
Murielle Kirsch, M.D. Coma Science Group, Cyclotron Research Center,
University of Liège, Liège, Belgium
Department of Anesthesiology, University Hospital of Liège,
Liège, Belgium
Steven Laureys, M.D., Ph.D. Coma Science Group, Cyclotron Research Center,
University of Liège, Liège, Belgium
Department of Neurology, University Hospital of Liège, Liège, Belgium
Didier Ledoux, M.D. Coma Science Group, Cyclotron Research Center,
University of Liège, Liège, Belgium
Department of Intensive Care, University Hospital of Liège, Liège, Belgium
Rémy Lehembre, M.Sc. Coma Science Group, Cyclotron Research Center,
University of Liège, Liège, Belgium
Zulay Lugo, M.Sc. Coma Science Group, Cyclotron Research Center,
University of Liège, Liège, Belgium
Department of Psychology, Biological Psychology,
Clinical Psychology and Psychotherapy, University of Würzburg,
Würzburg, Germany
Dorothee Lulé, Ph.D. Section of Neurophysiology,
Department of Neurology, University of Ulm, Ulm, Germany
Steve Majerus, Ph.D. Coma Science Group, Cyclotron Research Center,
University of Liège, Liège, Belgium
Marcello Massimini, Ph.D. Physiology Department,
University of Milan, Milan, Italy
Audrey Maudoux, M.D. Oto-Rhino-Laryngology Head and Neck
Surgery Department, University Hospital of Liège, Liège, Belgium
Coma Science Group, Cyclotron Research Center,
University of Liège, Liège, Belgium
Quentin Noirhomme, Ph.D. Coma Science Group,
Cyclotron Research Center, University of Liège, Liège, Belgium
Christophe L.M. Phillips, Ph.D. Cyclotron Research Center,
University of Liège, Liège, Belgium
Mario Rosanova, Ph.D. Physiology Department,
University of Milan, Milan, Italy
Contributors xix

Manuel Schabus, Ph.D. Department of Psychology, Laboratory

for Sleep and Consciousness Research, University of Salzburg,
Salzburg, Austria
Caroline Schnakers, Ph.D. Coma Science Group,
Cyclotron Research Center, University of Liège, Liège, Belgium
Jessica Schrouff, M.Sc. Cyclotron Research Center,
University of Liège, Liège, Belgium
Andrea Soddu, Ph.D. Coma Science Group,
Cyclotron Research Center, University of Liège, Liège, Belgium
Bettina Sorger, Ph.D. Coma Science Group, Cyclotron Research Center,
University of Liège, Liège, Belgium
Department of Cognitive Neuroscience, Faculty of Psychology and Neuroscience,
Maastricht University and Maastricht Brain Imaging Centre (M-BIC),
Maastricht, The Netherlands
Aurore Thibaut, M.Sc. Coma Science Group, Cyclotron Research Center,
University of Liège, Liège, Belgium
Marie Thonnard, M.Sc. Coma Science Group, Cyclotron Research Center,
University of Liège, Liège, Belgium
Jean-Flory Luaba Tshibanda, M.D. Coma Science Group,
Cyclotron Research Center, University of Liège, Liège, Belgium
Department of Radiology, University Hospital of Liège, Liège, Belgium
Audrey Vanhaudenhuyse, Ph.D. Coma Science Group,
Cyclotron Research Center, University of Liège, Liège, Belgium
Chapter 1
Behavioral Assessment and Diagnosis
of Disorders of Consciousness

Caroline Schnakers and Steve Majerus

Introduction

During these last years, there has been increasing fascination for the field of disor-
ders of consciousness. Due to progress in intensive care, more and more severely
brain-injured patients survive their initial brain insult, but many of these will go
through various states of impaired consciousness. The proportion of these patients
is estimated at 46 per million in the United States, 14 per million in Great Britain
[1], and 36 per million in Belgium [2]. Many patients will remain in a vegetative
state for a more or less extended period of time before regaining some level of con-
sciousness (minimally conscious state). Some of these cases have been widely
broadcasted, such as the case of Terri Shiavo (1963–2005) who stayed in a vegeta-
tive state for 15 years after a cardiac arrest or the case of Terry Wallis who emerged
from a minimally conscious state 19 years (1984–2003) after a severe traumatic
brain injury [3]. At the same time, prolonged hospitalization is expensive. In the
United States, the costs are estimated between 600,000 and 1,875,000 dollars per
year per patient with severe traumatic injury [4]. Questions regarding end-of-life
decisions are critical here, particularly in chronic vegetative patients. In a recent
European survey (n = 2,475), the majority of medical and paramedical professionals

C. Schnakers, Ph.D. (*)

Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du 6 Août, Sart Tilman B30, 4000 Liège, Belgium
e-mail: [email protected]
S. Majerus, Ph.D.
Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du 6 Août, Sart Tilman B30, 4000 Liège, Belgium
Department of Psychology, Cognition and Behavior, University of Liège,
Boulevard du Rectorat 3 B33, 4000 Liège, Belgium
e-mail: [email protected]

C. Schnakers, S. Laureys (eds.), Coma and Disorders of Consciousness, 1

DOI 10.1007/978-1-4471-2440-5_1, © Springer-Verlag London 2012
2 C. Schnakers and S. Majerus

(66%) agreed to withdraw life support for chronic vegetative patients while only
28% agreed for patients in a chronic minimally conscious state. Many clinicians
reported that they would not want to be life-supported if they happened themselves
to be in a chronic vegetative state (82%) or if they were in a chronic minimally con-
scious state (67%) [5]. As we can see, the social, economic, and ethical conse-
quences associated with disorders of consciousness, and particularly the vegetative
state, are gigantic.
The term “vegetative” indicates preserved physiological functions (cardiac,
respiratory, sleep/wake cycles) without clear signs of consciousness of either the
self or the environment. In a sense, the body works without the mind. One of the
few ways we have to differentiate these patients from conscious patients is to
observe their spontaneous behaviors and their reactions to stimuli occurring in their
environment. This behavioral assessment requests thorough expertise on behalf of
the clinician. It also depends on the physical and mental capacities (particularly,
the vigilance level) of the patient at the time of assessment. Missing signs of con-
sciousness is not a rare fact, and diagnostic errors are frequent (i.e., around 40%)
[6–8]. The diagnosis is, however, crucial. It influences the way the patient’s care
will be oriented and the way end-of-life decisions will be considered with the
patient’s family. Developing valid and sensitive behavioral scales to detect the
presence of signs of consciousness, even subtle, therefore, represents a real
challenge.

Disorders of Consciousness: Definition

We present here the different states of altered consciousness (i.e., brain death, coma,
vegetative state and minimally conscious state) that can occur in severely brain-
injured patients (Table 1.1).

Brain Death

In brain death, the organism does not function anymore without medical assistance
[9, 10]. Critical functions such as respiration, blood circulation, neuroendocrine,
and homeostatic regulation are absent. The patient is apneic and unreactive to envi-
ronmental stimulation. The term “brain death” requires the demonstration of irre-
versible cessation of all clinical functions of the brain, but also, of the brainstem
[11]. Brain death is classically caused by a brain lesion (e.g., massive traumatic
injury, intracranial hemorrhage, or anoxia) that results in increased intracranial
pressure (i.e., higher than the mean arterial blood pressure). The diagnosis can be
made within 6–24 h postinjury [12].
1 Behavioral Assessment and Diagnosis of Disorders of Consciousness 3

Table 1.1 Diagnostic criteria for brain death, coma, vegetative, and minimally conscious states
and locked-in syndrome
Consciousness level Diagnostic criteria
Brain death No arousal/eye-opening
No behavioral signs of awareness
Apnea
Loss of brain functions (brainstem reflexes)
Coma No arousal/eye-opening
No behavioral signs of awareness
Impaired spontaneous breathing
Impaired brainstem reflexes
No vocalizations
>1 h
Vegetative state Arousal/spontaneous or stimulus-induced eye opening
No behavioral signs of awareness
Preserved spontaneous breathing
Preserved brainstem reflexes
No purposeful behaviors
No language production or comprehension
Preservation (partial or complete) of hypothalamic and brain stem
autonomic functions
Preservation of sleep-wake cycle
>1 month: persistent vegetative
Compatible:
Grimaces to pain
Localization to sounds
Atypical but compatible:
Visual fixation
Response to threat
Inappropriate single words
Minimally Arousal/spontaneous eye-opening
conscious state Fluctuating but reproducible behavioral signs of awareness
Response to verbal order
Environmentally-contingent smiling or crying
Object localization and manipulation
Sustained visual fixation and pursuit
Verbalizations
Intentional but unreliable communication
Emergence from MCS:
Functional communication
Functional object use
Locked-in syndrome Arousal/spontaneous eye-opening
Preserved cognitive functions
Communication via eye gaze
Anarthria
Tetraplegia
4 C. Schnakers and S. Majerus

Coma

Brain injured patients can stay in coma for several weeks, showing no arousal or
awareness; the eyes are constantly closed and no voluntary behavioral responses are
observed [13]. In most cases, patients emerge from their comatose state within
2–4 weeks postinjury. The prognosis for recovery is influenced by different factors
such as etiology, the patient’s general medical condition, and age. Outcome is
known to be unfavorable if, during 3 consecutive days, there are no pupillary or
corneal reflexes, stereotyped or absent motor responses to noxious stimulation, and
isoelectrical or burst suppression EEG patterns [14]. Prognosis for recovery in trau-
matic coma survivors is better than in anoxic cases. Recovery from coma may lead
to a vegetative state, a minimally conscious state or, more rarely, to a locked-in
syndrome.

Vegetative State

The term “vegetative state” (VS) was defined to describe “an organic body capable
of growth and development but devoid of sensation and thought” [1]. This state
implies the preservation of autonomic functions (e.g., cardiovascular regulation,
thermoregulation) and sleep–wake cycles in the absence of awareness. Behaviorally,
patients in VS open their eyes spontaneously or in response to stimulation, but they
only show reflex behaviors, unrelated to the environment [15]. It is very important
to stress the difference between persistent and permanent VS, which are – both
being referred to as PVS – causing unnecessary confusion. When the term “persis-
tent vegetative state” was first described, it was emphasized that persistent did not
mean permanent; it is now recommended to omit “persistent” and to mention the
time during which a patient has been in a VS. In the absence of recovery after a
specified period (i.e., 3 months for traumatic etiology and 12 months for nontrau-
matic etiologies) the state can be declared to be permanent and, only then, the ethi-
cal and legal issues concerning treatment withdrawal can be discussed [16]. Given
the negative connotation of the term “vegetative state”, it has been recently pro-
posed to use the term “unresponsive wakefulness syndrome” rather than “vegetative
state” [17].

Minimally Conscious State

Patients in a minimally conscious state (MCS) are awake and show fluctuating but
reproducible signs of awareness. These patients can manifest emotional and ori-
ented behavioral responses such as response to verbal order, object manipulation,
oriented responses to noxious stimulation, visual pursuit, or fixation. However,
1 Behavioral Assessment and Diagnosis of Disorders of Consciousness 5

these behaviors can fluctuate in time, which makes the detection of awareness a
difficult endeavor. Recovery from MCS is defined by the emergence of functional
communication and/or functional object use. Even if prognosis is better as com-
pared to VS, some patients can remain in an MCS without fully recovering con-
sciousness for a prolonged period of time [18].

Locked-in Syndrome

The Locked-in Syndrome (LIS) is not a disorder of consciousness, although it may

present as one. Misdiagnosis is very frequent here [19]. LIS patients cannot move or
talk due to quadriplegia; only vertical eye movements and eye blinking may be used
to communicate with their environment. Typically, the diagnosis of LIS is not estab-
lished until 2.5 months post-onset. There are reports suggesting that family mem-
bers tend to detect signs of consciousness prior to medical staff [20]. This syndrome
is often due to a selective supranuclear motor de-efferentation, producing a paraly-
sis of all four limbs and the last cranial nerves without interfering with conscious-
ness or cognition. According to Bauer et al. [21], different categories of LIS can be
distinguished based on the extent of motor impairment: (1) Classical LIS consists
of a total immobility but preserved vertical eye movements and blinking; (2) Incomplete
LIS is characterized by remnant nonocular voluntary motions (e.g., head or fingers
movements); (3) Total LIS patients are completely immobile, unable to control any
eye movement, and nearly impossible to diagnose before the development of neu-
roimaging [21]. Patients with LIS may live for 12 or more years while remaining in
this state of complete immobility [22]. It is important to note that LIS patients rate
their quality of life to be similar to the quality of life in the healthy population [23].
In the absence of other structural or functional brain abnormalities, patients with
LIS are fully able to make their own decisions [24].

Behavioral Scales

Behavioral observation constitutes the major tool for detecting signs of conscious-
ness in severely brain injured patients. The detection of voluntary behaviors is nev-
ertheless difficult and signs of consciousness can easily be missed due to sensory
and motor disabilities, tracheostomy, fluctuating arousal levels, or ambiguous and
rapidly exhausted responses. Numerous behavioral rating scales have been devel-
oped and validated to assess levels of consciousness and to help establishing accu-
rate diagnoses. We have to distinguish here “arousal” from “consciousness.” Indeed,
a patient can be aroused but show no signs of consciousness, as in VS. Preservation
of arousal is, therefore, a necessary but insufficient condition for consciousness
(Fig. 1.1) [25]. In this section, we briefly review commonly used behavioral instru-
ments for the assessment of consciousness.
6 C. Schnakers and S. Majerus

Brain death, Vegetative Minimally Locked-in

coma state conscious syndrome
state

Awareness

Awareness

Awareness

Awareness
Arousal

Arousal

Arousal

Arousal
Fig. 1.1 Behavioral observation assesses two dimensions of consciousness: arousal and aware-
ness. In brain death and coma, both dimensions are absent. In the vegetative state, arousal level is
relatively preserved in the absence of signs of awareness. In the minimally conscious state, both
dimensions are present although behavioral signs often fluctuate. In the locked-in syndrome, both
dimensions are fully preserved despite complete loss of speech and motor functions

The Glasgow Coma Scale (GCS)

The GCS [26, 27] remains the most widely used tool in traumatic and acute-care
settings. The GCS was the first validated rating scale developed to monitor levels
of consciousness in the intensive care unit. This scale is relatively brief and can
be easily incorporated into routine clinical care. It includes three subscales that
address arousal level, motor function, and verbal abilities. Subscale scores are
added and yield a total score ranging from 3 to 15. The GCS has been exten-
sively investigated for its prognostic value [28]. Despite its widespread use, the
GCS has been criticized for variable inter-rater agreement and problems in deriv-
ing scores in patients with ocular trauma, tracheostomy, or ventilatory support
[29–34].

The Full Outline of UnResponsiveness Scale (FOUR)

The FOUR was recently developed to replace the GCS to assess severely brain-
injured patients in intensive care [35–37]. The scale includes four subscales
assessing motor and ocular responses, brainstem reflexes, and breathing. The total
score ranges from 0 to 16. Unlike the GCS, the FOUR does not assess verbal
functions to accommodate the high number of intubated patients in intensive care.
1 Behavioral Assessment and Diagnosis of Disorders of Consciousness 7

A score of 0 on the FOUR assumes the absence of brainstem reflexes and breath-
ing and, therefore, helps to diagnose brain death. The scale also monitors recov-
ery of autonomic functions and tracks emergence from VS. The FOUR is
specifically designed to detect patients in a locked-in syndrome as it uses oculo-
motor commands that detect vertical eye movements and eye blinks, both being
preserved in LIS.

The Wessex Head Injury Matrix (WHIM)

The WHIM [38, 39] was developed to capture changes in patients in VS until
emergence from posttraumatic amnesia. This tool is particularly sensitive to
detecting changes in patients in MCS not captured by traditional scales such as
the GCS. Shiel et al. longitudinally followed 97 severely brain injured patients
recovering from coma to create the WHIM. The 62-items were ordered accord-
ing to the mean sequence of recovery observed in these patients; they assess
arousal level and concentration, visual consciousness (i.e., visual pursuit), com-
munication, cognition (i.e., memory and spatiotemporal orientation), and social
behaviors. The WHIM score represents the rank of the most complex behavior
observed [40].

The JFK Coma Recovery Scale (CRS-R)

The CRS-R was originally developed by investigators from the JFK Johnson
Rehabilitation Institute in 1991. The scale was revised and published in 2004 as
the JFK Coma Recovery Scale-Revised (CRS-R) [41–46]. The purpose of the
CRS-R is to assist with differential diagnosis, prognostic assessment, and treat-
ment planning in patients with disorders of consciousness. The scale consists of
23 items that comprise six subscales addressing auditory, visual, motor, oromo-
tor, communication, and arousal functions. CRS-R subscales are comprised of
hierarchically-arranged items associated with brain stem, subcortical, and cor-
tical processes. The lowest item on each subscale represents reflexive activity
while the highest items represent cognitively mediated behaviors. Scoring is
standardized and based on the presence or absence of operationally defined
behavioral responses to specific sensory stimuli. Psychometric studies indicate
that the CRS-R meets minimal standards for measurement and evaluation tools
designed for use in interdisciplinary medical rehabilitation. The CRS-R can be
administered reliably by trained examiners and produces reasonably stable
scores over repeated assessments. Validity analyses have shown that the CRS-R
is capable of discriminating patients in MCS from those in VS, which is of
critical importance in establishing prognosis and formulating treatment
interventions [42, 44].
8 C. Schnakers and S. Majerus

Pain Assessment: The Nociception Coma Scale

Patients recovering from coma are unable to communicate their feelings and their
perception of pain. It is, hence, important to develop sensitive instruments to assess
pain perception in these patients. Even if pain scales were developed to assess pain
in different types of noncommunicative populations (such as demented elderly and
newborns), none of these are adapted to detect pain in severely brain-injured
patients. In this context, the Nociception Coma Scale (NCS) was recently devel-
oped [47, 48]. This scale includes the observation of motor, verbal, and visual
responses as well as facial expression following pain. The total score varies from 0
to 12. A validation study of the NCS was performed by observing the responses of
48 VS and MCS patients following a nociceptive stimulation (pressure applied to
the bed of the fingernail). Results showed a good inter-rater agreement and concur-
rent validity between NCS and other pain scales validated for demented elderly
patients and newborns. However, in contrast to these scales, the total score of the
NCS significantly differed according to the level of consciousness, with higher
scores in MCS versus VS patients, suggesting that the scale is particularly adapted
to assess pain in patients recovering from coma. These results show that the NCS is
a sensitive scale, which specifically assesses pain in noncommunicative severely
brain-injured patients.

Conclusion

Accurately diagnosing patients with a disorder of consciousness represents a real

challenge. A recent study showed that about 40% of patients clinically diagnosed as
being in a vegetative state present, in fact, signs of consciousness [8]. The recovery
of consciousness is usually gradual, sometimes marked by clear signs of conscious-
ness, but more often these signs are subtle. The assessment of residual cognitive
functioning is complicated by vigilance fluctuations, motor deficits, and other con-
founding factors. The knowledge of diagnostic criteria and the use of valid and sen-
sitive standardized scales are, therefore, essential to establish a precise diagnosis.

References

1. Jennett B. 30 years of the vegetative state: clinical, ethical and legal problems. In: Laureys S,
editor. The boundaries of consciousness: neurobiology and neuropathology, vol. 150.
Amsterdam: Elsevier; 2005. p. 541–8.
2. Demotte R. Politique de la santé à mener à l’égard des patients en état végétatif persistant ou
en état pauci-relationnel. Moniteur Belge. 2004;69334–40.
3. 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.
1 Behavioral Assessment and Diagnosis of Disorders of Consciousness 9

4. Consensus conference. Rehabilitation of persons with traumatic brain injury. NIH consensus
development panel on rehabilitation of persons with traumatic brain injury. JAMA.
1999;282(10):974–83.
5. Demertzi A, Ledoux D, Bruno MA, et al. Attitudes towards end-of-life issues in disorders of
consciousness: a European survey. J Neurol. 2011;258:1058–65.
6. Childs NL, Mercer WN, Childs HW. Accuracy of diagnosis of persistent vegetative state.
Neurology. 1993;43(8):1465–7.
7. Andrews K, Murphy L, Munday R, Littlewood C. Misdiagnosis of the vegetative state: retro-
spective study in a rehabilitation unit. BMJ. 1996;313(7048):13–6.
8. Schnakers C, Vanhaudenhuyse A, Giacino J, et al. Diagnostic accuracy of the vegetative and
minimally conscious state: clinical consensus versus standardized neurobehavioral assess-
ment. BMC Neurol. 2009;9:35.
9. Bernat JL. A defense of the whole-brain concept of death. Hastings Cent Rep. 1998;28(2):
14–23.
10. Haupt WF, Rudolf J. European brain death codes: a comparison of national guidelines.
J Neurol. 1999;246(6):432–7.
11. Pallis C, Harley DH. ABC of brainstem death. 2nd ed. London: BMJ Publishing Group;
1996.
12. Medical Consultants on the Diagnosis of Death. Guidelines for the determination of death.
Report of the medical consultants on the diagnosis of death to the President’s Commission for
the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research. JAMA.
1981;246(19):2184–6.
13. Plum F, Posner JB. The diagnosis of stupor and coma. 1st ed. Philadelphia: Davis, F.A; 1966.
14. Attia J, Cook DJ. Prognosis in anoxic and traumatic coma. Crit Care Clin.
1998;14(3):497–511.
15. The Multi-Society Task Force on PVS. Medical aspects of the persistent vegetative state (1).
N Engl J Med. 1994;330(21):1499–508.
16. Monti M, Laureys S, Owen A. The vegetative state. BMJ. 2010;341:3765.
17. Laureys S, Celesia G, Cohadon F, et al. Unresponsive wakefulness syndrome: a new name for
the vegetative state or apallic syndrome. BMC Med. 2010;8(1):68.
18. Giacino J, Ashwal S, Childs N, et al. The minimally conscious state: definition and diagnostic
criteria. Neurology. 2002;58(3):349–53.
19. American Congress of Rehabilitation Medicine. Recommendations for use of uniform nomen-
clature pertinent to patients with severe alterations of consciousness. Arch Phys Med Rehabil.
1995;76:205–9.
20. Laureys S, Pellas F, Van Eeckhout P, et al. The locked-in syndrome: what is it like to be con-
scious but paralyzed and voiceless? Prog Brain Res. 2005;150:495–511.
21. Bauer G, Gerstenbrand F, Rumpl E. Varieties of the locked-in syndrome. J Neurol.
1979;221(2):77–91.
22. Katz RT, Haig AJ, Clark BB, DiPaola RJ. Long-term survival, prognosis, and life-care plan-
ning for 29 patients with chronic locked-in syndrome. Arch Phys Med Rehabil.
1992;73(5):403–8.
23. Lule D, Zickler C, Hacker S, et al. Life can be worth living in locked-in syndrome. Prog Brain
Res. 2009;177:339–51.
24. Bruno M, Bernheim JL, Schnakers C, Laureys S. Locked-in: don’t judge a book by its cover.
J Neurol Neurosurg Psychiatry. 2008;79(1):2.
25. Majerus S, Gill-Thwaites H, Andrews K, Laureys S. Behavioral evaluation of consciousness in
severe brain damage. In: Laureys S, editor. The boundaries of consciousness: neurobiology
and neuropathology. Amsterdam: Elsevier; 2005. p. 397–413.
26. Teasdale G, Jennett B. Assessment of coma and impaired consciousness. A practical scale.
Lancet. 1974;2(7872):81–4.
27. Teasdale G. Acute impairment of brain function-1. Assessing ‘conscious level’. Nurs Times.
1975;71(24):914–7.
10 C. Schnakers and S. Majerus

28. McNett M. A review of the predictive ability of Glasgow Coma Scale scores in head-injured
patients. J Neurosci Nurs. 2007;39(2):68–75.
29. Jagger J, Jane JA, Rimel R. The Glasgow Coma Scale: to sum or not to sum? Lancet.
1983;2(8341):97.
30. Koziol JA, Hacke W. Multivariate data reduction by principal components, with application to
neurological scoring instruments. J Neurol. 1990;237(8):461–4.
31. Rowley G, Fielding K. Reliability and accuracy of the Glasgow Coma Scale with experienced
and inexperienced users. Lancet. 1991;337(8740):535–8.
32. Moskopp D, Stähle C, Wassermann HD. Problems of the Glasgow Coma Scale with early
intubated patients. Neurosurg Rev. 1995;18:253–7.
33. Wijdicks EF, Kokmen E, O’Brien PC. Measurement of impaired consciousness in the neuro-
logical intensive care unit: a new test. J Neurol Neurosurg Psychiatry. 1998;64:117–9.
34. Bhatty GB, Kapoor N. The Glasgow Coma Scale: a mathematical critique. Acta Neurochir.
1993;120(3–4):132–5.
35. Wijdicks EF. Clinical scales for comatose patients: the Glasgow Coma Scale in historical con-
text and the new FOUR score. Rev Neurol Dis. 2006;3(3):109–17.
36. Wijdicks EF, Bamlet WR, Maramattom BV, et al. Validation of a new coma scale: the FOUR
score. Ann Neurol. 2005;58(4):585–93.
37. 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.
38. Shiel A, Horn SA, Wilson BA, et al. The wessex head injury matrix (WHIM) main scale: a
preliminary report on a scale to assess and monitor patient recovery after severe head injury.
Clin Rehabil. 2000;14(4):408–16.
39. Majerus S, Van der Linden M. Wessex head injury matrix and Glasgow/Glasgow-liège Coma
Scale: a validation and comparison study. Neuropsychol Rehabil. 2000;10(2):167–84.
40. Wilson FC, Elder V, McCrudden E, Caldwell S. Analysis of wessex head injury matrix
(WHIM) scores in consecutive vegetative and minimally conscious state patients. Neuropsychol
Rehabil. 2009;19(5):754–60.
41. Giacino J, Kalmar K, Whyte J. The JFK coma recovery scale-revised: measurement character-
istics and diagnostic utility. Arch Phys Med Rehabil. 2004;85(12):2020–9.
42. Schnakers C, Majerus S, Giacino J, et al. A French validation study of the coma recovery
scale-revised (CRS-R). Brain Inj. 2008;22(10):786–92.
43. Lovstad M, Froslie KF, Giacino JT, et al. Reliability and diagnostic characteristics of the JFK
coma recovery scale-revised: exploring the influence of rater’s level of experience. J Head
Trauma Rehabil. 2010;25(5):349–56.
44. Schnakers C, Giacino J, Kalmar K, et al. Does the FOUR score correctly diagnose the vegeta-
tive and minimally conscious states? Ann Neurol. 2006;60(6):744–5.
45. 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.
46. Giacino J, Kezmarsky MA, DeLuca J, Cicerone KD. Monitoring rate of recovery to predict
outcome in minimally responsive patients. Arch Phys Med Rehabil. 1991;72(11):897–901.
47. 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.
48. Schnakers C, Chatelle C, Majerus S, et al. Assessment and detection of pain in non-communi-
cative severely brain-injured patients. Expert Rev Neurother. 2010;10(11):1725–31.
Chapter 2
Prognosis of Patients with Altered State
of Consciousness

Marie-Aurélie Bruno, Didier Ledoux, Audrey Vanhaudenhuyse,

Olivia Gosseries, Aurore Thibaut, and Steven Laureys

Introduction

Medical technologies and treatment developments led to an increase in patients

surviving severe brain damage. A patient’s prognosis is important not only for the
patient, but also for his family, as well as for the caregivers. The family has to deal
with the patient’s cognitive and motor deficits caused by the accident, the lack of
communication, and the emotions associated with the events. Healthcare profes-

M.-A. Bruno, Ph.D. (*) • A. Vanhaudenhuyse, Ph.D. •

A. Thibaut, M.Sc. • S. Laureys, M.D., Ph.D.
Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du 6 Août, Sart Tilman B30, 4000 Liège, Belgium
Department of Neurology, University Hospital of Liège,
Liège, Belgium
e-mail: [email protected]; [email protected]; [email protected];
[email protected]
D. Ledoux, M.D.
Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du 6 Août, Sart Tilman B30, 4000 Liège, Belgium
Department of Intensive Care, University Hospital of Liège,
Liège, Belgium
Department of Neurology, University Hospital of Liège,
Liège, Belgium
e-mail: [email protected]
O. Gosseries, M.Sc.
Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du 6 Août, Sart Tilman B30, 4000 Liège, Belgium
Department of Neurology, University Hospital of Liège,
Liège, Belgium
National Fund for Scientific Research, Liège, Belgium
e-mail: [email protected]

C. Schnakers, S. Laureys (eds.), Coma and Disorders of Consciousness, 11

DOI 10.1007/978-1-4471-2440-5_2, © Springer-Verlag London 2012
12 M.-A. Bruno et al.

Emergence from the minimally conscious state

Functional
recovery Functional communication
Functional object use

Minimally conscious state

Appropriate smiling/crying
Recovery of Localization to noxious stimulation
consciousness Visual pursuit
Command following
vegetative state/unresponsive
wakefulness syndrome

Arousal Eye opening

Reflex movements

Coma

Fig. 2.1 Behaviorally progressive transitions from coma, to vegetative state/unresponsive wake-
fulness syndrome, to minimally conscious state, to emergence from the minimally conscious state

sionals are involved in the daily care, in assessing the effectiveness of various medi-
cal treatments, and in any end-of-life decisions. A patient’s recovery can be
considered according to three dimensions: survival/mortality, recovery of con-
sciousness (minimally conscious state – MCS), and functional recovery
(the emergence from MCS). Several factors are associated with an increased mortal-
ity in patients with altered state of consciousness: (1) a Glasgow motor scale £2 [1],
(2) the absence of pupillary and/or corneal reflex, (3) a status epilepticus, (4) an iso-
electric EEG (“flat EEG”), (5) the absence of Somatosensory Evoked Potentials (N20),
and (6) an increased serum concentration of neuron-specific enolase (NSE) [2–4].
Medical complications such as hypotension, hyperthermia, hyperglycemia, infec-
tions, or prolonged mechanical ventilation also decrease the chances of survival [4–6].
Most common causes of death in the intensive care and rehabilitation units are urinary
and pulmonary infections, heart failure and cachexia, sudden death, and organ failure
[7, 8]. After the acute phase and once medical complications are stabilized, the poten-
tial degree of recovery should be evaluated. Patients may show signs of consciousness
recovery or functional recovery. Recovery of consciousness can be translated by the
presence of clear-cut signs of self- or environmental awareness (fluctuating voluntary
responses to a verbal and/or a written request, visual tracking, context-specific emo-
tional responses, etc. – MCS) [9]. Functional recovery is characterized by the recu-
peration of a functional communication and/or a functional use of objects and/or an
ability to learn and perform new tasks as well as to participate in personal, profes-
sional, or recreational activities (emergence from MCS) [10]. Recovery of conscious-
ness can occur without functional recovery. Those vegetative state patients/
unresponsive wakefulness syndrome (VS/UWS) who will recover usually evolve into
an MCS (consciousness recovery), then possibly present a functional recovery,
2 Prognosis of Patients with Altered State of Consciousness 13

Table 2.1 Clinical evolution at 3, 6, and 12 months post-onset for vegetative state/unresponsive
wakefulness syndrome (VS/UWS) patients and for minimally conscious state (MCS) patients (in %)
Diagnosis 1 month
post-injury 3 months (%) 6 months (%) 12 months (%)
VS-traumatic (n = 52)
Dead 21 31 42
VS 56 33 21
MCS 21 23 14
Emergence from MCS 2 13 23
VS-non-traumatic (n = 64)
Dead 30 52 70
VS 64 38 20
MCS 6 9 8
Emergence from MCS 0 1 2
MCS-traumatic (n = 35)
Dead 6 14 23
VS 17 6 0
MCS 60 40 29
Emergence from MCS 17 40 48
MCS-non-traumatic (n = 49)
Dead 14 18 33
VS 20 4 2
MCS 64 61 39
Emergence from MCS 2 17 26

reflecting the emergence from MCS and this with or without physical, psychological,
or neuropsychological disorders [9] (Fig. 2.1).
Several studies attempted to quantify the mortality rate and the degree of recov-
ery of patients with disorders of consciousness (Tables 2.2, 2.3, and 2.4). Although
these studies are informative, they have limitations, and for several reasons they
cannot guide healthcare professionals and families in making treatment decisions:
(1) most of these studies only included VS/UWS patients and very few studied
MCS patients; (2) the majority were conducted before the establishment of MCS
criteria (published in 2002) and, therefore, considered VS/UWS and MCS as a sin-
gle condition, which is not correct; (3) the inclusion criteria were poorly defined;
and finally (4) the number of patients included in these studies was limited, making
findings difficult to generalize.
Recently, the Belgian Federal Ministry of Public Health launched a project on
medical attitudes toward patients suffering from altered state of consciousness [11].
During the past 6 years, medical history, clinical, and epidemiological data were
collected in 15 Belgian neurorehabilitation expertise centers [11]. These data allowed
us to improve our knowledge of patients with altered state of consciousness – survival
and recovery. In order to determine patients’ consciousness state and to establish a
reliable diagnosis, the Coma Recovery Scale-Revised (CRS-R) [12] was performed
after 1, 3, 6, and 12 months post-injury. The prognosis was considered according to
14 M.-A. Bruno et al.

Table 2.2 Mortality and recovery rates of traumatic vegetative state/unresponsive wakefulness
syndrome patients
Duration
Authors N post-onset Follow-up Results
Higashi et al. [8, 17] 38 3 months 1–5 years 3-year cumulative mortality rate: 55%,
5-year: 66%
Bricolo et al. [14] 135 2 weeks 1 year 30% dead, 8% VS/UWS, 31% severe
disability, 18% moderate recovery,
13% good recovery. Age at onset
influences patients’ prognosis:
mortality rate (16% for patients
<20 years; 78% for patients
>60 years) and good recovery
(43% < 20 years and 0% > 60 years)
Braakman et al. [15] 140 1 month 1 year 51% dead, 11% VS/UWS, 26% severe
disability, 10% moderate disability
or recovery, 2% unknown (dead:
30% of after 3 months, 40% after
6 months). Duration of unconscious-
ness was negatively correlated with
the functional recovery. No patients
over 40 years showed any recovery
allowing him to living independently
Sazbon [18] 55 1 month 1 year 51% recovery of consciousness
Sazbon and 134 1 month 1 year 54% recovery of consciousness, 46%
Groswasser [5] VS/UWS
Sazbon and 134 1 month 1 year 3, 6, and 12-month cumulative
Groswasser [16] mortality rate: 7%, 19%, and 32%
5-year cumulative mortality rate: 68%
Multi-Society Task 434 1 month 1 year 33% dead, 15% VS/UWS, 52%
Force [10] recovery of consciousness: 28%
severe disability, 17% moderate
disability, 7% good recovery
Danze et al. [19] 522 1 month 1 year 19% dead, 20% VS/UWS, 61%
recovery of consciousness: 47%
severe disability, 12% moderate
disability, 2% good recovery.
Mortality rate was positively
correlated with the patient’s age
Dubroja et al. [20] 19 1 month 2 years 68% recovery of consciousness (58%
after 1 year, 5% after 2 years): 58%
moderate disability, 42% severe
disability
Giacino and Kalmar 30 11 ± 5 weeks 1 year 10% dead, 60% extremely severe
[21] disability or VS/UWS, 30%
moderate to severe disability, 0%
partial or good recovery
Estraneo et al. [22] 18 11 months 28 ± 12 39% dead, 17% VS/UWS, 44%
months recovery of consciousness and
functional recovery (17% MCS)
2 Prognosis of Patients with Altered State of Consciousness 15

Table 2.3 Mortality and recovery rates of non-traumatic vegetative state patients
Duration
Authors N Etiology since onset Follow-up Results
Levy et al. [25] 25 Non-traumatic 1 month 1 year 72% dead, 8% intelligible
words but no command
following, 12% visual
pursuit or localization
to sounds. None of the
patients can live
independently
Sazbon [18] 118 Non-traumatic 1 month 1 year 21% recovery of
consciousness
Sazbon 100 Anoxic 1 month 1–5 years 1 year post-onset: 46% dead,
et al. [24] 20% conscious (13%
after 3 months, 20% after
6 months, 17%
tetraparetic, 19%
cognitive disability, 15%
dysphasic and 13%
psychological disability),
33% unconscious, 1%
unknown
5 years post-onset: 68%
dead, 20% conscious,
7% unconscious, 5%
unknown
Cumulative mortality rate:
15% at 3 months, 31%
at 6 months
After 5 months of VS/UWS,
no recovery of conscious-
ness. Age at onset
influences motor status
and independence degree
in daily living but not the
cognitive status or
psychological disorders
Multi-Society 169 Non-traumatic 1 month 1 year 53% dead, 32% VS/UWS,
Task Force 11% severe disability,
[10] 3% moderate disability,
1% good recovery
Giacino and 34 Non-traumatic 11 ± 5 weeks 1 year 7% dead, 87% extremely
Kalmar [21] severe disability or VS/
UWS, 6% severe to
moderate disability, 0%
partial or total recovery
Estraneo 50 Hemorragic 12 months 26 ± 13 4% dead, 50% VS/UWS,
et al. [22] months 6% recovery of
consciousness (6%
MCS)
Anoxic 9 months 23 ± 12 43% dead, 36% VS/UWS,
months 20% recovery of
consciousness and
functional recovery
(7% MCS)
 16

Table 2.4 Mortality and recovery rates of minimally conscious state patients
Authors N Etiology Diagnostic Duration post-onset Follow-up Results
Grosswasser and 72 Traumatic MCS 1 month to 1 year More than 49% independent, 9% partially independent, 32%
Sazbon [27] 2 years dependant for daily activities, 40% no work,
49% sheltered workshops, 11% job similar to
the one practiced before the accident
Sazbon and 72 Traumatic MCS 1 month to 1 year 1 year VS/UWS < 2 months (n = 38): <50% independent
Groswasser [28] for daily living activities, <90% cognitive deficit
VS/UWS > 6 months (n = 4): 0% independent for
daily living activities, 100% cognitive deficits
Giacino 30 Traumatic MCS 9 ± 5 weeks 1 year 0% dead, 17% extremely severe disability or VS/
and Kalmar [21] UWS, 60% severe to moderate disability, 3%
partial or good recovery
10 Non-traumatic 20% dead, 30% extremely severe disability or VS/
UWS, 50% severe to moderate disability, 0%
total recovery
Lammi et al. [29] 18 Traumatic MCS 1 month 2–5 years 10% dead, 20% extremely severe disability or VS/
UWS, 55% severe to moderate disability, 15%
partial or good recovery
Katz et al. [30] 36 Traumatic and VS/UWS (11) 1 month 1–4 years Last evaluation showed that 72% emerged from the
Non- (3 non- MCS. Time to emerge was more important for
traumatic traumatic) VS/UWS patients as compared to MCS patients.
MCS (25) 73% of the VS/UWS patients evolved to a MCS
(11 non- (87% traumatic, 33% non-traumatic), 72% of the
traumatic) MCS patients emerged after 9 weeks (mean).
Luauté et al. [26] 39 Traumatic and MCS 1 year 5 years 36% dead, 33% severe disability, 23% MCS.
Non- Information no available for three patients
traumatic
M.-A. Bruno et al.
2 Prognosis of Patients with Altered State of Consciousness 17

Vegetative state/Unresponsive
wakefulness syndrome (n = 116) Minimally concscious state (n = 84)
% 100 %100
90 90
80 80
70 70
Traumatic

60 60
50 50
40 40
30 30
20 20
10 n = 52 10 n = 35
0 0
1 3 6 12 1 3 6 12

% 100 % 100
90 90
80 80
Non-traumatic

70 70
60 60
50 50
40 40
30 30
20 20
10 n = 64 10 n = 49
0 0
1 3 6 12 1 3 6 12
EMERGENCE MCS Dead VS/UWS

Fig. 2.2 Clinical evolution at 3, 6, and 12 months (abscissa) post-onset for vegetative state (VS/
UWS) patients and for minimally conscious state (MCS) patients. In red: VS/UWS patients; in
blue: MCS; in black: dead patients; in green: emergence from MCS

the three dimensions mentioned above – survival/mortality, recovery of conscious-

ness (MCS), and functional recovery (emergence from MCS). Our study is one of the
few to have established a clear distinction between recovery of consciousness and
functional recovery. This distinction allowed us to identify (1) VS/UWS patients (at
1 month post-injury) who progressed to MCS 1 year after the insult (recovery of
consciousness); (2) VS/UWS patients that emerged from MCS (functional recovery);
and finally (3) MCS patients (1 month after the accident) who emerged from this state
(functional recovery) 1 year later. Two hundred patients were included in our study:
116 patients were diagnosed VS/UWS and 84 patients were diagnosed MCS 1 month
after injury. Among VS/UWS patients, 52 were traumatic cases (traffic accident,
work accident, fall, violence, etc.) and 64 were Non-traumatic cases (anoxia, stroke,
intoxication, infection, brain tumor, etc.). Of the 84 MCS patients, etiology was trau-
matic in 35 and non-traumatic in 49 (Table 2.1, Fig. 2.2). Compared to previous
findings, our study presents some important strengths: (1) it includes a large number
of patients, (2) it properly assigns a standardized diagnosis to patients, (3) it differen-
tiates the prognosis for VS/UWS and MCS patients using etiology – traumatic versus
Non-traumatic – as adjustment factor.
18 M.-A. Bruno et al.

Mortality Rate

In our study, we found that in traumatic VS/UWS patients, 1 year post-onset mortal-
ity rate was 42% [13]. These results are in line with previous studies showing VS/
UWS patients mortality rates ranging from 30% to 50% [10, 14–16]. In 1980,
Bricolo and colleagues studied the prognosis of 135 patients who stayed in a
“nonresponsive state” at least 2 weeks post-onset and had a mortality rate of 30%
after 1 year [14]. Braakman and colleagues followed, for 1 year, 140 VS/UWS
patients and observed a mortality rate of 51% [15]. Finally, the Multi-Society Task
Force studied 434 patients and reported a mortality rate of 33% [10] (Table 2.2).
Regarding Non-traumatic VS/UWS patients’ mortality rate, previous studies
showed rates ranging from 50% to 70%. In 1978, Levy and colleagues reported a
mortality rate of 70% at 1-year post-injury (n = 25) [23] – this is in line with our
observation (mortality rate of 70%; n = 64). The Multi-Society Task Force [10],
Estraneo and colleagues [22], and Sazbon and colleagues [24] report mortality rates
of 53% (n = 169), 43% (n = 32), and 46% (n = 100), respectively. These observed dif-
ferences could be explained by the limited number of patients included in these stud-
ies [22, 23] and by the choice of inclusion criteria (duration of the non-response state
ranging from 1 month [23] to 1 year [22] (post-onset)) (Table 2.3). Finally, like in
previous studies, our study revealed a higher mortality rate in non-traumatic VS/
UWS patients as compared to traumatic VS/UWS patients [5, 8, 10, 17, 22, 24, 26].
To date, only a few studies have examined MCS patients’ mortality rates
(Table 2.4), and most of them presented some limitations. Indeed, due to the low
number of observations [21, 29] and the lack of information concerning patients’
etiology [26], it is challenging to generalize findings from the literature. In 2005,
Lammi and colleagues studied 18 traumatic MCS patients and observed, 1 year
post-injury, a 10% mortality rate [29]. In the works of Giacino and Kalmar, mortal-
ity rates were 20% for Non-traumatic MCS patients (n = 10), and there were no
death in MCS traumatic patients (n = 30) [21]. Finally, a recent study published by
Luauté and collaborators [26] showed a mortality rate of 36%; however, no distinc-
tion was made for etiologies (n = 36). In our recent study, we showed a 23% mortal-
ity rate in traumatic MCS patients (n = 35) and 33% in non-traumatic patients
(n = 49) [13]. These findings expand the knowledge on the prognosis of patients suf-
fering from disorders of consciousness by providing information on MCS patients’
mortality rate and are in line with previous studies showing prognosis differences
according to the etiology of VS/UWS or MCS [10, 21].

Recovery of Consciousness and Functional Recovery

VS/UWS Patients

In VS/UWS patients, consciousness recovery and functional recovery can be con-

sidered as a single category. We will then use the general term “recovery.” In the
literature, recovery rates range from 40% to 60% for traumatic VS/UWS patients.
2 Prognosis of Patients with Altered State of Consciousness 19

In our study, we found a recovery rate of 37% 1 year post-onset (n = 52) [13]. Results
obtained by Braakman and colleagues (n = 140) [31] and Estraneo and colleagues
(n = 18) [22] are in line with our study, showing recovery rates of 36% and 44%,
respectively. Sazbon and colleagues showed a rate of 51% (n = 55) [18], the Multi-
Society Task Force a rate of 52% (n = 434) [10], Giacino and Kalmar of 56% (n = 30)
[21], Dubroja and colleagues [20] of 58% (n = 19), and Bricolo and colleagues [14]
observed a recovery rate of 62% (n = 135). The variability of these results could be
explained by the patients’ inclusion criteria and by the definition of “recovery” used
by the different authors. Concerning non-traumatic VS/UWS patients, our results
are in line with previous studies showing recovery rate ranging from 0% to 21%,
1 year post-onset [10, 18, 23, 24, 32].
With regard to the recovery of consciousness in VS/UWS patients (VS/UWS
patients evolving to MCS), we observed that 14% of traumatic patients (n = 52) and
8% of non-traumatic patients (n = 64) recovered consciousness [13]. Our results are
in line with those of Estraneo and colleagues [22] who followed for 2 years 18 trau-
matic VS/UWS patients and 32 non-traumatic VS/UWS patients. They noted a con-
sciousness recovery rate of 17% for traumatic patients and 7% for non-traumatic
patients. If one focuses on the functional recovery in VS/UWS patients (patients
emerging from MCS), findings from previous studies are difficult to compare since
each study assessed functional recovery with different criteria and/or scales. Some
studies estimated recovery using the Glasgow Outcome Scale [33] while others
used the Disability Rating Scale [34]. The Glasgow Outcome Scale includes five
categories. A score of 5 corresponds to a good recovery (independent patients with
or without neurological deficit), a score of 4 is given to patients with moderate dis-
ability (independent patient with moderate neurological and/or intellectual deficit),
a score of 3 corresponds to severe disability (conscious patient but totally dependent
in daily living activities), a score of 2 characterizes VS/UWS patients, and finally a
score of 1 is assigned to dead patients. The Disability Rating Scale assesses the level
of functioning of patients by scoring their level of ability. A total score of 30 is
assigned to dead patients, a score between 25 and 29 correspond to patients who are
extremely vegetative, 22–24 for VS/UWS patients, 17–21 characterized patients
with extreme severe deficits, 12–16 characterized patients with severe deficits, 7–11
moderate to severe deficits, 4–6 moderate deficits, 2–3 partial or light deficits, and
finally a score of 0 is given to patients with no deficit. It has to be noted that with
these two scales, MCS patients and patients who emerged from MCS would be
classified in the same category; they would obtain the same score of 3 with the
Glasgow Outcome Scale and a score between 17 and 21 with the Disability Rating
Scale. This means that patients with different level of functional recovery would
receive the same scoring and be classified in a single category. This obviously raises
several difficulties in the quantification, interpretation, and generalization of results
reported by studies using such assessment tools.
In our research, we report a functional recovery rate – emergence from MCS 1 year
after initial brain injury – of 23% in traumatic patients who were VS/UWS 1 month
after onset [13]. A previous study on traumatic VS/UWS patients outcome, from
Bricolo and collaborators [14], reports 1 year posttrauma severe disability in 31%,
moderate recovery in 18%, and good recovery in 13% (n = 135) of observed patients.
20 M.-A. Bruno et al.

Braakman and colleagues noted 26% of severe disability and 10% of moderate
or complete recovery (defined as patients being able to live independently) 1 year
post-onset (n = 140). In 1994, out of 522 patients, Danze and colleagues reported
that 14% of patients recovered cognitive and motor capacities allowing them to live
independently [19]. Giacino and Kalmar [21] reported a recovery rate of 30%
(n = 30), and finally the Multi-Society Task Force found that 28% of studied patients
presented severe deficits, 17% moderate deficits, and 7% good recovery 1 year post-
onset (n = 434) [10]. Although these results are difficult to generalize, we can esti-
mate that functional recovery rate in traumatic VS/UWS patients ranges between
10% and 30%.
It is, however, important to mention that the rate of patients living independently
is very limited – ranges between 0% and 14%. Our study showed that non-traumatic
VS/UWS patients presented a functional recovery rate as low as 2% 1 year after the
onset.
Similarly, Levy and colleagues [23] observed that, out of their 25 studied patients,
no patient recovered a functional status that would allow independent living. The
Multi-Society Task Force showed that, out of the 169 patients studied, 3% presented
moderate deficits and only 1% had a good recovery [10]. One should note that late
recovery (more than a year post-injury) remains possible and occurs more frequently
for traumatic patients [8, 10, 35–37].

MCS Patients

Our study showed that the rate of traumatic MCS patients emerging from MCS –
functional recovery – is about 48%. Giacino and Kalmar [21] assessed the degree of
recovery in 30 MCS patients using the Disability Rating Scale, and reported 23% of
complete or partial recovery and 60% of poor to moderate recovery. Grosswasser
and Sazbon [27] assessed functional recovery in 72 traumatic MCS patients (out of
a group of 134 patients) and found that 49% of them could live independently, while
19% were partly dependent and 32% remained dependent in daily activities.
Regarding the patients’ occupational status, 11% found an employment similar to
the employment they had before the accident, 49% of the patients are employed in
sheltered workshops, and 40% were unable to reenter in the world of work.
In 2005, Lammi and colleagues [29] studied the recovery of 18 traumatic MCS
patients and showed that 55% of them presented moderate to severe deficits whereas
15% presented good recovery or minor deficits.
For non-traumatic MCS patients, our study showed that functional recovery occurred
in 26% of the cases. Only one study looked at the functional recovery rate of non-
traumatic MCS patients and showed moderate to severe deficits in 50% of the cases
while no good recovery was observed (n = 10) [21]. However, this study included only
a few numbers of patients making results impossible to interpret. We also found that
functional recovery rate (emergence from MCS – functional communication and/or
functional objects use) was higher for MCS patients as compared to VS/UWS patients,
2 Prognosis of Patients with Altered State of Consciousness 21

regardless of etiology. Indeed, 23% of the traumatic VS/UWS patients as compared to

48% of MCS patients recovered a functional communication and/or a functional use of
objects 1 year post-onset.
Our results are similar for non-traumatic patients, 2% of functional recovery for
VS/UWS patients and 26% for MCS patients – in line with previous studies evaluat-
ing the functional recovery of VS/UWS and MCS patients [21, 26, 30].
Note also that a study including 72 traumatic patients showed that the dura-
tion of unconsciousness in the acute stage influenced the prognosis 1 year post-
onset. Patients who did not emerge from the VS/UWS within 6 months had a
worse disability level than those who showed signs of consciousness during
this period [28].
Finally, other studies have demonstrated the importance of the age at the time of
the accident. Children between 5- and 6-years old are more likely to present favor-
able prognosis than adults [8, 10, 15, 38–41]. Patients older than 40 years have less
chance to present favorable prognosis than young patients [10, 27, 42]. However,
these results need to be confirmed by further studies based on a larger number of
patients, with clearly defined inclusion criteria, and with a better distinction between
degrees of recovery (MCS or emergence from MCS).

Conclusion

Data from our study showed that mortality rate in VS/UWS patients was higher than
in MCS patients, regardless of the etiology. In addition, patients with non-traumatic
brain injury have a higher mortality rate than traumatic patients. The same observa-
tions were noted regarding the patients’ degree of functional recovery. We observed
that, whatever the etiology, MCS patients have a better functional recovery than VS/
UWS patients. Similarly, traumatic patients were more likely to present functional
recovery than patients with a non-traumatic etiology [13].
In patients with altered state of consciousness, a reliable prognosis is a funda-
mental aspect of the care, from the intensive care units to the long-term life institu-
tion. Surgical, medical, or ethical decisions will depend upon this information.
While it is legitimate to set up a high medical and therapeutic cares for patients with
a good prognosis, it may not be desirable for most medical and paramedical teams
to promote such treatments when the predictable prognosis is bad. Scientific
advances allow a better understanding of the mortality rate and the degree of recov-
ery in VS/UWS and MCS patients; however, medical staff and the family should be
cautious regarding ethical decisions. We demonstrated that the MCS entity is differ-
ent from the VS/UMS, the outcome of MCS patients being significantly better than
that of VS/UWS patients. In addition, we know that the etiology of brain injury
influences patients’ prognosis. Traumatic patients are more likely to recover func-
tional communication than non-traumatic patients. However, further studies still
need be done to better characterize the functional recovery of patients who emerged
from MCS and to better quantify the cognitive impairments of these patients.
22 M.-A. Bruno et al.

References

1. Teasdale G, Jennett B. Assessment of coma and impaired consciousness. A practical scale.

Lancet. 1974;2:81–4.
2. Boveroux P, Kirsch M, Boly M, et al. Neurologic prognosis assessment in postanoxic enceph-
alopathy. Reanimation. 2008;17:613–7.
3. Kirsch M, Boveroux P, Massion P, et al. Comment prédire l’évolution du coma post-anoxique
? Rev Med Liege. 2008;63:263–8.
4. Wijdicks EF, Hijdra A, Young GB, et al. Practice parameter: prediction of outcome in comatose
survivors after cardiopulmonary resuscitation (an evidence-based review): report of the Quality
Standards Subcommittee of the American Academy of Neurology. Neurology. 2006;67:203–10.
5. Sazbon L, Groswasser Z. Outcome in 134 patients with prolonged posttraumatic unawareness.
Part 1: parameters determining late recovery of consciousness. J Neurosurg. 1990;72:75–80.
6. Jennett B, Teasdale G, Braakman R, et al. Predicting outcome in individual patients after
severe head injury. Lancet. 1976;1:1031–4.
7. The Multi-Society Task Force on PVS. Medical aspects of the persistent vegetative state (1).
N Engl J Med. 1994;330:1499–508.
8. Higashi K, Sakata Y, Hatano M, et al. Epidemiological studies on patients with a persistent
vegetative state. J Neurol Neurosurg Psychiatry. 1977;40:876–85.
9. Giacino JT, Ashwal S, Childs N, et al. The minimally conscious state: definition and diagnostic
criteria. Neurology. 2002;58:349–53.
10. The Multi-Society Task Force on PVS. Medical aspects of the persistent vegetative state (2).
N Engl J Med. 1994;330:1572–9.
11. Demotte R. Politique de la santé à mener à l’égard des patients en état végétatif persistant ou
en état pauci-relationnel. Moniteur Belge 2004:69334–40.
12. Giacino JT, Kalmar K, Whyte J. The JFK coma recovery scale-revised: measurement charac-
teristics and diagnostic utility. Arch Phys Med Rehabil. 2004;85:2020–9.
13. Bruno MA, Gosseries O, Vanhaudenhuyse A, et al. État végétatif et état de conscience mini-
male: un devenir pire que la mort ? In: Puybasset L, editor. Enjeux éthiques en réanimation.
Paris: Springer; 2010. p. 119–28.
14. Bricolo A, Turazzi S, Feriotti G. Prolonged posttraumatic unconsciousness: therapeutic assets
and liabilities. J Neurosurg. 1980;52:625–34.
15. Braakman R, Jennett WB, Minderhoud JM. Prognosis of the posttraumatic vegetative state.
Acta Neurochir. 1988;95:49–52.
16. Sazbon L, Groswasser Z. Medical complications and mortality of patients in the postcomatose
unawareness (PC-U) state. Acta Neurochir (Wien). 1991;112:110–2.
17. Higashi K, Hatano M, Abiko S, et al. Five-year follow-up study of patients with persistent
vegetative state. J Neurol Neurosurg Psychiatry. 1981;44:552–4.
18. Sazbon L. Prolonged coma. Prog Clin Neurosci. 1985;2:65–81.
19. Danze F, Veys B, Lebrun T, et al. Prognostic factors of post-traumatic vegetative states: 522
cases. Neurochirurgie. 1994;40:348–57. discussion 357–348.
20. Dubroja I, Valent S, Miklic P, Kesak D. Outcome of post-traumatic unawareness persisting for
more than a month. J Neurol Neurosurg Psychiatry. 1995;58:465–6.
21. Giacino JT, Kalmar K. The vegetative and minimally conscious states: a comparison of clinical
features and functional outcome. J Head Trauma Rehabil. 1997;12:36–51.
22. Estraneo A, Moretta P, Loreto V, et al. Late recovery after traumatic, anoxic, or hemorrhagic
long-lasting vegetative state. Neurology. 2010;75:239–45.
23. Levy DE, Knill-Jones RP, Plum F. The vegetative state and its prognosis following nontrau-
matic coma. Ann N Y Acad Sci. 1978;315:293–306.
24. Sazbon L, Zagreba F, Ronen J, et al. Course and outcome of patients in vegetative state of
nontraumatic aetiology. J Neurol Neurosurg Psychiatry. 1993;56:407–9.
25. Levy DE. Coma after cardiac arrest–how Much care is enough? [editorial]. West J Med.
1987;147:191–3.
2 Prognosis of Patients with Altered State of Consciousness 23

26. Luaute J, Maucort-Boulch D, Tell L, et al. Long-term outcomes of chronic minimally con-
scious and vegetative states. Neurology. 2010;75:246–52.
27. Groswasser Z, Sazbon L. Outcome in 134 patients with prolonged posttraumatic unawareness.
Part 2: functional outcome of 72 patients recovering consciousness. J Neurosurg. 1990;72:81–4.
28. Sazbon L, Groswasser Z. Time-related sequelae of TBI in patients with prolonged post-coma-
tose unawareness (PC-U) state. Brain Inj. 1991;5:3–8.
29. Lammi MH, Smith VH, Tate RL, Taylor CM. The minimally conscious state and recovery
potential: a follow-up study 2 to 5 years after traumatic brain injury. Arch Phys Med Rehabil.
2005;86:746–54.
30. Katz DI, Polyak M, Coughlan D, et al. Natural history of recovery from brain injury after
prolonged disorders of consciousness: outcome of patients admitted to inpatient rehabilitation
with 1–4 year follow-up. Prog Brain Res. 2009;177:73–88.
31. Braakman R, Gelpke GJ, Habbema JD, et al. Systematic selection of prognostic features in
patients with severe head injury. Neurosurgery. 1980;6:362–70.
32. Giacino JT. The vegetative and minimally conscious states: consensus-based criteria for estab-
lishing diagnosis and prognosis. NeuroRehabilitation. 2004;19:293–8.
33. Jennett B, Bond M. Assessment of outcome after severe brain damage. Lancet. 1975;1:480–4.
34. Rappaport M, Hall KM, Hopkins K, et al. Disability rating scale for severe head trauma: coma
to community. Arch Phys Med Rehabil. 1982;63:118–23.
35. Avesani R, Gambini MG, Albertini G. The vegetative state: a report of two cases with a long-
term follow-up. Brain Inj. 2006;20:333–8.
36. Childs NL, Mercer WN. Late improvement in consciousness after post-traumatic vegetative
state. N Engl J Med. 1996;334:24–5.
37. Voss HU, Uluc AM, Dyke JP, et al. Possible axonal regrowth in late recovery from the mini-
mally conscious state. J Clin Invest. 2006;116:2005–11.
38. Kriel RL, Krach LE, Jones-Saete C. Outcome of children with prolonged unconsciousness and
vegetative states. Pediatr Neurol. 1993;9:362–8.
39. Johnston RB, Mellits ED. Pediatric coma: prognosis and outcome. Dev Med Child Neurol.
1980;22:3–12.
40. Bruce DA, Schut L, Bruno LA, et al. Outcome following severe head injuries in children.
J Neurosurg. 1978;48:679–88.
41. Ashwal S. The persistent vegetative state in children. Adv Pediatr. 1994;41:195–222.
42. Strauss DJ, Shavelle RM, Ashwal S. Life expectancy and median survival time in the perma-
nent vegetative state. Pediatr Neurol. 1999;21:626–31.
Chapter 3
Functional Imaging and Impaired Consciousness

Audrey Vanhaudenhuyse, Melanie Boly, Marie-Aurélie Bruno, Olivia Gosseries,

Athena Demertzi, Murielle Kirsch, Jean-Flory Luaba Tshibanda,
Marie Thonnard, Vanessa Charland-Verville, Camille Chatelle,
Aurore Thibaut, Steven Laureys, and Andrea Soddu

A. Vanhaudenhuyse, Ph.D. (*) • M.-A. Bruno, Ph.D. • A. Demertzi, M.Sc.

V. Charland-Verville, M.Sc. • S. Laureys, M.D., Ph.D. • M. Thonnard, M.Sc.
Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du 6 Août, Sart Tilman B30, 4000 Liège, Belgium
Department of Neurology, University Hospital of Liège, Liège, Belgium
e-mail: [email protected]; [email protected]; [email protected];
[email protected]; [email protected]; [email protected]
M. Boly, M.D., Ph.D. • A. Soddu, Ph.D.
Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du 6 Août, Sart Tilman B30, 4000 Liège, Belgium
e-mail: [email protected]; [email protected]
O. Gosseries, M.Sc. • C. Chatelle • A. Thibaut
Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du 6 Août, Sart Tilman B30, 4000 Liège, Belgium
Department of Neurology, University Hospital of Liège, Liège, Belgium
National Fund for Scientific Research, Liège, Belgium
e-mail: [email protected]
M. Kirsch, M.D.
Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du 6 Août, Sart Tilman B30, 4000 Liège, Belgium
Department of Anesthesiology, University Hospital of Liège, Sart Tilman, 4000 Liège, Belgium
e-mail: [email protected]
J.-F.L. Tshibanda, M.D.
Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du 6 Août, Sart Tilman B30, 4000 Liège, Belgium
Department of Radiology, University Hospital of Liège, Sart Tilman, 4000 Liège, Belgium
e-mail: [email protected]

C. Schnakers, S. Laureys (eds.), Coma and Disorders of Consciousness, 25

DOI 10.1007/978-1-4471-2440-5_3, © Springer-Verlag London 2012
26 A. Vanhaudenhuyse et al.

Introduction

Differentiating a voluntary from a reflexive behavior in patients unable to communi-

cate remains a big challenge. Currently, 40% of patients with disorders of conscious-
ness are misdiagnosed as being in vegetative state when assessed with nonstandardized
scales [1]. Along with more traditional electroencephalographic techniques, mag-
netic resonance imaging (MRI) and positron emission tomography (PET) offer today
the opportunity to study the neural correlates of consciousness in patients with altered
states of consciousness as compared to healthy subjects. While some methods are
devoted to detecting cortical responses during sensory stimulation (tactile, auditory,
visual, etc.), others propose to measure neural activity at rest or during “active” para-
digms, requiring participation of the subject who have to perform mental imagery
tasks. In this chapter, we propose a differentiated approach to study brain functions
of patients in an altered state of consciousness, by combining multimodal assessment
techniques such as PET and functional MRI (fMRI), to distinguish patients in a veg-
etative state (unconscious) from patients in a minimally conscious state (Fig. 3.1).

Brain Activity at Rest: Is There Any Implication

for Consciousness?

Several studies using PET have shown a massive decrease in global cerebral metab-
olism for patients in vegetative state and coma, ranging from 40% to 50% (for a
review, see [2]). However, patients recovering from a vegetative state do not show a

Resting state cerebral metabolism

Spontaneous brain activity correlated to the

level of consciousness

(stimulus independent)

Passive paradigmes
Cerebral function
assessment with fMRI Brain activity induced by stimuli
and/or PET
(automatic functions integrity)

Active paradigmes

Brain activity reflecting command following

(sign of consciousness)

Fig. 3.1 Differentiated approach to study brain functions in patients with an altered state of con-
sciousness combining multimodal assessment techniques such as positron emission tomography
(PET) and functional magnetic resonance imaging (fMRI) (Adapted from [2])
3 Functional Imaging and Impaired Consciousness 27

systematic recovery of global brain metabolism to normal values [3]. Similarly,

some healthy subjects demonstrate a metabolism comparable to what is observed in
vegetative state patients [4]. Furthermore, vegetative state is characterized by a dys-
function of local rather than global cerebral areas, and specifically of the regions
encompassing bilateral frontal, parieto-temporal, posterior parietal, mesiofrontal,
posterior cingulate, and precuneal cortices [5, 6]. These brain areas are consistently
more hypometabolic in vegetative state than in minimally conscious state patients
[6], and are usually preserved in patients emerging from the minimally conscious
state who recovered the ability to communicate [7]. Moreover, awareness seems
related not exclusively to the activity in the frontoparietal network but, as impor-
tantly, to the functional connectivity within this network and the thalami. Indeed,
long-range cortico-cortical and cortico-thalamo-cortical “functional disconnec-
tions” have been identified in the vegetative state [8]. The vegetative state seems
also characterized by hypermetabolism in the reticular ascending system and a
deficient functional connectivity between this system and the precuneus [9]. Finally,
in a recent study measuring cerebral metabolism, we were able to demonstrate that
the presence of visual fixation in nontraumatic brain-injured patients otherwise
showing typical clinical characteristics of vegetative state was not accompanied by
a different brain activity compared to patients in vegetative state showing no visual
fixation [10]. Patients without visual fixation showed hypometabolism in the fronto-
temporo-parietal cortex and thalami, similar to patients with visual fixation. No dif-
ference in recovery of consciousness could be identified between the two groups.
These recent findings suggest that sustained visual fixation in nontraumatic dis-
orders of consciousness should be considered carefully when giving a diagnosis
of consciousness. In addition to the interest in measuring brain metabolism, an
increasing attention is also drawn to spontaneous brain activity and its possible rel-
evance for cognitive behavior [11]. Gusnard and Raichle [12] proposed the concept
of default mode network after observing that most of the studies in PET and func-
tional MRI (fMRI) reported a set of brain areas exhibiting task-induced deactivations,
encompassing precuneus, posterior parietal lobe, and medial prefrontal cortex,
which are more active at rest than when subjects were involved in attention-demand-
ing cognitive tasks. These decreases in activity are considered as a reflection of
suppression in information processing for the brain regions not involved in the task
[13], and have a quite consistent spatial pattern [12]. If the exact function of the
default mode network is still debated, some studies suggest that it could be involved
in conscious perception processing. Moreover, the default mode network is known
to be implicated in specific cognitive processes such as internal and stimuli indepen-
dent thoughts [14, 15]. The study of the default mode network has a clinical interest
in patients in an altered state of consciousness, since it allows the investigation of
higher-order cognitive networks without requiring patients’ active participation,
particularly important in vegetative and minimally conscious patients. A recent
work showed that no default mode network was detected in a brain death patient [2,
16] (Fig. 3.2), while vegetative state was characterized by a partial preservation of
connectivity in the regions of the default mode network [16, 19]. Moreover, we
demonstrated that there was a correlation between the level of connectivity within
28 A. Vanhaudenhuyse et al.

Controls Brain death 7

−8.00 6
5
4
3
−3.45
8.00 2
1
0
3.45
−1
t(167)
Controls MCS VS COMA

Fig. 3.2 Default mode network observed in healthy subjects (yellow). Brain death patients are
characterized by a total absence of the network. The contours in black and white show the default
mode network observed in a group of independent control subjects. Chart represents the nonlinear
correlation between the level of consciousness of patients and the degree of connectivity of the
precuneus with all the other regions of the default mode network (Adapted from [2, 51]). Locked-in
syndrome: red dot, MCS minimally conscious state, VS vegetative state (Adapted from [2, 18])

the default mode network and the patients’ level of consciousness (patients in coma,
vegetative state, minimally conscious state and locked-in syndrome) [18] (Fig. 3.2).
Specifically, the highest the drop in the level of awareness (coma, vegetative state),
the most the degree of connectivity in the default mode network decreases. In addi-
tion, posterior cingulate cortex/precuneus was the region which better disentangles
patients in vegetative and minimally consciousness state [18]. These results seem to
sustain the hypothesis that the precuneus has a specific role in the architecture of the
default mode network, as already indicated by functional [17] and structural [20]
findings. All these results suggest that the default mode network is involved in the
process of consciousness, completing previous studies, which show a partial preser-
vation of the default mode network in light sleep [21] or in monkeys under general
anesthesia [22]. These results also confirm recent studies showing a decrease in
network connectivity in the default mode network in sedated healthy subjects [23]
and deep sleep [24]. Finally, it is shared opinion that if the described findings seem
very encouraging, most still need to be done to develop interpretable paradigms at
the individual level, and therefore applicable for a clinical diagnosis of patients with
severe brain injury, by developing techniques considering physiological and move-
ment artifacts on the default mode network [25].

Brain Activation and External Stimuli: Conscious Perception?

Sensory perceptions (i.e., somatosensory, auditory, and visual) are related to con-
scious perception of stimulation. By definition, patients in vegetative state cannot
consciously perceive these stimuli. However, we cannot exclude from behavioral
observations that the stimuli presented are not perceived. In this context, various
neuroimaging studies have investigated the brain activity observed during the
3 Functional Imaging and Impaired Consciousness 29

presentation of different stimuli in disorders of consciousness patients. Several

fMRI studies [26–29] have confirmed previous PET results, showing preserved
brain activity in primary sensory cortex, but this activation is disconnected from all
higher associative areas in patients in vegetative state both for auditory [30–32] and
somatosensory [33, 34] stimulations. These studies also report that patients in mini-
mally conscious state, compared to vegetative patients, have a much more distrib-
uted brain activity and cortico-cortical connectivity in response to stimuli, including
nociceptive stimuli. These results emphasize the importance to adapt analgesic
treatment in minimally conscious patients considering that their brain response dur-
ing painful stimuli is very similar to the one observed in healthy subjects [33, 34].
In addition, for patients in minimally conscious state, auditory stimuli with emo-
tional valence (i.e., cry of a baby, own name, and story narrated by a familiar voice)
involve more areas than the ones observed in response to neutral stimuli in accor-
dance to that observed in healthy individuals [35–37]. The activity of the anterior
cingulate cortex has recently been demonstrated to correlate with the level of con-
sciousness during the presentation of self-referential stimuli (patient’s own name
[38]). The highest the level of consciousness – in a minimally conscious state as
compared to vegetative state patients – the greatest is the intensity of activation of
the cingulate cortex. The fact that higher order areas are activated in minimally
conscious state patients suggests that the content of our words may be relevant for
the patient. Finally, some patients in vegetative state, even if exceptionally, may also
show some activations in higher order areas in response to auditory [27, 32] or
visual [32, 39] stimuli. The presence of unusual activity has been proposed as a pos-
sible prognostic marker for recovery of consciousness [40].

Active fMRI Paradigms: Help in the Diagnosis of Altered

States of Consciousness?

Without a full understanding of brain mechanisms underlying consciousness, brain

activation in patients similar to that observed in healthy subjects cannot be attributed
to a conscious perception. The only conclusion one can draw is that these specific
regions are able to process sensory stimuli [2]. Therefore, the question is how to
distinguish an automatic brain response to a conscious perception. To overcome the
problematic of disentangling automatic from conscious neuronal activity, Owen et al.
[41] have proposed a different approach consisting in applying an active fMRI
paradigm. A patient initially diagnosed as being in vegetative state was instructed to
perform two mental imagery tasks, previously validated in a series of healthy volun-
teers [42]. In this study, when the patient was asked to “imagine playing tennis,” an
activation of the supplementary motor area was recorded; while the activation of
parahippocampal gyri was observed when the patient was asked to “imagine visiting
her own house.” Such brain activations, similar to those observed in healthy volun-
teers, showed that the patient, on one hand, was able to understand instructions and, on
the other hand, she was able to perform the imagery tasks. Note that these activations
30 A. Vanhaudenhuyse et al.

could be interpreted only as coming from an intentional participation of the patient.

Indeed, even if the words “tennis” and “home” have already been reported as causing
automatic brain activity – not aware – this activity is known to last only few millisec-
onds and appears only in areas related to word processing [43]. Brain activity lasted
the entire period during which the patient was asked to perform the task (30 s) before
resting, and was localized in areas known for their involvement in motor imagery
tasks [42, 44, 45]. Owen et al. [45] conclude that such activations cannot be explained
in terms of automatic responses. Interestingly, 6 months after this study, the patient
showed behavioral signs of consciousness. Subsequently, this paradigm was applied
in 54 patients in an altered state of consciousness (23 in a vegetative state, 31 in a
minimally conscious state [46]). Among the studied patients, five (four initially diag-
nosed as being in a vegetative state and one in a minimally conscious state) showed
classical brain activation observed for mental imagery tasks – in the supplementary
motor area when patients were asked to perform the motor imagery task, and in the
parahippocampal gyrus during the visuospatial task. This paradigm was then adapted
into a yes/no answer communication system, allowing a patient with posttraumatic
etiology to communicate without using the traditional communication channels
(motor or language). This patient had received diagnosis of vegetative state 17 months
after his traumatic accident. Five years after his injury, behavioral assessments
showed reproducible signs of consciousness, reflecting a minimally conscious state.
When inside the fMRI scanner, the patient was asked various biographical questions
and instructed to answer “yes” by imagining playing tennis, or “no” by imagining
visiting his home. The brain activity observed during the question sessions was simi-
lar to that observed in the training session in which the patient showed the ability to
perform both imagery tasks [46], and was, respectively, translated in the yes/no
answers for five out of the six questions (last question remained unanswered). The
interpretation of these findings does not necessarily support the hypothesis that the
patient was initially misdiagnosed. Indeed, it is plausible that the patient was in a
vegetative state in the years before the fMRI scanning and that only gradually moved
away from the vegetative state without, however, being able to behaviorally demon-
strate his consciousness. Finally, 49 of the patients included in the latter study did not
show any brain activation consistent with performing an imagery task. It is now
difficult to say whether this absence is the result of a lack of sensitivity in the method,
or whether it really reflects the inability or limitations in cognitive performance from
the patients. Some patients could also be unconscious or sleeping during the fMRI
session, as the arousal level can quickly fluctuate in these patients. For others, even
in a minimally conscious state, the tasks could be cognitively too complex to be real-
ized. Indeed, deficits in processes related to language comprehension, working mem-
ory, decision-making, or executive functions are all factors that can hinder a success
of such tasks [47]. Conversely, one can conclude that patients’ ability to perform
these tasks demonstrates at least some preservation of cognitive processes. Voluntary
brain modulation in fMRI was also recently demonstrated as a device of nonfunc-
tional communication in a minimally conscious patient [48]. These studies show
clearly a means to detect potentially existing dissociations between the behavior
observed at the patients’ bedside and their residual brain activity, reflecting preserved
3 Functional Imaging and Impaired Consciousness 31

cognitive function [6, 49, 50]. This finding indicates that, in some patients, motor
function can be so impaired that bedside assessments based on the presence or
absence of a behavioral response may not reveal awareness, regardless of how thor-
oughly and carefully they are administered. These “active” paradigms must, there-
fore, be considered as an aid for diagnosis in patients with severe brain injury.

Conclusion

The difficulty in determining an accurate diagnosis and prognosis in patients surviv-

ing a coma has important ethical implications. Precise information based on
scientific findings on the diagnosis and prognosis of the patient is essential to make
the family understand and accept more easily medical decisions. The limits of
behavioral assessments must encourage us to consider other biomarkers, which may
be more objective for the detection of consciousness and may help clinicians in their
evaluation. The various paradigms presented in this chapter demonstrate the value
of neuroimaging techniques like PET or fMRI as diagnostic tools for patients in
altered states of consciousness. Paradigms that study spontaneous activity may
detect a residual level of consciousness without requiring the cooperation of patients.
Active paradigms allow the detection of cognitive functions without involving lan-
guage or motor responses. These paradigms give us the ability to differentiate veg-
etative from minimally conscious patients and can help to clarify the diagnosis of
patients when behavioral assessments are ambiguous. PET allows identifying brain
regions which present preserved metabolism, while fMRI allows identifying whether
these regions are functionally interconnected or are more active during a task than
at rest. By combining both techniques, more precise information regarding the func-
tionality of cortical and subcortical regions could be obtained and potentially
improve the diagnosis in patients with impaired consciousness. We also demon-
strated the possibility of establishing a basic communication paradigm through two
distinct (yes/no) imagery tasks inside the fMRI scanner. Unfortunately, this tech-
nique is impractical in everyday life but could definitely be used to allow patients
expressing themselves on relevant issues such as their quality of life or decisions
about their own care. Other methods, mostly based on electroencephalography,
which may be adapted to everyday life, have already been used and are nowadays
under development. By using these techniques, patients could express themselves
on pain sensation and feelings. This would allow clinicians to adapt the treatment of
patients, such as analgesic management for example. Finally, scientific advances in
the area of disorders of consciousness and the possible clinical applications, makes
possible to improve the diagnosis, refine the prognosis, and thus respond to ethical
issues. However, neuroimaging techniques are still limited by artifacts induced, e.g.,
by the patient’s movements inside the scanner, by artificial respiration, or by
implanted ferromagnetic elements. Future work is, therefore, needed to reduce as
much as possible these artifacts and new analyses need to be developed to eliminate
residual contamination of the data.
32 A. Vanhaudenhuyse et al.

References

1. Schnakers C, Vanhaudenhuyse A, Giacino J, et al. Diagnostic accuracy of the vegetative and

minimally conscious state: clinical consensus versus standardized neurobehavioral assess-
ment. BMC Neurol. 2009;9:35.
2. Bruno M, Soddu A, Demertzi A, et al. Disorders of consciousness: moving from passive to rest-
ing state and active paradigms. Cogn Neurosci. 2010;1(3):193–203.
3. Laureys S, Lemaire C, Maquet P, et al. Cerebral metabolism during vegetative state and after
recovery to consciousness. J Neurol Neurosurg Psychiatry. 1999;67:121.
4. Laureys S. The neural correlate of (un) awareness: lessons from the vegetative state. Trends
Cogn Sci. 2005;9:556–9.
5. Laureys S, Goldman S, Phillips C, et al. Impaired effective cortical connectivity in vegetative
state. Neuroimage. 1999;9:377–82.
6. Laureys S, Owen AM, Schiff ND. Brain function in coma, vegetative state, and related disor-
ders. Lancet Neurol. 2004;3:537–46.
7. Nakayama N, Okumura A, Shinoda J, et al. Relationship between regional cerebral metabo-
lism and consciousness disturbance in traumatic diffuse brain injury without large focal
lesions: an FDG-PET study with statistical parametric mapping analysis. J Neurol Neurosurg
Psychiatry. 2006;77:856–62.
8. Laureys S, Faymonville ME, Luxen A, et al. Restoration of thalamocortical connectivity after
recovery from persistent vegetative state. Lancet. 2000;355:1790–1.
9. Silva S, Alacoque X, Fourcade O, et al. Wakefulness and loss of awareness: brain and brain-
stem interaction in the vegetative state. Neurology. 2010;74:313–20.
10. Bruno MA, Vanhaudenhuyse A, Schnakers C, et al. Visual fixation in the vegetative state: an
observational case series PET study. BMC Neurol. 2010;10:35.
11. Raichle ME. Neuroscience. The brain’s dark energy. Science. 2006;314:1249–50.
12. Gusnard DA, Raichle ME. Searching for a baseline: functional imaging and the resting human
brain. Nat Rev Neurosci. 2001;2:685–94.
13. Drevets WC, Burton H, Videen TO, et al. Blood flow changes in human somatosensory cortex
during anticipated stimulation. Nature. 1995;373:249–52.
14. Mason MF, Norton MI, Van Horn JD, et al. Wandering minds: the default network and stimu-
lus-independent thought. Science. 2007;315:393–5.
15. Vanhaudenhuyse A, Demertzi A, Schabus M, et al. Two distinct neuronal networks mediate the
awareness of environment and of self. J Cogn Neurosci. 2011;23:570–8.
16. Boly M, Tshibanda L, Vanhaudenhuyse A, et al. Functional connectivity in the default network
during resting state is preserved in a vegetative but not in a brain dead patient. Hum Brain
Mapp. 2009;30:2393–400.
17. Fransson P, Marrelec G. The precuneus/posterior cingulate cortex plays a pivotal role in the
default mode network: evidence from a partial correlation network analysis. Neuroimage.
2008;42:1178–84.
18. Vanhaudenhuyse A, Noirhomme Q, Tshibanda LJ, et al. Default network connectivity reflects
the level of consciousness in non- communicative brain-damaged patients. Brain. 2010;133:
161–71.
19. Cauda F, Micon BM, Sacco K, et al. Disrupted intrinsic functional connectivity in the vegeta-
tive state. J Neurol Neurosurg Psychiatry. 2009;80:429–31.
20. Hagmann P, Cammoun L, Gigandet X, et al. Mapping the structural core of human cerebral
cortex. PLoS Biol. 2008;6:e159.
21. Horovitz SG, Fukunaga M, de Zwart JA, et al. Low frequency BOLD fluctuations during
resting wakefulness and light sleep: a simultaneous EEG-fMRI study. Hum Brain Mapp.
2008;29:671–82.
22. Vincent JL, Patel GH, Fox MD, et al. Intrinsic functional architecture in the anaesthetized
monkey brain. Nature. 2007;447:83–6.
3 Functional Imaging and Impaired Consciousness 33

23. Greicius MD, Kiviniemi V, Tervonen O, et al. Persistent default-mode network connectivity
during light sedation. Hum Brain Mapp. 2008;29:839–47.
24. Horovitz SG, Braun AR, Carr WS, et al. Decoupling of the brain’s default mode network dur-
ing deep sleep. Proc Natl Acad Sci USA. 2009;106:11376–81.
25. Soddu A, Vanhaudenhuyse A, Bahri MA, Bruno MA, Boly M, Demertzi A, Tshibanda JF,
Phillips C, Stanziano M, Ovadia-Caro S, Nir Y, Maquet P, Papa M, Malach R, Laureys S,
Noirhomme Q. Identifying the default-mode component in spatial IC analyses of patients with
disorders of consciousness. Hum Brain Mapp. 2012;33:778–96.
26. Coleman MR, Rodd JM, Davis MH, et al. Do vegetative patients retain aspects of language
comprehension? Evidence from fMRI. Brain. 2007;130:2494–507.
27. Di HB, Yu SM, Weng XC, et al. Cerebral response to patient’s own name in the vegetative and
minimally conscious states. Neurology. 2007;68:895–9.
28. Fernandez-Espejo D, Junque C, Vendrell P, et al. Cerebral response to speech in vegetative and
minimally conscious states after traumatic brain injury. Brain Inj. 2008;22:882–90.
29. Moritz CH, Rowley HA, Haughton VM, et al. Functional MR imaging assessment of a non-
responsive brain injured patient. Magn Reson Imaging. 2001;19:1129–32.
30. Boly M, Faymonville ME, Peigneux P, et al. Auditory processing in severely brain injured
patients: differences between the minimally conscious state and the persistent vegetative state.
Arch Neurol. 2004;61:233–8.
31. Laureys S, Faymonville ME, Degueldre C, et al. Auditory processing in the vegetative state.
Brain. 2000;123(8):1589–601.
32. Owen AM, Menon DK, Johnsrude IS, et al. Detecting residual cognitive function in persistent
vegetative state. Neurocase. 2002;8:394–403.
33. Boly M, Faymonville ME, Schnakers C, et al. Perception of pain in the minimally conscious
state with PET activation: an observational study. Lancet Neurol. 2008;7(11):1013–20.
34. Laureys S, Faymonville ME, Peigneux P, et al. Cortical processing of noxious somatosensory
stimuli in the persistent vegetative state. Neuroimage. 2002;17(2):732–41.
35. Laureys S, Perrin F, Faymonville ME, et al. Cerebral processing in the minimally conscious
state. Neurology. 2004;63:916–8.
36. Bekinschtein T, Leiguarda R, Armony J, et al. Emotion processing in the minimally conscious
state. J Neurol Neurosurg Psychiatry. 2004;75:788.
37. Schiff ND, Rodriguez-Moreno D, Kamal A, et al. FMRI reveals large-scale network activation
in minimally conscious patients. Neurology. 2005;64:514–23.
38. Qin P, Di H, Liu Y, et al. Anterior cingulate activity and the self in disorders of consciousness.
Hum Brain Mapp. 2010;31:1993–2002.
39. Menon DK, Owen AM, Williams EJ, et al. Cortical processing in persistent vegetative state.
Lancet. 1998;352:200.
40. Di H, Boly M, Weng X, et al. Neuroimaging activation studies in the vegetative state: predic-
tors of recovery? Clin Med. 2008;8:502–7.
41. Owen AM, Coleman MR, Boly M, et al. Detecting awareness in the vegetative state. Science.
2006;313:1402.
42. Boly M, Coleman MR, Davis MH, et al. When thoughts become action: an fMRI paradigm to
study volitional brain activity in non- communicative brain injured patients. Neuroimage.
2007;36:979–92.
43. Hauk O, Johnsrude I, Pulvermuller F. Somatotopic representation of action words in human
motor and premotor cortex. Neuron. 2004;41:301–7.
44. Weiskopf N, Mathiak K, Bock SW, et al. Principles of a brain- computer interface (BCI) based
on real-time functional magnetic resonance imaging (fMRI). IEEE Trans Biomed Eng.
2004;51:966–70.
45. Owen A, Coleman M, Boly M, et al. Response to comments on detecting awareness in the
vegetative state. Science. 2007;315:1221.
46. Monti MM, Vanhaudenhuyse A, Coleman MR, et al. Willful modulation of brain activity in
disorders of consciousness. N Engl J Med. 2010;362:579–89.
34 A. Vanhaudenhuyse et al.

47. Majerus S, Bruno MA, Schnakers C, et al. The problem of aphasia in the assessment of
consciousness in brain-damaged patients. Prog Brain Res. 2009;177:49–61.
48. Bardin JC, Fins JJ, Katz DI, Hersh J, Heier LA, Tabelow K, Dyke JP, Ballon DJ, Schiff ND,
Voss HU. Dissociations between behavioural and functional magnetic resonance imaging-
based evaluations of cognitive function after brain injury. Brain. 2011;134:769–82.
49. Monti MM, Coleman MR, Owen AM. Neuroimaging and the vegetative state: resolving the
behavioral assessment dilemma? Ann NY Acad Sci. 2009;1157:81–9.
50. Owen AM, Coleman MR. Functional neuroimaging of the vegetative state. Nat Rev Neurosci.
2008;9:235–43 Schnackers.
51. Tshibanda L, Vanhaudenhuyse A, Boly M, et al. Neuroimaging after coma. Neuroradiology.
2010;52:15–24.
Chapter 4
Multivariate Pattern Recognition Analysis:
Brain Decoding

Jessica Schrouff and Christophe L.M. Phillips

Introduction

Two of the most fundamental questions in the field of neuroscience are how infor-
mation is represented in different brain structures, and how this information evolves
over time. Various tools, such as magnetic resonance imaging (MRI) and positron
emission tomography (PET), have been developed over the last few decades to
record brain activity and investigate these questions. In particular, functional MRI
(fMRI) tracks changes of the blood oxygenation level-dependent (BOLD) signal,
which is a good indicator of brain activity [1], with a spatial resolution of a few
cubic millimeters and a typical temporal resolution in the order of 1 or 2 s.
Until recently, the methods used to analyze such data focused on characterizing
the individual relationship between a cognitive or perceptual state and each image
voxel, i.e., following a massively univariate statistical approach. A well-known uni-
variate technique is statistical parametric mapping (SPM) [2]. SPM relies on the
general linear model [3] to detect which voxels show a statistically significant
response to the (combination of) experimental conditions of interest. However, there
are limitations on what can be learned about the representation of information by
examining voxels in a univariate fashion. For instance, spatially distributed sets of
voxels considered as nonsignificant by a SPM analysis of one experimental condi-
tion might still carry information about the presence or absence of that condition.
Furthermore, classic voxel-based analytic techniques are also mainly designed to
perform group-wise comparisons and would, therefore, be unsuitable to evaluate the
state of the disease of each individual.
On the other hand, multi-voxel pattern analyses (MVPA) [4–6] allow an increased
sensitivity to detect the presence of a particular mental representation. These multivariate

J. Schrouff, M.Sc. (*) • C.L.M. Phillips, Ph.D.

Cyclotron Research Center, University of Liège,
Allée du 6 Août, Sart Tilman B30, 4000 Liège, Belgium
e-mail: [email protected]; [email protected]

C. Schnakers, S. Laureys (eds.), Coma and Disorders of Consciousness, 35

DOI 10.1007/978-1-4471-2440-5_4, © Springer-Verlag London 2012
36 J. Schrouff and C.L.M. Phillips

methods, also known as brain decoding or mind reading, attempt to link a particular
cognitive, behavioral, perceptual, or medical state to specific patterns of voxels’ activity.
Application of these methods made it possible to decode the category of a seen object
[7–9] or the orientation of a stripped pattern seen by the subject [10, 11] from the brain
activation of the imaged subject. Advances in pattern-classification algorithms also
allowed the decoding of less-controlled conditions such as memory retrieval tasks [12,
13]. Advanced mathematical tools are still under development to allow the classification
of more complicated experimental data sets, such as examining the content of mind wan-
dering or detecting the state of consciousness of a patient showing no response to a
command.

Methodology

Multivariate pattern analysis derives from the fields of pattern recognition and
machine learning, which are concerned with the automatic discovery of regularities
in data. Those regularities then serve as the basis for the classification of new data.
A classical example of pattern recognition is the automatic classification of hand-
written digits (Fig. 4.1): each digit is represented by a grey scale image of 28 × 28
pixels and the goal is to build an algorithm capable of classifying each image into
the correct category (i.e., 0, 1,…, 9). We, therefore, need to build an “algorithmic
machine” that will take images as inputs and produce their corresponding digit as
outputs. Due to the large variability in handwritings, this operation is not trivial and
the use of machine learning is necessary. This means that the computer has to learn
which pattern in the images corresponds to which digit. This learning is achieved by
providing a “learning set,” which is a set comprising both images (inputs) and cor-
responding digit (outputs). This is called “supervised learning.” The machine can
then build the required function using this learning set and finally assign outputs to
new inputs. Similarly, instead of classifying data into discrete categories or outputs,
a machine can be taught to regress out a continuous scalar from a series of inputs –
after learning, the trained function can then predict the (continuous) outcome from

Fig. 4.1 Examples of

handwritten digits. Each digit
is discretized as a 28 × 28
pixel grey scale image.
Practically speaking, more
than those five examples of
handwritten digits would be
necessary to build an efficient
classifier
4 Multivariate Pattern Recognition Analysis: Brain Decoding 37

Data Set:
objects label variables
var 1 var 2 var 3 var m
1 1
Learning set

2 −1
Trained
3 −1
classifer

i 1

i+1 1 1
training set

i+2 1 −1

n −1 −1
True Predicated
label Accuracy assessment label

Fig. 4.2 Graphical representation of the data set, split in training and test sets, and of the cross-
validation procedure. Each of the n objects or data points is represented by a vector of m variables
and its label. The training set, variables and label, is used to train the classifier. The trained classifier
is then applied on the test set data and the predicted labels are compared to the true labels to assess
the classifier’s accuracy

a new input. For the rest of this chapter, we only focus on the discrete classification
problem, typically into two categories.
The learning set is generally in the form of a matrix (Fig. 4.2): each “data point”
is represented by a vector, which is the collection of variables to feed in the machine,
and a label, the output of the function. Usually, the data set comprises n observa-
tions or objects to classify, based on the values of m variables. The ratio between the
number of objects and the number of variables that describe it plays an important
role in the building of the function: if m is large (many variables) compared to n
(few observations), then there is a risk of “over-fitting” [14]. If this dimensionality
issue, also known as the “curse of dimensionality” [15], is not accounted for prop-
erly, the optimized machine can achieve perfect classification on the training data
but will classify poorly any new data point: the resulting classifier does not general-
ize to similar but slightly different data points.
To validate and assess the generalizability of the trained classifier, the data set is
usually divided into two parts: the “learning set,” on which the classifier is trained,
and the “testing set,” used to compute the accuracy of the classifier. Depending on
the size of the data set available, this training-then-testing procedure can be repeated
multiple times with different splits of the data set. The final accuracy of the model
is computed as the mean of the model accuracy at each division. This is referred to
as “cross-validation procedure.”
The accuracy of the model is generally computed in terms of correct or
misclassifications for each category (Table 4.1; Eq. 4.1). For example, with the
38 J. Schrouff and C.L.M. Phillips

Table 4.1 Different types of classification errors for the case of two classes
Accuracy = TP + TN / (TP + FP + FN + TN ) (4.1)
Sensitivity = TP / (TP + FN ) (4.2)
Specificity = TN / (TN + FP ) (4.3)
Class 1 (true label: −1) Class 2 (true label: +1)
Class 1 (predicted label: −1) True negative (TN) False negative (FN)
Class 2 (predicted label: +1) False positive (FP) True positive (TP)
The ‘−1’ and ‘+1’ labels correspond to Class 1 and Class 2, respectively, e.g., the absence (−1) or
presence (+1) of a disease for a test subject

handwritten digits, 80% classification accuracy for 1’s means that four times out of
five the trained machine correctly recognizes the handwritten digit, and is incorrect
in one case out of five, i.e., the digit “1” is not recognized or another digit is recog-
nized as “1”. More specific parameters can also be derived, such as the sensitivity
and specificity of the model (Eqs. 4.2 and 4.3), which allows more insight into the
characteristics of the classification algorithm. Still, with the handwritten digits to
classify, for example, a sensibility of 90% and specificity of 60% would mean that:
if “1” is actually handwritten, then it is correctly recognized 9 times out of 10
(1 “false negative”) but for any 10 other handwritten digits except “1”, 4 of them
are erroneously classified as “1” (4 “false positive”).
This can be very important, e.g., in the case of pathology classification, it is
crucial to correctly detect all diseased patients and limit to a minimum the number
of false negatives, i.e., maximize the sensitivity even if it reduces a bit the
specificity.
There are many ways of building the input–output function, depending on how
the data are mathematically modeled. Among the various existing techniques, com-
mon ones are the kernel-based approaches [16] such as support vector machines
(SVM) [17], relevant vector machine (RVM) [18], and Gaussian processes (GP)
[19] as well as other approaches like linear discriminant analysis [15] or Gaussian
naïve bayes [2, 20].
When it comes to applying data mining tools to brain imaging data, SVM is one
of the most widespread methods. This technique is based on a simple and logical
idea: when discriminating between objects from different categories, the larger the
distance between objects from different categories, the better the classification.
SVM is mostly a binary classifier, i.e., it discriminates between only two categories
of patterns, and thus aims at finding the best hyper-plane separating the data of the
two categories, i.e., the margin between data points from each category and the
hyper-plane is maximized (as illustrated in Fig. 4.3 for the case of 2D data points).
In practice, the data points located on the margin are the only ones defining the
hyper-plane and are called “support vectors,” hence, the name “support vector
machine.” SVM (like RVM) is a sparse technique relying on a form of “automatic
relevance determination” [21], i.e., the automatic selection of relevant or represen-
tative data points among the whole data set.
4 Multivariate Pattern Recognition Analysis: Brain Decoding 39

× ×
× ×
× ×
× ×

× ×
Maximum margin
Separting
hyperplanes

Fig. 4.3 “Support vector machine” principle: The objects considered have two variables, their
value being represented by both axes, and come from two categories (× and ο here). The object
families are separated by a hyper-plane. Among all the possible hyper-planes (three are shown in
the left image), the one that maximizes the margin (right image) is picked up by SVM. The data
points on the margin are called “support vectors”

Applications

Data mining tools can be applied to a large variety of data sets, from classification
of handwritten digits to classification of brain activity induced by visualized images.
The simple form of the required inputs, a vector of variables, leads to this large
spectrum of applicability: all objects that can be characterized by a set of variables
stacked in a vector can be classified. For example, an fMRI data set consists of a
series of 3D images. Each volume can, therefore, be considered as an observation
and can be vectorized according to its voxels values. Volumes corresponding to dif-
ferent mental conditions, e.g., watching images of faces versus those of buildings,
can then be classified. In a similar way, grey matter density extracted from structural
MR images of patients and controls subjects can be classified [22]. Positron emis-
sion tomography (PET) images [23] and electro- or magnetoencephalographic
(EEG or MEG) data [24] have also been classified using this perspective.
Multivariate decoding of neuroimaging data can be used to achieve two different
objectives: firstly and obviously predict the perceptual, cognitive, or medical state of
one or many subjects but also, secondly, reveal the pattern of voxels leading to the
discrimination of these states. With a linear SVM (and other linear kernel machines),
these two goals can be reached simultaneously: the estimated weight associated to each
voxel reveals the patterns of voxels considered as important by the model to perform
the classification. Or else, the trained machine can be treated as a “black box” that
predicts the category of any new data fed in. This last application can be viewed as a
diagnostic tool in the case of a disease-versus-healthy classification (or any variation).

Diagnostic Tool

The application of multivariate analysis as a diagnostic tool could become particu-

larly useful, especially, when the concerned disease is difficult to diagnose using
classical clinical exams. Alzheimer’s disease (AD), one of the commonest causes of
40 J. Schrouff and C.L.M. Phillips

dementia, is a good example: a definitive diagnosis of AD can only be obtained

using postmortem histopathological analysis. Currently, AD is diagnosed using
clinical exams, neuropsychological testing, and manual measurements on brain
images (MRI or PET), leading to time-consuming criteria and accuracies of the
diagnosis around 80%, at best [25]. AD is, therefore, often misdiagnosed; although
an early treatment would be more effective. Multivariate analyses performed on
structural MR images of patients and controls allowed the construction of SVM,
able to automatically distinguish between healthy subjects and AD patients with
accuracies between 86% [22] and 96% [26], depending on the sample size and
information used. Moreover, in [26], data from different scanners were used, sug-
gesting that a trained classifier could be applied across centers.
These studies relied on SVM and, therefore, led to binary 0/1 classifications: the
input image is classified in one category or the other without further information
about the classification reliability. Nevertheless, when using a multivariate tech-
nique as a diagnostic tool, knowing the probability associated with the prediction
would be very valuable. Consider two new images, from two patients A and B,
classified as “healthy,” but the one from Patient A has a probability of 99% of being
so and the one from B only 51%. In both cases, under a binary classification scheme,
the corresponding subjects could be considered as “healthy,” but the classification
probability of the B image is very close to the “50% chance threshold,” which might
suggest further exams are needed to confirm or infirm the diagnostic of Patient B.
“Relevance Vector Machine” (RVM), akin to SVM, provides such probabilistic
prediction by returning the posterior probability of being in one category versus the
other – thanks to its Bayesian formulation. RVM was recently applied on
fluorodeoxyglucose PET (FDG-PET) data, i.e., images of cerebral glucose metabo-
lism of patients (vegetative state: VS, and locked-in syndrome: LIS) and healthy
controls [23].
First, an RVM was trained to discriminate VS patients (n = 13) from healthy sub-
jects (n = 37) using their FDG-PET images. With these well-defined and separate
categories, i.e., unconscious versus conscious subjects, cross-validation of this
“consciousness classifier” showed 100% accuracy. Then, the trained RVM was
applied on the FDG-PET images of eight LIS patients: the probabilities returned
ranged between 61% and 100% to be in the “conscious” category. This suggests that
LIS patients could be automatically and correctly classified as conscious, contrary
to VS patients, based only on the patterns of their cerebral metabolism.

Relevant Patterns of Voxels

In the case of a linear kernel classifier, the relevance of each voxel can be estimated
as a weighted linear combination of the same voxel values of the images used for
training. This “voxel relevance” is not a statistical value, per se, but simply reflects
how much any voxel contributes to the classification of the input image into one
output category or the other: a large (respectively small) value (in absolute value)
4 Multivariate Pattern Recognition Analysis: Brain Decoding 41

0.02

0.01

−0.01

−0.02

Fig. 4.4 Distribution over the brain volume of the voxel relevance for a “ relevance vector
machine” trained to discriminate between FDG-PET images of VS patients and healthy subjects.
A positive value (yellow–red) indicates that relatively large metabolic activity in those voxels will
drive the classification towards the “healthy subject” category. Conversely, negative values (blue–
purple) pushes towards the “VS patient” category. The voxels with little relevance (green) hardly
contribute to the classification of data (Adapted from Phillips et al. [23])

indicates that this voxel has relatively much (respectively little) influence on the
classification. Since there is a value per voxel, it is possible to present a “relevance
map” as an image in brain space (see Fig. 4.4 for the relevance maps of the control
versus VS classification) [23]. Such a map can bring insight on the location of the
discriminating areas and, therefore, help neuroscientists build more efficient criteria
of diagnosis but also orient them during the elaboration of new hypotheses concern-
ing the origins or evolution of a certain disease or disorder.

Perspectives

In view of the recent advances in multivariate pattern analysis, these techniques

certainly will become more common to study consciousness. Their application as a
diagnostic tool to differentiate patients presents obvious advantages such as objec-
tiveness, automation, and the fact that a posterior probability can be provided with
the final prediction, compared to current time-consuming and subjective criteria.
Other applications could also be envisaged, e.g., classification techniques could
be applied to “response to command” experiences in fMRI [27], leading to repro-
ducible user-independent and possibly more accurate results than the current “gen-
eral linear model” analysis used. Furthermore, multivariate analysis could be applied
online as the data are acquired, i.e., during the recording – the model is updated in
real-time with each new image. Online processing of “response to command” fMRI
experiments offers a new communication channel relying solely on brain activation
[28]. Such “brain–computer interface” schemes could certainly benefit, in term of
accuracy and speed, from more advanced “brain decoding” tools.
Finally, “brain reading” could, maybe one day, offer a partial view on the mental
content of patients in altered states of consciousness. First, a model would be trained
42 J. Schrouff and C.L.M. Phillips

with data acquired from healthy subjects thinking about different semantic categories
of pictures and words, e.g., faces, buildings, animals or emotions [29]. Then, this
model would be applied on data from minimally conscious state or VS patients.

References

1. Ogawa S, Lee TM, Kay AR, Tank DW. Brain magnetic resonance imaging with contrast
dependent on bloodoxygenation. Proc Nat Acad Sci USA. 1990;87:9868–72.
2. Friston KJ, Ashburner JT, Kiebel SJ, et al. Statistical parametric mapping: the analysis of
functional brain images. London/Burlington, San Diego: Elsevier/Academic Press; 2007.
3. Holmes A, Poline JB, Friston KJ. Characterizing brain images with the general linear model.
In: Human brain function. San Diego: Academic Press; 1997. p. 59–84.
4. Norman KA, Polyn SM, Detre GJ, Haxby JV. Beyond mind-reading: multi-voxel pattern anal-
ysis of fMRI data. Trends Cogn Sci. 2008;10:424–30.
5. Friston K, Chu C, Mourão-Miranda J, et al. Bayesian decoding of brain images. Neuroimage.
2008;39:181–205.
6. Haynes JD, Rees G. Decoding mental states from brain activity in humans. Nat Rev Neurosci.
2006;7:523–34.
7. Spiridon M, Kanwisher N. How distributed is visual category information in human occipito-
temporal cortex? An fMRI study. Neuron. 2002;35:1157–65.
8. Cox DD, Savoy RL. Functional magnetic resonance imaging (fMRI) ‘brain reading’: detecting
and classifying distributed patterns of fMRI activity in human visual cortex. Neuroimage.
2003;19:261–70.
9. Shinkareva S, Mason RA, Malave VL, et al. Using fMRI brain activation to identify cognitive
states associated with perception of tools and dwellings. PLoS One. 2008;3(3):1394.
10. Kamitani Y, Tong F. Decoding the visual and subjective contents of the human brain. Nat
Neurosci. 2005;8:679–85.
11. Haynes J, Rees G. Predicting the orientation of invisible stimuli from activity in human pri-
mary visual cortex. Nat Neurosci. 2005;8:686–91.
12. Polyn S, Natu V, Cohen J, Norman K. Category-specific cortical activity precedes retrieval
during memory search. Science. 2005;310:1963–6.
13. Chadwick M, Hassabis D, Weiskopf N, Maguire E. Decoding individual episodic memory
traces in the human hippocampus. Curr Biol. 2010;20:1–4.
14. Pitt MA, Myung IJ. When a good fit can be bad. Trends Cogn Sci. 2002;6:421–5.
15. Bishop CM. Pattern recognition and machine learning. Springer; 2006. http://www.springer.
com/computer/image+processing/book/978-0-387-31073-2
16. Müller KR, Mika S, Rätsch G, et al. An introduction to kernel-based learning algorithms.
IEEE Trans Neural Netw. 2001;12:181–202.
17. Burges C. A tutorial on support vector machines for pattern recognition. Data Mining Knowl
Discov. 1998;2:121–67.
18. Tipping ME. Sparse Bayesian learning and the relevance vector machine. J Mach Learn Res.
2001;1:211–44.
19. Rasmussen CE, Williams CKI. Gaussian processes for machine learning. Cambridge: The
MIT Press; 2006.
20. Mitchell TM, Hutchinson R, Niculescu RS, et al. Learning to decode cognitive states from
brain images. Mach Learn. 2004;57:145–75.
21. Neal RM. Bayesian learning for neural networks. New York: Springer; 1996.
22. Vemuri P, Gunter JL, Senjem ML, et al. Alzheimer’s disease diagnosis in individual subjects
using structural MR images: validation studies. Neuroimage. 2008;39:1186–97.
4 Multivariate Pattern Recognition Analysis: Brain Decoding 43

23. Phillips C, Bruno MA, Maquet P, et al. “Relevance vector machine” consciousness classifier
applied to cerebral metabolism of vegetative and locked-in patients. Neuroimage. 15 May
2011;56(2):797–808.
24. Chan AM, Halgren E, Marinkovic K, Cash SS. Decoding word and category-specific spa-
tiotemporal representations from MEG and EEG. Neuroimage. 2011;54:3028–39.
25. Knopman DS, DeKosky ST, Cummings JL, et al. Practice parameter: diagnosis of dementia
(an evidence-based review). report of the quality standards Subcommittee of the American
Academy of Neurology. Neurology. 2001;56:1143–53.
26. Kloppel S, Stonnington CM, Chu C, et al. Automatic classification of MR scans in Alzheimer’s
disease. Brain. 2008;131:681–9.
27. Monti MM, Vanhaudenhuyse A, Coleman MR, et al. Willful modulation of brain activity in
disorders of consciousness. N Engl J Med. 2010;362:579–89.
28. Sorger B, Dahmen B, Reithler J, et al. Another kind of ‘BOLD response’: answering multiple-
choice questions via online decoded single-trial brain signals. In: Progress in brain research.
Vol 177. Elsevier; 2009. p. 275–92. http://www.elsevier.com/wps/find/bookdescription.cws_
home/720425/description#description
29. Mitchell TM, Shinkareva SV, Carlson A, et al. Predicting human brain activity associated with
the meanings of nouns. Science. 2008;30:1191–5.
Chapter 5
Spectroscopy and Diffusion Tensor Imaging
in Disorders of Consciousness

Jean-Flory Luaba Tshibanda, Athena Demertzi, and Andrea Soddu

Introduction

Medical advances in (neuro) intensive care currently allow patients suffering from
severe brain lesions, such as traumatic brain injury or anoxia, to survive. These patients
can evolve toward brain death, a vegetative state, a minimally conscious state, or toward
recovery of normal consciousness level [1]. In most cases, this evolution is based on
behavioral testing, which is explained in Chapter 1. This examination implies a part of
subjectivity and does not always lead to an accurate diagnosis. It has been shown that
about 40% of patients considered as being in a vegetative state were, in fact, in a mini-
mally conscious state when assessed more carefully [2–4]. This high rate of misdiag-
nosis is unacceptable for all medical and paramedical professionals in-charge and
(might influence) for the efficacy of the patient’s management, the accuracy of the
information provided to the family, and the balanced management of financial resources.
On the other hand, a consensus on end-of-life decisions has to be considered as regards
the debates currently happening within the medical field.
Magnetic resonance imaging (MRI) may help to improve the diagnosis of patients
with disorders of consciousness and provide prognosis of the recovery of patients’
brain functioning. This technique allows not only a better assessment of structural
lesions but also a better understanding of disorders of consciousness.

J.-F.L. Tshibanda, M.D. (*)

Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du 6 Août, Sart Tilman B30, 4000 Liège, Belgium
Department of Radiology, University Hospital of Liège,
Liège, Belgium
e-mail: [email protected]
A. Demertzi, M.Sc. • A. Soddu, Ph.D.
Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du 6 Août, Sart Tilman B30, 4000 Liège, Belgium

C. Schnakers, S. Laureys (eds.), Coma and Disorders of Consciousness, 45

DOI 10.1007/978-1-4471-2440-5_5, © Springer-Verlag London 2012
46 J.-F.L. Tshibanda et al.

Magnetic Resonance Imaging

Technological advances of these last 10 years in the field of magnetic resonance

imaging have radically changed our understanding of coma and of post-comatose
states of consciousness.
MRI techniques allow the realization of structural, metabolic, and functional
measures of brain structures involved in arousal and awareness. These refer to the
reticular activating system located in the posterior upper part of the pons, the cere-
bral peduncles, the thalamus, the hypothalamus, the inferior part of the frontal lobes,
and the precuneus [5–9]. A good knowledge of brain lesions caused by traumatic
brain injury and stroke, which are the main etiologies responsible for impairments
of consciousness, will allow clinicians to choose adequately the MRI sequences.
Traumatic brain injury causes brain contusions and diffuse axonal lesions.
Contusions (often located in temporal and frontal areas) are the consequence of the
impact of the parenchyma against the skull; while diffuse axonal lesions are caused
by rotational acceleration–deceleration events. These axonal lesions are often
located between structures of different density, such as the juxtacortical or periven-
tricular white matter, the corpus callosum, the basal ganglia, and the brainstem
(midbrain and pons). These lesions can be ischemic or hemorrhagic, which can
influence the prognosis of posttraumatic cerebral functioning. Anoxic/hypoxic
lesions resulting from a stroke are often accompanied by a cytotoxic edema. These
lesions can evolve toward recovery or toward tissue necrosis and hemorrhage.
It is important to choose the right moment to scan a patient recovering from coma
in order to benefit from an MRI examination. An MRI scan performed in a strict
dorsal decubitus can lead to intracranial hypertension in acute patients suffering a
severe brain injury. Additionally, this technique can allow the detection of reversible
lesions such as edema; but can miss secondary brain lesions (linked to the patient’s
global deterioration or linked to the presence of intracranial hypertension), which are
likely to aggravate the functional prognosis of the brain [10–12]. On the other hand,
a late MRI examination allows the detection of sequels and atrophies [13]; but has no
impact on the patient’s treatment or on the prognostic value. The ideal moment for a
MRI scan seems to be the third week after the brain edema disappears [14].
A three-dimensional T1 sequence is performed in order to estimate cerebral atro-
phy and is used as an anatomical reference for the areas of cortical activation observed
with functional imaging. The “fluid attenuated inversion recovery” (FLAIR) sequence
is useful to detect ischemic lesions, contusions, and diffuse axonal (not hemorrhagic)
lesions. These lesions are often invisible using tomodensitometry and can explain
disorders of consciousness in those patients where tomodensitometrical results were
normal. Diffuse axonal hemorrhagic lesions are revealed by the T2* sequence, which
shows microbleedings as well as hypointense lesions, due to the presence of hemo-
siderin (Fig. 5.1). The number of lesions detected by these two morphological T2*
sequences is correlated to patients’ long-term prognosis [15, 16]. The prognosis is
particularly negative in the case of bilateral lesions involving the brainstem [17, 18].
Although these two sequences are very sensitive in the detection of lesions, they do
not explain disorders of consciousness presented by patients who do not show any
5 Spectroscopy and Diffusion Tensor Imaging in Disorders of Consciousness 47

a b

Fig. 5.1 Traumatic patient in a minimally conscious state. The axial T2* sequence images (a)
show hemorrhagic diffuse axonal lesions invisible on the axial FLAIR sequence images (b)

abnormality of the brainstem. Therefore, they are not sufficient to establish a prog-
nosis. There is, hence, a growing interest for using innovative techniques, such as
spectroscopy, tensor imaging, or functional imaging.

Magnetic Resonance Spectroscopy

In contrast to conventional MRI, which provides anatomical information based on

the tissue water signal, magnetic resonance spectroscopy (MRS) provides chemical
information based on the signal of tissue metabolites. Atoms studied in vivo with
MRS are proton (1H), phosphorus (31P), carbon (13C), sodium (23NA), potassium
(39K), and fluorine (19F).
Spectroscopy of proton conduction is currently the most used technique in the
exploration of post-comatose disorders of consciousness. Conducted with a 1.5 or
3 T MRI, this noninvasive technique allows the study of the brain function by the
quantification of several metabolites. The detection of these metabolites constitutes a
real challenge, considering their very weak concentration in comparison to peri- and
intracerebral water. This problem is solved by using techniques that allow the aboli-
tion of the tissue water signal. Besides, the heterogeneity of the parenchyma thick-
ness requires focal measurements, which lead to use voxels of 1 or 2 cm³ and to a low
signal-to-noise ratio. The MRS performed in vivo with abolition of the tissue water
signal and with a 135 or 270 ms echo-time focuses on four metabolites: [1] Choline
(Cho), appearing at 3.2 ppm, is a marker of membrane synthesis and catabolism and
48 J.-F.L. Tshibanda et al.

is more abundant in white matter than in gray matter. Its rate increases in case of cell
proliferation, in inflammatory and demyelinating processes; [2] Creatine (Cr),
appearing at 3.0 ppm, is a relatively stable marker of aerobic energetic metabolism;
[3] The N-acetyl-aspartate (NAA), appearing at 2.0 ppm, is present in neural cyto-
plasm and in axons. The NAA is equally distributed in gray and white matter [19].
Several studies suggest that this metabolite is a cerebral osmolytes likely to introduce
reversible variations [20, 21]. It is reduced in all diseases that are accompanied by
neural or axonal loss, such as neurodegenerative pathologies, stroke, traumatic brain
injury, and glial tumors; and, hence, is a marker of neural and axonal integrity [22].
The possibility to quantify neural damages in vivo is certainly one of the most impor-
tant applications of the MRS; and [4] lactate, appearing at 1.3 ppm, is a marker of
anaerobic metabolism and is usually difficult to detect. Its rate increases in case of
hypoxia, ischemia, or severe traumatic injury.
The exploration of post-comatose patients includes two MRS acquisitions. A
sequence of mono-voxel proton MRS (single voxel spectroscopy or SVS) is per-
formed at the level of the posterior upper section of the pons (Fig. 5.2) as it contains
the reticular activating system known for its role in arousal. Besides, the pons is
often damaged by diffuse axonal lesions invisible on FLAIR and T2* sequences and
can detect secondary lesions due to a temporal herniation. A multi-voxel proton
MRS (chemical shift imaging or CSI) is performed at the level of the basal ganglia
and allows assessing metabolites in several anatomical structures involved in arousal
and awareness processes, such as thalami, lenticular nucleus, mesiofrontal areas,
and occipital periventricular white matter (Fig. 5.3). The difficulty in conducting an
absolute quantification of these metabolites leads to use values based on the ratio of
a metabolite divided by the sum of metabolites or by another metabolite considered
to be stable. Creatine plays the role of reference to calculate the NAA/Cr or Cho/Cr
ratios. A recent review has underlined several studies showing the efficacy of the
MRS in predicting recovery of brain-injured patients [23], confirming the prelimi-
nary study of Choe et al. [24], who suggested that the decrease of NAA/Cr ratio
could underline axonal loss in the case of traumatic brain injury and could constitute
a metabolic criterion for recovery in these patients. Usually, NAA/Cr ratio is supe-
rior or equal to 1.50 and NAA/NAA + Cho + Cr ratio is equal to 0.40.
In case of traumatic brain injury, a reduction of the NAA/Cr ratio is a reliable
indicator of negative prognosis [25]. This ratio decreases rapidly few hours after the
onset and reaches its lowest value within 48 h. It remains stable for a month after-
ward allowing the realization of MRI scans in the third week postinjury when the
cerebral edema has decreased. It, nevertheless, strongly fluctuates 6 weeks to 1 year
postinjury. Within this period of time, some authors observed a constant reduction
of NAA/Cr ratio [26, 27] while others observed a biphasic evolution (i.e., increase/
decrease phase) of this one [28]. The variability of the NAA/Cr ratio is a potential
limit of this technology. Moreover, the use of creatine as a reliable reference to cal-
culate the values of NAA and Cho stresses the problem of stability of creatine con-
centration. Indeed, it is not established that creatine remains stable in the case of
traumatic brain injury. Some studies showed that its concentration can increase in
hypometabolic states and decrease in hypermetabolic states [29, 30].
5 Spectroscopy and Diffusion Tensor Imaging in Disorders of Consciousness 49

svs_se_135_avec_WS
19/07/2010
I : Integral
15:17:48
SE:30 IM:1

Cho
I : 6.21

0.6

Cr
I: 4.60 NAA
I : 4.37
0.4

Cr2
I: 2.54
0.2

0.0

ppm
4 3 2 1
TR:
TE: EC:
thk/ sp
W 4095 : L 2050 C:HEA;HEP;NE2_c_5

Fig. 5.2 Vegetative patient. The sequence of single voxel spectroscopy (SVS) performed in the
pons shows a decrease of the NAA/Cr ratio

The MRS contributes to improve the establishment of functional prognosis of

severe brain-injured patients and permits to follow their clinical evolution in paral-
lel to the NAA/Cr ratio, a decrease of this ratio being indicative of poor
prognosis.

Diffusion Tensor Imaging

Diffusion tensor imaging (DTI) is an MRI technique that allows the study in vivo of
tissue microstructure, which was developed based on the study of the diffusion of
water molecules. This diffusion is free when molecules randomly move in a
50 J.-F.L. Tshibanda et al.

Fig. 5.3 Illustration of the

areas where the metabolites
are measured using chemical
shift imaging (CSI):
thalamus, lenticular nucleus,
insula and periventricular
white matter

homogeneous environment without restriction such as the cerebrospinal fluid

(Brownian motion). Restricted isotropic diffusion happens when the movement of
water molecules is restricted in all directions and restricted anisotropic diffusion
happens when this restriction is only present in some directions, creating heteroge-
neity in the speed of movements of water molecules or anisotropy. The axonal orga-
nization is at the origin of anisotropic diffusion in the white matter – the diffusion
of water molecules is preferentially directed in parallel to the axon and restricted
perpendicularly to this one. The detection of water diffusion direction around axons
allows the study of functional connectivity. It is, hence, possible to study cognitive
deficits according to the anatomical connectivity between brain areas coding for
these cognitive processes.
Diffusion tensor imaging allows measuring the fractional anisotropy (FA). Its
value is equal to zero when the diffusion is isotropic or is high when the diffusion is
anisotropic. It allows creating a map of the white matter with color codification for
the orientation of axons (Fig. 5.4). The diffusion tensor imaging is the only tech-
nique giving in vivo overview of the axonal tracts (tractography). In case of disor-
ders of consciousness, the FA is measured in the posterior section of the pons, in the
5 Spectroscopy and Diffusion Tensor Imaging in Disorders of Consciousness 51

cerebral peduncles, in the internal and external capsules, in the periventricular white
matter, and in the semioval center (Fig. 5.4). The destruction of the axons leads to a
reduction of the anisotropic diffusion and, therefore, of the FA. This one constitutes
a noninvasive marker of the integrity of axonal tracts using T2* and FLAIR
sequences [31]. The study performed by Perlbarg et al. [32] showed a reduction of
the FA measured in the cerebral peduncles, the internal capsule, and the corpus cal-
losum in the group of traumatic brain injured patients with a bad outcome as com-
pared to those with a good outcome. By studying the recovery of traumatic patients
at 1 year postonset, Tollard et al. [33] observed a reduction of FA in all measured
brain areas. They also noticed that combining spectroscopy (NAA, Cr) and diffu-
sion (FA) methods allowed the prediction of bad outcome at 1 year with a specificity
of 97% against 75% only for the spectroscopy and 85% only for the diffusion. For
the three cases, the sensitivity is, respectively, 86%, 75%, and 79%. These two stud-
ies confirm that the FA constitutes a good biomarker for detecting lesions in the
white matter.

Fig. 5.4 (a) Diffusion tensor imaging with color codification for the axonal orientation. (b)
Illustration of the areas where the fractional anisotropy (FA) is measured: pons, cerebral peduncles,
internal and external capsules, periventricular white matter and semioval center
52 J.-F.L. Tshibanda et al.

Fig. 5.4 (continued)

b

Conclusions

Spectroscopy and diffusion tensor imaging detect invisible tissue lesions using
FLAIR and T2* sequences. These techniques allow a better diagnosis and prognosis
of patients with disorders of consciousness (particularly, when they are combined)
and permit a better monitoring of the deterioration or recovery of posttraumatic
brain functioning by considering the fluctuation of the NAA/Cr and FA values in
time. To our mind, in the future, these noninvasive techniques will play a more
important role in the monitoring of post-comatose patients.

References

1. Laureys S, Boly M. The changing spectrum of coma. Nat Clin Pract Neurol. 2008;4:544–6.
2. Childs NL, Mercer WN, Childs HW. Accuracy of diagnosis of persistent vegetative state.
Neurology. 1993;43:1465–7.
3. Andrews K, Murphy L, Munday R, et al. Misdiagnosis of the vegetative state: retrospective
study in a rehabilitation unit. BMJ. 1996;313:13–6.
4. Schnakers C, Vanhaudenhuyse A, Giacino J, et al. Diagnostic accuracy of the vegetative and
minimally conscious state: clinical consensus versus standardized neurobehavioral assess-
ment. BMC Neurol. 2009;9:35.
5 Spectroscopy and Diffusion Tensor Imaging in Disorders of Consciousness 53

5. Kampfl A, Schmutzhard E, Franz G, et al. Prediction of recovery from post-traumatic vegeta-

tive state with cerebral magnetic-resonance imaging. Lancet. 1998;351:1763–7.
6. Shibata Y, Matsumura A, Meguro K, et al. Differentiation of mechanism and prognosis of
traumatic brain stem lesions detected by magnetic resonance imaging in the acute stage. Clin
Neurol Neurosurg. 2000;102:124–8.
7. Firsching R, Woischneck D, Klein S, et al. Brain stem lesions after head injury. Neurol Res.
2002;24:145–6.
8. Uzan M, Albayram S, Dashti SG, et al. Thalamic proton magnetic resonance spectroscopy in
vegetative state induced by traumatic brain injury. J Neurol Neurosurg Psychiatry.
2003;74:33–8.
9. Salazar AM, Grafman JH, Vance SC, et al. Consciousness and amnesia after penetrating head
injury: neurology and anatomy. Neurology. 1986;36:178–87.
10. Huisman TA, Schwamm LH, Schaefer PW, et al. Diffusion tensor imaging as potential biomarker
of white matter injury in diffuse axonal injury. AJNR Am J Neuroradiol. 2004;25:370–6.
11. Ross BD, Ernst T, Kreis R, et al. 1H MRS in acute traumatic brain injury. J Magn Reson
Imaging. 1998;8:829–40.
12. Marino S, Zei E, Battaglini M, et al. Acute metabolic brain changes following traumatic brain
injury and their relevance to clinical severity and outcome. J Neurol Neurosurg Psychiatry.
2007;78:501–7.
13. Trivedi MA, Ward MA, Hess TM, et al. Longitudinal changes in global brain volume between
79 and 409 days after traumatic brain injury: relationship with duration of coma. J Neurotrauma.
2007;24:766–71.
14. Weiss N, Galanaud D, Carpentier A, et al. Clinical review: prognostic value of magnetic reso-
nance imaging in acute brain injury and coma. Crit Care. 2007;11:230.
15. Carpentier A, Galanaud D, Puybasset L, et al. Early morphologic and spectroscopic magnetic
resonance in severe traumatic brain injuries can detect “invisible brain stem damage” and
predict “vegetative states”. J Neurotrauma. 2006;23:674–85.
16. Yanagawa Y, Tsushima Y, Tokumaru A, et al. A quantitative analysis of head injury using T2*-
weighted gradient-echo imaging. J Trauma. 2000;49:272–7.
17. Wedekind C, Fischbach R, Pakos P, et al. Comparative use of magnetic resonance imaging and
electrophysiologic investigation for the prognosis of head injury. J Trauma. 1999;47:44–9.
18. Firsching R, Woischneck D, Diedrich M, et al. Early magnetic resonance imaging of brainstem
lesions after severe head injury. J Neurosurg. 1998;89:707–12.
19. Danielsen ER, Ross B. Magnetic resonance spectroscopy diagnosis of neurological diseases.
1st ed. New York: Marcel Dekker, Inc; 1999.
20. Baslow MH, Suckow RF, Gaynor K, et al. Brain damage results in down-regulation of
N-acetylaspartate as a neuronal osmolyte. Neuromolecular Med. 2003;3:95–104.
21. Moffett JR, Ross B, Arun P, et al. N-acetylaspartate in the CNS: from neurodiagnostics to
neurobiology. Prog Neurobiol. 2007;81:89–131.
22. Sullivan EV, Adalsteinsson E, Spielman DM, et al. N-acetylaspartate–a marker of neuronal
integrity. Ann Neurol. 2001;50:823. author reply 4–5.
23. Tshibanda L, Vanhaudenhuyse A, Galanaud D, et al. Magnetic resonance spectroscopy and
diffusion tensor imaging in coma survivors: promises and pitfalls. Prog Brain Res.
2009;177:215–29.
24. Choe BY, Suh TS, Choi KH, et al. Neuronal dysfunction in patients with closed head injury
evaluated by in vivo 1 H magnetic resonance spectroscopy. Invest Radiol. 1995;30:502–6.
25. Cecil KM, Hills EC, Sandel ME, et al. Proton magnetic resonance spectroscopy for detection
of axonal injury in the splenium of the corpus callosum of brain-injured patients. J Neurosurg.
1998;88:795–801.
26. Garnett MR, Blamire AM, Corkill RG, et al. Early proton magnetic resonance spectroscopy in
normal-appearing brain correlates with outcome in patients following traumatic brain injury.
Brain. 2000;123(Pt 10):2046–54.
27. Holshouser BA, Tong KA, Ashwal S, et al. Prospective longitudinal proton magnetic reso-
nance spectroscopic imaging in adult traumatic brain injury. J Magn Reson Imaging.
2006;24:33–40.
54 J.-F.L. Tshibanda et al.

28. Brooks WM, Stidley CA, Petropoulos H, et al. Metabolic and cognitive response to human
traumatic brain injury: a quantitative proton magnetic resonance study. J Neurotrauma.
2000;17:629–40.
29. Castillo M, Kwock L, Mukherji SK. Clinical applications of proton MR spectroscopy. AJNR
Am J Neuroradiol. 1996;17:1–15.
30. Wood SJ, Berger G, Velakoulis D, et al. Proton magnetic resonance spectroscopy in first epi-
sode psychosis and ultra high-risk individuals. Schizophr Bull. 2003;29:831–43.
31. Huisman T, Sorensen A, Hergan K, et al. Diffusion-weighted imaging for the evaluation of
diffuse axonal injury in closed head injury. J Comput Assist Tomogr. 2003;27:5–11.
32. Perlbarg V, Puybasset L, Tollard E, et al. Relation between brain lesion location and clinical
outcome in patients with severe traumatic brain injury: a diffusion tensor imaging study using
voxel-based approaches. Hum Brain Mapp. 2009;30:3924–33.
33. Tollard E, Galanaud D, Perlbarg V, et al. Experience of diffusion tensor imaging and 1H spec-
troscopy for outcome prediction in severe traumatic brain injury: preliminary results. Crit Care
Med. 2009;37:1448–55.
Chapter 6
Electrophysiology and Disorders
of Consciousness

Quentin Noirhomme and Rémy Lehembre

Electroencephalography

Electroencephalography is the measure of the brain’s electrical activity using

electrodes placed on the surface of the skull. It directly reflects neuronal activity
with a high temporal resolution. However, the spatial resolution is poor for two
main reasons: (1) it is limited by interelectrode distance and (2) because of vol-
ume conduction, each sensor measures a sum of different brain sources, hence
sensors have a correlated signal. The number of electrodes used depends on the
application. For instance, when monitoring the level of anesthesia, only two elec-
trodes are needed to obtain an electroencephalographic trace, while in clinical
environments at least ten are used. In research, modern electrode caps have up to
256 electrodes. Electrode positioning on the cranial surface follows international
nomenclatures (10–20 system for up to 19 electrodes and 10–10 system for more
than 19 electrodes [1]), which aim to cover homogeneously most of the cranial
surface. Electrodes are named according to their position on the scalp and have
the letter F for frontal, C for central, P for parietal, O for occipital, or a letter
combination such as FC for a position between frontal and central. Additionally,
these letters are followed by a number (even for electrodes on the right hemi-
sphere and odd for electrodes on the left hemisphere) or by the letter Z (for elec-
trodes on the midline).
The measured signal results from a potential difference between two electrodes.
It is, therefore, not possible to use only one electrode. There are two categories of
montages: bipolar montages in which the electrodes are paired two by two; and

Q. Noirhomme, Ph.D. (*) • R. Lehembre, M.Sc.

Coma Science Group, Cyclotron Research Centre, University of Liège,
Allée du 6 Août 8, Sart Tilman B30, 4000 Liège, Belgium
e-mail: [email protected]; [email protected]

C. Schnakers, S. Laureys (eds.), Coma and Disorders of Consciousness, 55

DOI 10.1007/978-1-4471-2440-5_6, © Springer-Verlag London 2012
56 Q. Noirhomme and R. Lehembre

referential montages where all the electrodes are coupled to a single electrode
called the reference. In a bipolar montage, it can be considered that the recorded
signal stems from an imaginary position located between the two electrodes. In a
referential montage, the reference should not be located in a region where a signal
of interest has to be recorded. Common positions for the reference are the ear
lobes, the mastoids, possibly coupled, the nose, or a position on the midline. The
choice of the reference electrode has an influence on the shape of the recorded
signal, notably for evoked potentials. Bipolar montages are less sensitive to arti-
facts but will not detect events that are common to two coupled electrodes. A ref-
erential montage does not have this drawback but is, however, more sensitive to
artifacts [2].
The electroencephalogram (EEG) does not only detect electrical fields gener-
ated by cerebral activity but also fields generated by muscular activity such as eye
or eyelids movements, or fields generated by electrical apparatus. Patients in
altered states of consciousness are often surrounded by many different electronic
equipments withstanding their vital functions. They have little control of their
movements and can be spastic. Furthermore, they do not control their level of
sudation, which can potentially be the cause of artifacts, which should be mini-
mized during the recording. It is good practice to simultaneously record respira-
tion, heartbeats, and muscular activity in order to better track artifacts and
eventually remove them from the signal. It is also important to have full knowl-
edge of all drugs prescribed to the patient as some can have a sedative effect,
which can result in a slower EEG, or others such as benzodiazepines can add
additional fast frequencies to the signal. Some artifacts can be eliminated by data
filtering. A notch filter removes the 50 Hz line noise (60 Hz in the United States
and other regions in the world). The EEG spectrum covers frequencies ranging
from less than 1 Hz to several hundreds of Hz. The use of filters should, therefore,
depend on the frequency bands of interest. Too much filtering of low frequencies
might hide slow waves activity while filtering high frequencies might hide spin-
dles and spike waves. Heavy filtering results in a clean EEG trace but might
remove some of the signal of interest [2]. Taking these considerations into account,
the EEG signal is often observed after filtering between 1 and 30 Hz, notably in
sleep studies or for evoked potentials. These limits can then be adjusted in order
to include more frequencies.
There are three main approaches in the study of EEG, which are presented in
separate sections. The first approach consists in the visual inspection of the electro-
encephalographic trace in order to unveil background EEG activity and characteris-
tic events. It gives global information on cerebral activity. The second approach,
based on evoked potentials, measures the brain’s response to an external stimulation
such as a sound or a flash. It is useful in the study of auditory, visual, or sensory–
motor areas. Finally, the third approach is quantitative and aims to extract character-
istic EEG features of cerebral activity. Such features are, e.g., the power in different
frequency bands, an index of wakefulness, or a measure of connectivity between
cerebral areas. The goal of this approach is to provide an automated EEG analysis,
but it is still under development.
6 Electrophysiology and Disorders of Consciousness 57

EEG Interpretation

The visual interpretation of an EEG trace gives information on the global cerebral
activity of a patient. The EEG can help to establish the origin and the severity of the
injury (e.g., in the case of a focal lesion or a diffuse dysfunction) and differentiate
states that are apparently similar to coma such as an epileptic absence, a psychogenic
coma, a noncooperative patient, or a locked-in syndrome. Combined with the etiol-
ogy, the EEG can give an indication on the patient’s prognosis. Finally, the EEG can
track the patient evolution and the effect of drugs such as anti-epileptics [3–5].
The EEG of a control subject is mainly composed of an alpha rhythm (8–12 Hz)
with a posterior and symmetric distribution. Its amplitude increases when the sub-
ject is resting and eyes closed. An additional beta rhythm (13–30 Hz) of smaller
amplitude can be irregularly observed. These rhythms are underlined by slower
rhythms which are more visible when the subject becomes tired. In tired subjects, a
theta rhythm (4–7 Hz) becomes the dominant rhythm. Once asleep, subjects have a
predominance of delta rhythms (1–3 Hz). In patients with disorders of conscious-
ness, a slower EEG can be observed, which correlates with the level of conscious-
ness, along with the apparition of characteristic patterns.
To allow a good interpretation, the EEG recording of a patient must last long
enough without stimulations to correctly assess the background activity of the EEG
and to detect potential variations [3]. It is important to test the reactivity of the EEG to
eye opening/closing and external stimulations. A reactive EEG reflects a lighter coma
and is associated with a better prognosis [3, 6]. Auditory or nociceptive stimulations
can be used. Reactivity can be characterized by a drop in the background activity or
by an increase in amplitude. Synek [7] suggested a scale to assess the severity of the
coma. This scale has then been adapted by Young et al. in order to improve its repro-
ducibility [6]. Young’s scale is presented in Table 6.1 and gives information on the
level of the coma. Following a lesion of the brainstem, some comatose patients can
display an EEG that is comparable to a normal wake EEG with a predominant alpha
rhythm – the alpha coma. In the case of an alpha coma, EEG changes can be observed
in response to sensory stimuli and to eye openings [3, 8]. These patients must be dif-
ferentiated from patients suffering a locked-in syndrome or patients in a psychogenic
coma. Indeed, in both the latter cases, the EEG can appear normal [3].
The EEG is useful for detecting a loss of consciousness due to non convulsive
epileptic seizures or epileptic spikes in patients that are unconscious for other rea-
sons than epilepsy [5]. A nonconvulsive epileptic seizure does not present the usual
signs of a complex partial seizure like oculomotor and masticatory muscles contrac-
tion, and the patient can appear confused, drowsy, or comatosed. The EEG will
show a continuous epileptic activity. If the patient does not react to administered
antiepileptic drugs and if the cause of the seizures cannot be treated, the prognosis
will be poor [9]. The EEG can also detect epileptic seizures due to a cerebral lesion.
In this case, the seizures can be treated.
In case of a loss of consciousness due to a focal supratentorial cerebral lesion, the
EEG is significantly abnormal. Different types of abnormalities can be observed,
58 Q. Noirhomme and R. Lehembre

Table 6.1 EEG classification introduced by Young and collaborators [6]

Category Subcategory
1. Delta/theta >50% of record (no theta coma) (a) Reactivity
(b) No reactivity
2. Triphasic waves
3. Burst-suppression (a) With epileptiform activity
(b) Without epileptiform activity
4. Alpha/theta/spindle coma (unreactive)
5. Epileptiform activity (no burst-suppression pattern) (a) Generalized
(b) Focal or multifocal
6. Suppression (a) <20 mV but >10 mV
(b) <10 mV
Guidelines:
1. Burst-suppression pattern should present generalized flattening at standard sensitivity for
more than 1 s at least every 20 s
2. Suppression: for this category, voltage criteria should be met for the entire recording; there
should be no reactivity
3. When more than 1 category is present, select the most critical one

such as polymorphic delta activity or epileptic spikes. Nevertheless, the EEG gives
little information about the location of the lesion; it can just hardly determine the
damaged hemisphere [3].
The last stage of the coma is characterized by burst suppression, when there is no
cerebral activity higher than 2 mV. An inactive EEG that lasts for more than 6 h in a
patient who is not in hypothermia suggests prosencephalon death but not necessar-
ily cerebral death as the EEG does not reflect the activity of the brainstem [5]. In
some rare occasions, patients in permanent vegetative state can have an inactive
EEG [3]. Similarly, drug intoxication can lead to an inactive EEG, but it is often
reversible.
The interpretation of an EEG recording does not allow a prognostic statement if
it is not combined with the etiology. Indeed, the characteristics features of the EEG
are not specific to one etiology. Here are some examples of poor prognosis. In case
of a cardiac arrest, a periodic generalized pattern is of poor prognosis. Following a
hypoxia or metabolic encephalopathy, the apparition of suppression periods lasting
several seconds that are not followed by a burst are of poor prognosis. A pattern
such as alpha coma or alpha-theta coma is associated with different prognosis based
on the etiology. For instance, if it is associated with a brainstem lesion, it is of poor
prognosis. For a detailed review of prognosis associated with different patterns, we
recommend the article by Brenner [3].

Evoked Potentials

Evoked potentials allow an objective evaluation of patients’ sensory and cognitive

functions. Changes due to sensory, motor, or cognitive activity are invisible when
recording the raw EEG because of low signal-to-noise ratio. To observe them, the
6 Electrophysiology and Disorders of Consciousness 59

stimulus must be repeated several times. Auditory evoked potentials are most often
elicited by sounds (clics) presented through headphones. Somato-sensory evoked
potentials are elicited by stimulation of the median nerve, near the wrist. Visual
evoked potentials elicited by flashes are less common because they do not always
trigger a response, even in healthy controls [10]. To observe an evoked potential, the
EEG must be windowed around the stimulus, each window having the same size
and starting at the same moment with respect to the stimulus. These windows
are then averaged and, therefore, activities that systematically arise at the same time
interval after the stimulus will add up while other cerebral activity not related to the
stimulus will be averaged out. This procedure reveals positive and negative varia-
tions in the EEG called “components.” These components are markers of the activ-
ity of cerebral areas related to stimulus processing.
Components obtained in response to a sensory stimulation reflect the processing
of the stimulus through time, from low level peripheral receptive structures up to
high level associative cerebral areas. The faster components, linked to the physical
properties of the stimulus are called “exogenous” and reflect the activation of neu-
rons projecting toward the primary cortex. Belated components are linked to the
psychological signification of the stimulus, experimental conditions, and the level
of awareness. They are called “endogenous” components and reflect the activity of
subcortical and cortical structures, including associative areas.
Somesthetic-evoked potentials (SEP) are obtained by transcutaneous electric
stimulation of median nerves in the wrist. These potentials reflect the conduction
of the nervous influx through the brachial plexus and its access to the primary
somatosensory cortex [11]. A bilateral absence of the N20 in a comatose patient is
highly associated with an absence of consciousness recovery (in 99–100% of
cases) [12–15]. Brainstem auditory evoked potentials (BAEP) are useful to study
the conduction of the auditory signal via the auditory nerve and the protuberance.
They appear within 10 ms. The absence of these potentials is associated with a
poor recovery in patients with severe cerebral lesions but without peripheral audi-
tory lesions [16, 17]. Nevertheless, this component has a lower predictive value
than the N20 response [13].
The absence of exogenous components is a marker of poor prognosis [18]. Most
of the patients for whom these components are bilaterally absent either die or stay
in permanent vegetative state. However, the presence of exogenous components is
not informative enough to be a marker of good prognosis. Clear exogenous compo-
nents can be observed in patients that never recover.
More advanced cerebral processes, possibly reflecting the presence of conscious-
ness, can be studied using cognitive-evoked potentials. Until now they were exclu-
sively studied with auditory tasks because comatose patients do not have eye-gaze
control. They differ from exogenous-evoked potentials in the sense that they are highly
dependent on the experimental conditions. It is, thus, important to record these poten-
tials when the patient is most vigilant and to ensure that the paradigm is optimized for
recording the best potentials while minimizing the number of repetitions to avoid a
habituation effect. There are five main components usually studied: the N100 compo-
nent in response to a stimulus, the mismatch negativity and the P3 in response to
novelty, and N400 and P600 components in response to semantic changes.
60 Q. Noirhomme and R. Lehembre

The N100 component, a negative inflection that appears 100 ms after the start of
the stimulus, indicates a response of the auditory cortex. This component is elicited
by all types of stimuli and reveals that the auditory cortex is properly functioning.
Its predictive value is, however, highly debated [19–22]. Aside BAEPs, the N100
component would yet have a lower predictive value than the N20 response as regards
consciousness recovery in comatose patients [13].
The mismatch negativity (MMN) is a negative component, which appears
between 100 and 200 ms after a change or odd sound following a series of monoto-
nous sounds. This component has low amplitude, which implies that a high number
of repetitions are necessary for a good visualization. Since the MMN does not
require the subject’s attention, it indicates an automatic response triggered by the
difference between the dissonant sound and the other preceding sounds which are
still recorded in memory. Previous data obtained with an MMN paradigm in coma-
tose patients suggest that this component beholds important predictive value inde-
pendent of the etiology. Indeed, an absence of this response was related to an absence
of recovery in 91–93% of cases [13, 23]. This component, likewise the N20 response,
has a better predictive value than AEPCTs or the N100 response [13, 24].
Kotchoubey et al. have shown that the MMN component could be present in
minimally conscious patients (34%) and in patients in a vegetative state as well
(65%) [25]. However, recently, Wijnen et al. demonstrated that for ten vegetative
patients, the amplitude of the MMN was significantly higher in patients who evolved
later to a minimally conscious state [26].
The P3 is a positive inflexion generated when the subject detects a rare and unex-
pected stimulus. For an auditory potential, it appears approximately 300 ms after the
stimulus, while for a visual stimulus it can appear 500 or 600 ms after the stimulus
presentation. In case of cerebral lesions, its latency can be higher [27, 28]. The
MMN and P3 are two different cerebral responses elicited by similar stimuli (devi-
ant or novel), but they differ according to the time interval between stimuli. The
MMN is generated when the stimuli are close to one another but disappears when
the interval between two stimuli is longer than 2 s. The MMN originates from the
superior temporal gyrus and from the frontal cortex. The P3 relates to the activation
of a network of cerebral areas including frontoparietal regions [29]. The P3 is fre-
quently linked to cognitive processes of higher complexity than the N100 and MMN
components, such as categorization, decision-making, or updates in working mem-
ory. If simple sounds are sufficient for the generation of an MMN or a P3, the latter
can also be generated by more complex stimuli. The emotional valence of these
stimuli will have an impact on the amplitude. A stimulus such as the own name will
more likely trigger a P3 than a simple sound [30, 31] (Fig. 6.1). The presence of an
MMN and/or a P3 in acute minimally conscious patients is considered as a sign of
good prognostic but cannot disentangle patients in a vegetative state from patients
in a minimally conscious state [32–35]. From a behavioral viewpoint, the distinc-
tion between these two states can be made based on response to command. Hence,
active-evoked potentials are used to better evaluate consciousness in a patient as it
requires his/her active participation, which differs from passive listening. In an
experience where patients were asked to count occurrences of their own name,
6 Electrophysiology and Disorders of Consciousness 61

a b

P3

P3
c d

P3
P3

500 1,000 500 1,000

Fig. 6.1 Auditory evoked potentials in response to the own name in (a) healthy controls (n = 5),
(b) locked in syndrome patients (n = 4), (c) minimally conscious patients (n = 6) and (d) vegetative
patients (n = 5). The area in grey represents the difference in activation between the presentation of
the own name and the presentation of other names. A P3 response can be observed even in some
patients in a vegetative state. Electrode Pz (Adapted from [30])

presented along with seven other names, some minimally conscious patients dis-
played a P3 of greater amplitude than when passively listening to their name. On the
other hand, vegetative patients who showed a P3 in response to their name did not
display higher amplitudes when asked to actively count their names [31]. These
tasks have a similar goal than active tasks used in fMRI and are presented in another
chapter of this book.
The N400 is a negative inflexion which appears roughly 400 ms after a word
presentation. Its amplitude is increased if the stimulus is discordant (semantic
or phonologic discordances) based on the context (word or sentence). Care
must be taken, however, as a semantic incongruence can also lead to a P600, a
positive inflexion which appears 600 ms after stimulus presentation. Any
change, negative or positive, can, thus, be considered as incongruence process-
ing. These inflexions can also be observed in inattentive healthy controls, and
therefore it can be assumed that like the MMN and P3, they result from an
automatic process [33].
Evoked potentials are complementary to behavioral studies in patients. They can
be used according to a hierarchical approach [36] where low level functionalities are
first evaluated with exogenous potentials and then higher level processing is tested
with cognitive potentials. The latter is presented in passive and then in active tasks.
If the patient answers to the active task, he demonstrates a response to command. At
this stage, it becomes important to test communication tools with the patient such as
brain–computer interfaces, which are the scope of another chapter included in this
book.
62 Q. Noirhomme and R. Lehembre

Quantitative Electroencephalography

Quantitative electroencephalography (qEEG) consists in the use of algorithms in

order to extract complex measures likely to add objective information that cannot be
extracted from visual inspection of the EEG traces. For instance, one can compute
the power spectral density of the signal at each electrode location, assess the level
of connectivity between electrodes, or use measures based on signal complexity as
employed in anesthesia to establish the level of consciousness of a patient. Practically,
a qEEG analysis consists in the extraction of non artifacted temporal windows, last-
ing from one to several seconds, from a recordings lasting several minutes. The
different required parameters are computed on each of these windows and then
averaged in order to provide a global value. Furthermore, building large EEG data-
sets enable the use of statistical tools allowing a diagnostic or prognostic evaluation
of groups of patients.
The power spectral density can be obtained from the fast Fourier transform or a
wavelet transform computed on the extracted windows. It is expressed in absolute
power (uV2) or in relative power (%) for each individual electrode or for a subgroup
of electrodes corresponding to a region of interest. In comatose patients, the power
spectral density is higher in lower frequencies while it is lower in higher frequencies
as compared to healthy controls. Such findings can also be observed visually in
EEG traces (Fig. 6.2) [37, 38].
Tools initially developed for anesthesia monitoring have been proposed to evalu-
ate the level of consciousness in severely brain-injured patients. These tools are
used in the clinical field to measure the depth of anesthesia and to prevent the
patient’s arousal during a surgical procedure, while allowing drug savings and
a faster postoperative awakening thanks to a better control of the depth of
anesthesia.

0.18

0.16

0.14 Control
MCS
0.12 VS
Fig. 6.2 Normalized power
0.1
spectral density computed at
Cz in five different frequency
0.08
bands in healthy controls
(n = 5), patients in a 0.06
minimally conscious state
(n = 12, MCS) and patients in 0.04
a vegetative state (n = 10,
VS). Patients have more 0.02
power in low frequency
bands and less power in 0
higher frequency bands Delta Theta Alpha Beta Gamma
6 Electrophysiology and Disorders of Consciousness 63

Furthermore they are easy to use and interpret. For example, the bispectral index
(BIS) is a unitless measure ranging from 0 (inactive EEG) to 100 (normal activity),
which results from a combination of temporal and frequency parameters [39].
During sleep, BIS values gradually decrease [40]. In case of disorders of conscious-
ness, vegetative patients have a lower BIS value than patients in a minimally con-
scious state, but this value cannot systematically differentiate patients in a vegetative
state from patients in a minimally conscious state at an individual level [41]. Similar
results were obtained using the spectral entropy (EEG-entropy) [42, 43].
The EEG can also help to quantify connectivity between cerebral areas [44].
Connectivity is disrupted in patients suffering from disorders of consciousness [45].
Computing the coupling between electrodes provides a connectivity measure of
underlying brain areas. This measure gives complementary information that can be
used for diagnostic and prognostic purposes. To correctly interpret this measure, it
is important to take into account two inherent limitations: the volume conduction
problem and the choice of the reference. Indeed, the reference induces a common
component to the signals at each electrode location which will, therefore, appear
artificially coupled. It is, thus, advised to use bipolar montages in connectivity stud-
ies. Similarly, the diffusion of electrical currents through the cerebrospinal fluid and
the bone will result in the fact that different electrodes will record redundant infor-
mation and, again, will appear artificially connected. Correlation methods and
coherence, the spectral equivalent of correlation, are subject to these drawbacks.
They can nonetheless provide information, but their interpretation must take into
account the reference and volume conduction problem. A study on a single patient
in a vegetative state with a damaged right hemisphere showed a diminution of the
connectivity in the right hemisphere. Such decrease was not visible using spectral
power measures [46]. A coherence decrease has also been observed in minimally
conscious patients [47, 48]. Other measures of connectivity have been proposed in
the past 10 years to answer the volume conduction problem [44, 49]. The first results
with some of these new methods also tend to show a drop of connectivity [38, 48].
Finally, qEEG may also include source reconstruction methods, which estimate
the potential brain areas where the electrical signal is coming from [50]. These
methods are based on approximations of the brain structure. A core approximation
will model the brain as three concentric spheres and a finer approximation will use
magnetic resonance images of the patient’s brain in order to obtain a realistic model
of the brain. All the parameters computed at the scalp level can also be computed at
the source level, which provides a better spatial resolution and allows a three dimen-
sion visualization of the brain’s activity. However, these methods require at least 32
electrodes, 64 or more electrodes being recommended [50].
To conclude, qEEG allows various objective measurements of the cerebral activ-
ity providing information on the patient’s consciousness level. Encouraging results
have shown differences between groups of patients according to the diagnosis and
the prognosis. It is, however, not yet possible to target one single parameter allow-
ing an accurate diagnosis at the individual level. Future challenges consist in linking
the excellent temporal resolution of the EEG with the spatial resolution of the fMRI
in order to have a better comprehension of disorders of consciousness.
64 Q. Noirhomme and R. Lehembre

Conclusion

While the interpretation of the EEG trace can give some information on the patient’s
consciousness level and detect epileptic activity, exogenous-evoked potentials can
give useful information as regards the patient’s prognosis (e.g., N20 or MMN) and
remnant stimulus processing. Active cognitive evoked potentials can also detect
signs of consciousness such as response to command. Finally, quantitative measures
offer a new interpretation of the EEG and provide new objective information (e.g.,
by using coherence and source reconstruction methods). They, nevertheless, require
more testing and improvements before being applied in clinical settings.

References

1. Guideline seven: a proposal for standard montages to be used in clinical EEG. American
Electroencephalographic Society. J Clin Neurophysiol. 1994;11:30–6. (http://www.ncbi.nlm.
nih.gov/pubmed/8195414)
2. Krauss GL, Fisher RS, editors. The Johns Hopkins atlas of digital EEG: an interactive training
guide. Baltimore: The Johns Hopkins University Press; 2006.
3. Brenner RP. The interpretation of the EEG in stupor and coma. Neurologist. 2005;11:271–84.
4. Young GB, et al. Initial assessment and management of the patient with impaired alertness. In:
Young GB, editor. Coma and impaired consciousness. New York: McGraw-Hill; 1998. p. 97–103.
5. Posner JB, Saper CB, Schiff ND, Plum F. Plum and Posner’s diagnosis of stupor and coma.
New York: Oxford University Press; 2007.
6. Young GB, McLachlan RS, Kreeft JH, Demelo JD. An electroencephalographic classification
for coma. Can J Neurol Sci. 1997;24:320–5.
7. Synek VM. Prognostically important EEG coma patterns in diffuse anoxic and traumatic
encephalopathies in adults. J Clin Neurophysiol. 1988;5:161–74.
8. Westmoreland BF, Klass DW, Sharbrough FW, Reagan TJ. Alpha-coma. Electroencephalographic,
clinical, pathologic, and etiologic correlations. Arch Neurol. 1975;32:713–8.
9. Kaplan PW. The clinical features, diagnosis, and prognosis of nonconvulsive status epilepti-
cus. Neurologist. 2005;11:348–61.
10. Vanhaudenhuyse A, Laureys S, Perrin F. Cognitive event-related potentials in comatose and
post-comatose states. Neurocrit Care. 2007;8:262–70.
11. Guerit JM. Evoked potentials in severe brain injury. Prog Brain Res. 2005;150:415–26.
12. Amantini A, Grippo A, Fossi S, et al. Prediction of ‘awakening’ and outcome in prolonged
acute coma from severe traumatic brain injury: evidence for validity of short latency SEPs.
Clin Neurophysiol. 2005;116:229–35.
13. Fischer C, Luaute J, Nemoz C, et al. Improved prediction of awakening or nonawakening from
severe anoxic coma using tree-based classification analysis. Crit Care Med. 2006;34:1520–4.
14. Lew HL, Dikmen S, Slimp J, et al. Use of somatosensory-evoked potentials and cognitive
event-related potentials in predicting outcomes of patients with severe traumatic brain injury.
Am J Phys Med Rehabil. 2003;82:53–61; quiz 62–4, 80.
15. Robinson LR, Micklesen PJ, Tirschwell DL, Lew HL. Predictive value of somatosensory
evoked potentials for awakening from coma. Crit Care Med. 2003;31:960–7.
16. de Sousa LC, Colli BO, Piza MR, et al. Auditory brainstem response: prognostic value in
patients with a score of 3 on the Glasgow Coma Scale. Otol Neurotol. 2007;28:426–8.
17. Haupt WF, Pawlik G, Thiel A. Initial and serial evoked potentials in cerebrovascular critical
care patients. J Clin Neurophysiol. 2006;23:389–94.
6 Electrophysiology and Disorders of Consciousness 65

18. Laureys S, Perrin F, Schnakers C, et al. Residual cognitive function in comatose, vegetative
and minimally conscious states. Curr Opin Neurol. 2005;18:726–33.
19. Fischer C, Luaute J, Adeleine P, Morlet D. Predictive value of sensory and cognitive evoked
potentials for awakening from coma. Neurology. 2004;63:669–73.
20. Glass I, Sazbon L, Groswasser Z. Mapping “cognitive” event-related potentials in prolonged
postcoma unawareness state. Clin Electroencephalogr. 1998;29:19–30.
21. Guerit JM, Verougstraete D, de Tourtchaninoff M, et al. ERPs obtained with the auditory odd-
ball paradigm in coma and altered states of consciousness: clinical relationships, prognostic
value, and origin of components. Clin Neurophysiol. 1999;110:1260–9.
22. Mutschler V, Chaumeil CG, Marcoux L, et al. Auditory P300 in subjects in a post-anoxic
coma. Preliminary data. Neurophysiol Clin. 1996;26:158–63.
23. Naccache L, Puybasset L, Gaillard R, et al. Auditory mismatch negativity is a good predictor
of awakening in comatose patients: a fast and reliable procedure. Clin Neurophysiol.
2005;116:988–9.
24. Daltrozzo J, Wioland N, Mutschler V, Kotchoubey B. Predicting coma and other low respon-
sive patients outcome using event-related brain potentials: a meta-analysis. Clin Neurophysiol.
2007;118:606–14.
25. Kotchoubey B, Lang S, Mezger G, et al. Information processing in severe disorders of conscious-
ness: vegetative state and minimally conscious state. Clin Neurophysiol. 2005;116:2441–53.
26. Wijnen VJ, van Boxtel GJ, Eilander HJ, de Gelder B. Mismatch negativity predicts recovery
from the vegetative state. Clin Neurophysiol. 2007;118:597–605.
27. Munte TF, Heinze HJ. Brain potentials reveal deficits of language processing after closed head
injury. Arch Neurol. 1994;51:482–93.
28. Granovsky Y, Sprecher E, Hemli J, Yarnitsky D. P300 and stress in mild head injury patients.
Electroencephalogr Clin Neurophysiol. 1998;108:554–9.
29. Pegado F, Bekinschtein T, Chausson N, et al. Probing the lifetimes of auditory novelty detec-
tion processes. Neuropsychologia. 2010;48:3145–54.
30. 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.
31. Schnakers C, Perrin F, Schabus M, et al. Voluntary brain processing in disorders of conscious-
ness. Neurology. 2008;71:1614–20.
32. Fischer C, Dailler F, Morlet D. Novelty P3 elicited by the subject’s own name in comatose
patients. Clin Neurophysiol. 2008;119:2224–30.
33. Vanhaudenhuyse A, Laureys S, Perrin F. Cognitive event-related potentials in comatose and
post-comatose states. Neurocrit Care. 2008;8:262–70.
34. Kotchoubey B. Event-related potential measures of consciousness: two equations with three
unknowns. Prog Brain Res. 2005;150:427–44.
35. Fischer C, Luaute J, Morlet D. Event-related potentials (MMN and novelty P3) in permanent
vegetative or minimally conscious states. Clin Neurophysiol. 2010;121:1032–42.
36. Kubler A, Kotchoubey B. Brain-computer interfaces in the continuum of consciousness. Curr
Opin Neurol. 2007;20:643–9.
37. Leon-Carrion J, Martin-Rodriguez JF, Damas-Lopez J, et al. Brain function in the minimally con-
scious state: a quantitative neurophysiological study. Clin Neurophysiol. 2008;119:1506–14.
38. Lehembre R, Bruno MA, Vanhaudenhuyse A, et al. Resting state EEG study of comatose
patients: a connectivity and frequency analysis to find differences between vegetative and
minimally conscious states. Funct Neurol 2012 (in press).
39. Johansen JW, Sebel PS. Development and clinical application of electroencephalographic
bispectrum monitoring. Anesthesiology. 2000;93:1336–44.
40. Noirhomme Q, Boly M, Bonhomme V, et al. Bispectral index correlates with regional cerebral
blood flow during sleep in distinct cortical and subcortical structures in humans. Arch Ital Biol.
2009;147:51–7.
41. Schnakers C, Ledoux D, Majerus S, et al. Diagnostic and prognostic use of bispectral index in
coma, vegetative state and related disorders. Brain Inj. 2008;22:926–31.
66 Q. Noirhomme and R. Lehembre

42. Gosseries O, Schnakers C, Ledoux D, et al. Automated EEG entropy measurements in coma,
vegetative state/unresponsive wakefulness syndrome and minimally conscious state. Funct
Neurol. 2011;26(1):25–30.
43. Viertio-Oja H, Maja V, Sarkela M, et al. Description of the entropy algorithm as applied in the
datex-ohmeda S/5 entropy module. Acta Anaesthesiol Scand. 2004;48:154–61.
44. Pereda E, Quiroga RQ, Bhattacharya J. Nonlinear multivariate analysis of neurophysiological
signals. Prog Neurobiol. 2005;77:1–37.
45. Laureys S. The neural correlate of (un)awareness: lessons from the vegetative state. Trends
Cogn Sci. 2005;9:556–9.
46. Davey MP, Victor JD, Schiff ND. Power spectra and coherence in the EEG of a vegetative
patient with severe asymmetric brain damage. Clin Neurophysiol. 2000;111:1949–54.
47. Schiff N. Large scale brain dynamics and connectivity in the minimally conscious state. In:
Handbook of brain connectivity. New York: Springer; 2007. p. 505–20.
48. Pollonini L, Pophale S, Situ N, et al. Information communication networks in severe traumatic
brain injury. Brain Topogr. 2010;23:221–6.
49. Stam CJ. Nonlinear dynamical analysis of EEG and MEG: review of an emerging field. Clin
Neurophysiol. 2005;116:2266–301.
50. Michel CM, Murray MM, Lantz G, et al. EEG source imaging. Clin Neurophysiol.
2004;115:2195–222.
Chapter 7
Brain-Computer Interface:
A Communication Aid?

Camille Chatelle, Zulay Lugo, Quentin Noirhomme, Bettina Sorger,

and Dorothee Lulé

Introduction

A brain-computer interface (BCI) is a system allowing for communication between

the brain and the external environment. It is independent from any peripheral neural
or muscular activity and it directly converts brain activity into a computerized com-
mand [1]. A BCI is based on cerebral activity measured using techniques, such as
electroencephalography (EEG), functional magnetic resonance imagery (fMRI),

C. Chatelle, M.Sc. (*) • Q. Noirhomme, Ph.D.

Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du Six août, 8, 4000 Liège, Belgium
e-mail: [email protected]; [email protected]
Z. Lugo, M.Sc.
Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du Six août, 8, 4000 Liège, Belgium
Department of Psychology I, Biological Psychology, Clinical Psychology
and Psychotherapy, University of Würzburg, Würzburg, Germany
e-mail: [email protected]
B. Sorger, Ph.D.
Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du Six août, 8, 4000 Liège, Belgium
Department of Cognitive Neuroscience, Faculty of Psychology and Neuroscience,
Maastricht University and Maastricht Brain Imaging Centre (M-BIC),
Maastricht, The Netherlands
e-mail: [email protected]
D. Lulé, Ph.D.
Section of Neurophysiology, Department of Neurology, University of Ulm,
Albert-Einstein-Allee 47, Ulm, Germany
e-mail: [email protected]

C. Schnakers, S. Laureys (eds.), Coma and Disorders of Consciousness, 67

DOI 10.1007/978-1-4471-2440-5_7, © Springer-Verlag London 2012
68 C. Chatelle et al.

implanted electrodes or functional near-infrared spectroscopy (fNIRS) in order to

control the environment [2]. A BCI is not a “mind-reading” device. Its primary
function is to provide the subject a virtual keyboard, whose keys are pressed using
brain activity. Each key press constitutes a choice of an item from a set of items.
This choice is made through the control of neuroelectrical activity [3–5]. A specific
algorithm translates the extracted features into commands that represent the users’
intent. These commands can control effectors to select items (e.g., words). Recent
development has shown the usefulness of BCIs in controlling motor prosthesis, cur-
sors, access to internet, and communication [5–10]. Here, we focused on systems
allowing functional communication with the surrounding. These tools, extensively
validated in healthy volunteers, could be of interest particularly for communicating
with patients whose cognitive functions are intact but are paralyzed and anarthric
following a neurological or muscular damage, e.g., patients with a locked-in syn-
drome (LIS) [11]. Several methods have been developed to allow communication
with these patients. The simplest is based on eye movements or blinks [e.g., one eye
blink for yes, two blinks for no]. There are other methods allowing the expression
of more complex ideas using alphabetic systems [12]. These methods are based on
the patient’s residual motor ability. However, in some cases, it is necessary to use a
communication system that does not involve motor skills at all. Those motor-inde-
pendent systems are not only useful for using alphabetic systems and expressing
more complex ideas [12]. BCIs could be the key to providing access to the outside
world for a LIS patient [13]. Finally, beyond communication, it is also possible to
use these systems as a diagnostic aid in patients with disorders of consciousness
(DOC). In fact, it could be used to detect a response to a command in the absence of
discernible behavior at the bedside [14]. Here, we present the recent progress in the
development of BCIs. Moreover, we discuss clinical applications in LIS patients
and studies performed in patients recovering from coma.

BCI Techniques

Electroencephalography – EEG

BCIs based on EEG use evoked potentials (ERPs), more precisely components such
as the P3 or steady-state visually evoked potentials (SSVEPs), sensorimotor rhythms
(SMRs), slow cortical potentials (SCPs) and the alpha rhythm. The use of ERPs is
a common approach in BCI research. The most widely used component is the P3, a
positive deflection occurring about 300 ms after the presentation of a target stimu-
lus. The advantage of using P3 BCIs is that these may include up to 36 different
commands. Donchin and his colleagues have developed a visual BCI using a 6 × 6
matrix composed of letters and signs [15]. The rows and columns are successively
illuminated. The subject has to focus his/her attention on the letter he/she wants to
spell, eliciting a P3. With this type of BCI, users would be able to spell up to 7–8
7 Brain-Computer Interface: A Communication Aid? 69

400

350

x: 14.16
Y:295.3
300

250
Power in µV2

200

150

100

x: 28.32
50 Y:31.91
x: 10.74 x: 21.24 x: 42.48
x: 31.98 Y:7.052
Y:4.475 Y:5.651
Y:0.4626
0
0 10 20 30 40 50 60
Frequencies in Hz

Fig. 7.1 SSVEP. Electrical brain responses (fast Fourier transformed) observed in response to two
visual stimuli flashing at a frequency of 10 (yellow) and 14 Hz (red). The subject had to focus on
the 14 Hz stimulus. Here, the first three harmonics in response to these stimuli are represented
(Illustration by Nicolas Partoune)

letters per minute with an accuracy of 80–90%. This paradigm has also been adapted
to the auditory modality, although the performance achieved is lower and the system
has been reported as requiring higher concentration [16]. Those nonvisual P3-based
BCIs have been developed for patients who lost control of their eye movements or
who are blind [3, 16–18]. Sellers and Donchin [3] tested a BCI for binary commu-
nication. In this paradigm, the user hears a sequence of four stimuli (yes, no, stop,
go) presented in randomized order. He/she should, then, focus on either yes or no in
each sequence. The task required to count the number of times the target (either yes
or no) was presented. The probability of the selected target stimulus was 0.25, which
is sufficient to elicit a P3 [15]. The authors have shown a reliable response during
ten sessions in healthy volunteers. Halder also showed the reliability of an auditory
BCI for binary communication [19], with an efficiency of 2.46 bits/min and a
classification accuracy of 78.5%. Some researchers have focused on other compo-
nents of ERPs such as the SSVEPs. These are visual ERPs and are elicited in occipi-
tal and parieto-occipital areas in response to visual stimulation. These are of small
amplitude, consistent [20] (Fig. 7.1) and not susceptible to be influenced by ocular
70 C. Chatelle et al.

and electromyographic artefacts [21]. The SSVEP-based BCIs offer the possibility
of using 2–13 different commands without affecting the classification accuracy
(64–96.5%) [21]. Those BCIs have shown their reliability with respect to the con-
trol of a cursor to execute commands or to control prosthesis [21]. A recent study
has shown its applicability for communication using a spelling words system based
on letters flashing at different frequencies [22]. The authors reported an average
accuracy of 92.25% and an efficiency of 37.62 bits/min (or 5.51 letters/min).
Changes in SMRs or m rhythms have also been used for BCI purposes. SMRs
refer to EEG activity of 8–15 Hz that can be recorded in primary sensorimotor areas
[1] and which is usually accompanied by a beta activity (18–26 Hz). This activity
can be reduced or desynchronized by preparing, executing, or imagining a move-
ment (event-related desynchronization), particularly in the contralateral motor
region. An increase in the SMR, or synchronization, occurs following the execution
of a movement and during relaxation [23]. The advantage here is that these compo-
nents do not require the actual execution of the movement but solely the kinaes-
thetic mental imagery of this [24]. However, it is not possible to use more than two
commands, the increase to three or more leading to a decrease in the classification
accuracy. In healthy subjects, many BCIs have shown satisfying results in produc-
ing words based on visual [25] and auditory [26] input. The lowest frequencies of
signals generated by the cortex and recorded at the scalp are the SCPs. The negative
SCPs are usually associated with movements and other functions that involve corti-
cal activation, while positive SCPs are usually associated with a reduction in corti-
cal activity [27]. This system is also limited to two (or less) commands. It has been
shown that it is possible to teach participants to control their brain activity (i.e., the
SCPs) to move an object on a screen [28]. In this context, communication systems
have been developed for patients with LIS (see below).
Alpha rhythms have also been implemented for BCIs. Vidal was the first to
show that it was possible to determine the position of eye gaze and the direction in
which the user wants to move a cursor through the activity generated in the visual
cortex [29]. Implicit changes in visual attention are associated with a desynchroni-
zation in alpha rhythm in the posterior regions, contralateral [30] to the target posi-
tion and with an ipsilateral synchronization [31]. A BCI was then developed by van
Gerven and colleagues [32] based on implicit spatial attention. They showed the
possibility of using the alpha band power to classify subjects’ responses (n = 15)
with an accuracy of 29–69% (25% chance level). This paradigm has also shown a
rate of 20 bits/min.

Neuroimaging

Even though it is more expensive than EEG, fMRI is also a technique used for BCIs,
especially due to its high spatial resolution. It measures changes in the blood oxy-
genation level in the brain (hemodynamic activity) and is related to neuronal activ-
ity [33]. Boly and colleagues proposed a paradigm based on motor and visual
7 Brain-Computer Interface: A Communication Aid? 71

imagery (i.e., to imagine playing tennis versus to imagine visiting the own house).
Both tasks showed clearly distinct brain activation in the supplementary motor and
parahippocampal areas, respectively [34]. To establish a communication, Monti and
colleagues asked participants to respond to autobiographical questions by yes or no
using one of these mental imagery tasks. At the end of each sequence (5 min), the
researchers were able to infer the responses of 16 participants with an accuracy of
100% [35]. Recently, another paradigm has been proposed to allow more complex
communication than binary yes/no answers. Sorger and colleagues [2] have tested a
communication paradigm in eight healthy volunteers using two mental tasks (e.g.,
motor imagery and mental arithmetic) and different temporal answer parameters
(task onset and duration). The questions were related to multiple choice answers (A,
B, C, or D). For answering a question, the participants had to look at the letter cor-
responding to the target response and perform the mental task associated to this
letter as long as this one was illuminated on the screen (Fig. 7.2). Across all partici-
pants, 37 out of 39 responses were correctly decoded (94.7%). In six participants,
the accuracy was of 100%, and in the other two participants, it remained above the
level of chance (85.7% and 75%, chance level being 25%).
Another promising BCI technique is fNIRS. Like fMRI, it measures the hemo-
dynamic response of the brain [36]. The main advantage of the fNIRS as opposed to
fMRI is that it is portable and less sensitive to motion artefacts. However, it mea-
sures activity in the cortex, and cannot reach deep areas (more than 3 cm deep) in
contrast to fMRI. Coyle and colleagues have proposed a BCI (Mindswitch) based on
one channel allowing generating two commands (yes/no) and then offering a binary
communication [37]. Several studies have also shown the usefulness of the fNIRS-
based BCIs using multiple channels [38, 39] and reporting accuracy between 80%
and 89% in healthy subjects (50% chance level).

Development of BCIs as Diagnostic and Communication

Tools in DOC

EEG-Based BCIs

EEG-based BCIs have been applied in patients with amyotrophic lateral sclerosis
(ALS) [5, 40]. However, there is variability in the results. One study showed that it
was possible to establish communication [40] in five out of six ALS patients using
the P3-based paradigm developed by Donchin [15]. Four of them used this system
later for spelling words and have demonstrated functional communication. Sellers
and Donchin showed reliable responses in ALS patients with their auditory para-
digm based on four stimuli (yes, no, stop, go), despite lower accuracies than in
healthy participants [3]. Kübler has adapted the use of a matrix in the auditory
modality. Five rows and five columns represented the letters of the alphabet [17].
The five lines were associated with a number between 1 and 5 and the five columns
72 C. Chatelle et al.

a Patient reads queation and selects answer option “A”

Which color do you like most?

motor imagery A C
mental calculation B D
A red
B blue
C green
D black

b Patient hemodynamically encodes letter “A”

answer encoding display evoked brain fMRI reponse within

activation (black) ‘motor imaery’ region

motor imagery A C
mental calculation B D
0−5 s

motor imagery A C
mental calculation B D
5−10 s

time
motor imagery A C
mental calculation B D
10−15 s

motor imagery A C
mental calculation B D
15−20 s

c Classifier automatically decodes selected answer option.

fMRI response

A red
B blue
“A”
C green
D black

0 5 10 15 20
time [s]
7 Brain-Computer Interface: A Communication Aid? 73

with a number between 6 and 10. The numbers were auditorily presented and the
patient selected the row and the column of the target letter. Four ALS patients were
evaluated with this system, demonstrating adequate performance for visual (more
than 70%) but not for auditory (just above chance level) communication. Moreover,
users reported more difficulty to concentrate in the auditory condition. To our
knowledge, Lulé and colleagues performed the first study in patients with DOC
[41]. They used the Sellers and Donchin’s paradigm to test its reliability as a diag-
nostic tool. The results have shown its benefits in improving the detection of con-
sciousness in patients with DOC (Fig. 7.3). Two LIS patients were also evaluated,
only one of them having reliably used the BCI. This shows that the system cannot
ensure the absence of consciousness in case of negative results [41].
Using SMRs, Neuper and colleagues [43] have trained a paralyzed patient to use
a language support program (LSP, [44]) in order to communicate. The spelling
involved the selection of a letter in successive steps using a virtual keyboard. A
predefined set of letters was divided in two subsets and presented at the top and at
the bottom of the screen. The patient was instructed to select one of this subset by
either relaxing or by using motor imagery. When the patient had selected the subset
containing the target letter, this subset was itself divided into two parts, and this
until the patient selected the target letter. After several months, the patient was able
to control the keyboard with an accuracy of 70%. Another study showed the possi-
bility for a patient with ALS to use a keyboard by the control of SMRs [45]. These
evoked potentials were also recently used with DOC patients. Cruse and colleagues
have detected command following (i.e., to imagine squeezing the right hand versus
to imagine moving all the fingers or toes) in a patient clinically diagnosed as being
in a vegetative state [46]. As with SMRs, the LSP can be used by patients via the
control of SCPs rhythms [47, 48]. Patients with advanced ALS showed their abili-
ties, after training, to use the LSP in increasing or decreasing their brain activity to
select the block corresponding to the target letter and then to write words [42].

Fig. 7.2 Answering multiple-choice questions based on intentionally generated single-trial fMRI
responses (example). (Panel a) A question and four answer options are displayed on a screen. The
participant is asked to select an answer and identify the corresponding letter in a letter encoding
display (see gray table). In the present example, the patient selects letter “A” (located in first row
and second column in the table; see white circle). (Panel b) During answer encoding, the patient
performs the specific mental task that is shown in the same row as the desired letter (“motor imag-
ery”, in the present example). The patient performs this task as long as the desired letter is high-
lighted by a light gray bar that moves from the left to the right side of the display during each
answer encoding trial. In the present example, in order to encode the letter A, the patient has to
perform “motor imagery” for the first 10 s of the trial. The patient’s performance results in a
specific spatiotemporal pattern of brain activation, which, in the present example, is unique to the
letter A (see brain activations in right column). (Panel c) During answer decoding, an algorithm
classifies the patient’s answer using multifilter correlations. More specifically, the obtained brain
activation pattern (in the present example, within a “motor imagery” region) is compared to four
different template patterns, with each template corresponding to one of the four possible answers.
In this way, the patient’s answer is decoded fully automatically and immediately after each trial
(i.e., answer)
74 C. Chatelle et al.

r2 769.5313 ms r2
500 ms

0.04

0.02

CP4 CP3
20 10
Yes Yes
Amplitude (µV)

Amplitude (µV) 5
10
0
0
−5
No, go, stop No, go, stop
−10 −10
0 200 400 600 800 1,000 0 200 400 600 800 1,000
Time (ms)

Fig. 7.3 P3 in a healthy volunteer (left) and a patient in a minimally conscious state (right) in
response to the target stimulus yes. On the top, distribution of the observed response to target
stimuli can be seen (yes). The colors in the images in the upper row represent the difference in the
observed response between the target stimuli and nontarget stimuli. The greater the difference, the
more the region is colored red–orange. Below, averaging over all responses for the other three
stimuli (nontargets no, stop, go; in green) and for all responses to target stimuli (yes; in brown) is
shown (Adapted from [42])

fMRI-Based BCIs

Functional MRI is particularly useful for the diagnosis of DOC rather than for com-
munication in LIS patients. The paradigm of mental imagery developed by Boly and
colleagues [34] was administered to a patient diagnosed as being in a vegetative
state. When she was asked to imagine playing tennis and visiting her house, the
patient showed an activity similar to controls on both tasks in the supplementary
motor areas and parahippocampal regions, suggesting she understood and followed
the command [14]. A few months later, she evolved into a minimally conscious
state. This case illustrates the importance of alternative techniques in the diagnosis
of patients recovering from coma. In fact, it appears that the patient was at that time
unable to show signs of consciousness through any other technique. Subsequently,
54 patients were studied using this protocol and five of them showed a response
similar to controls (four being diagnosed in a vegetative state and one in a mini-
mally conscious state) [49]. A yes/no communication was established in one of
these patients using fMRI. The patient was able to adequately respond five out of six
7 Brain-Computer Interface: A Communication Aid? 75

questions using mental imagery tasks, supporting the usefulness of this technique in
establishing a functional communication in patients unable to communicate at the
bedside.
The fNIRS is a fairly new technique that has not been much studied in BCI
research. To our knowledge, there is only one study demonstrating the usefulness
of fNIRS as a communication tool for patients. Naito and colleagues used fNIRS
in 40 ALS patients, including 17 total LIS [50]. They used a single channel and
two wavelengths (770 and 840 nm) measuring the activity in the frontal area. The
subjects had to answer various questions with a “yes” by performing mental arith-
metic or singing (and, hence, increasing their brain activity in frontal regions), or
a “no” by relaxing. The results show that 70% of incomplete LIS patients and
40% of patients with a complete LIS were able to respond yes/no with an accuracy
of 80%.

Invasive BCIs Systems

So far, we have presented BCIs using noninvasive systems. As many systems are
based on EEG signals measured on the scalp, the quality of recordings is relatively
low (distorted signal and low amplitude), the spatial resolution is limited and train-
ing is necessary. Therefore, some studies have focused on invasive recording meth-
ods. Recordings are performed either directly at the neuronal level [51, 52] or on the
brain surface in the case of an electrocorticographic recording [53–55]. BCIs based
on intracortical microelectrodes can directly record the activity of neurons and pro-
vide a stronger signal. These allow users to control devices such as computer cur-
sors more quickly and accurately [55]. While this technique has not been tested in
healthy subjects, patients showed good performances in the context of complex
communication (spelling words) [52, 54]. This technique, nevertheless, implies
clinical risks and its reliability is limited. Therefore, there is still a long way to go
before this device could be used in LIS patients on a daily basis [1].

Conclusion and Perspectives

We have presented a review of the literature on various BCIs tested in patients with
severe motor disabilities. The success in using a BCI still varies considerably
between individuals. If some systems work in healthy subjects, this does not neces-
sarily mean that it will work in specific patient-target groups, e.g., in severely para-
lyzed patients [56]. In about 20% of potential users, the accuracy obtained is not
sufficient to allow the patient to control an application [57]. Therefore, it seems
important to validate BCIs directly in the target population and to adapt these to the
needs and the problems specific to this population (e.g., to use auditory paradigms
if visual control problems occur). In the future, it is important to develop systems
76 C. Chatelle et al.

that are reliable and easy to use in the everyday life. New algorithms should include
the automatic detection of artefacts, single-trial classification, and the possibility to
classify a session without training sessions.
BCIs might provide useful information about patients recovering from coma.
These could be used to improve the clinical diagnosis as it has already been sug-
gested by fMRI and EEG studies using imagery tasks. However, the typical vigi-
lance fluctuation observed in DOC patients is a major confounding factor for this
BCI application [58]. Indeed, fMRI could demonstrate reliable detection of signs of
consciousness in patients diagnosed as being in a vegetative state. However, many
patients have been evaluated and only a few have shown signs of consciousness with
this paradigm, including patients showing signs of consciousness at the bedside.
Further research is needed to know whether this is due to a lack of awareness in
some patients, or due to vigilance fluctuation. Currently available BCI systems will
have to be tested in such patients. Active tasks such as motor imagery seem to be
promising in the detection of signs of consciousness [2, 13, 14, 16, 19, 34, 40].

References

1. Wolpaw JR, Birbaumer N, McFarland DJ, et al. Brain-computer interfaces for communication
and control. Clin Neurophysiol. 2002;113(6):767–91.
2. Sorger B, Dahmen B, Reithler J, et al. Another kind of ‘BOLD Response’: answering multiple-
choice questions via online decoded single-trial brain signals. Prog Brain Res. 2009;177:
275–92.
3. Sellers EW, Donchin E. A P300-based brain-computer interface: initial tests by ALS patients.
Clin Neurophysiol. 2006;117(3):538–48.
4. Sellers EW, Kubler A, Donchin E. Brain-computer interface research at the University of
South Florida Cognitive Psychophysiology Laboratory: the P300 speller. IEEE Trans Neural
Syst Rehabil Eng. 2006;14(2):221–4.
5. Kübler A. Brain-computer interfaces for communication in paralysed patients and implica-
tions for disorders of consciousness. In: Laureys S, Tononi G, editors. The neurology of con-
sciousness. London/Burlington/San Diego: Academic Press/Elsevier; 2008. p. 217–34.
6. Citi L, Poli R, Cinel C, Sepulveda F. P300-based BCI mouse with genetically-optimized ana-
logue control. IEEE Trans Neural Syst Rehabil Eng. 2008;16(1):51–61.
7. Yoo SS, Fairneny T, Chen NK, et al. Brain-computer interface using fMRI: spatial navigation
by thoughts. Neuroreport. 2004;15(10):1591–5.
8. Mugler EM, Ruf CA, Halder S, et al. Design and implementation of a P300-based brain-
computer interface for controlling an internet browser. IEEE Trans Neural Syst Rehabil Eng.
2010;18:599–609.
9. Sellers EW, Vaughan TM, Wolpaw JR. A brain-computer interface for long-term independent
home use. Amyotroph Lateral Scler. 2010;11:449–55.
10. Lee JH, Ryu J, Jolesz FA, et al. Brain-machine interface via real-time fMRI: preliminary study
on thought-controlled robotic arm. Neurosci Lett. 2009;450(1):1–6.
11. Schnakers C, Majerus S, Goldman S, et al. Cognitive function in the locked-in syndrome. J
Neurol. 2008;255(3):323–30.
12. Bruno MA, Schnakers C, Damas F, et al. Locked-in syndrome in children: report of five cases
and review of the literature. Pediatr Neurol. 2009;41(4):237–46.
13. Kubler A, Neumann N. Brain-computer interfaces–the key for the conscious brain locked into
a paralyzed body. Prog Brain Res. 2005;150:513–25.
7 Brain-Computer Interface: A Communication Aid? 77

14. Owen AM, Coleman MR, Boly M, et al. Detecting awareness in the vegetative state. Science.
2006;313(5792):1402.
15. Donchin E, Spencer KM, Wijesinghe R. The mental prosthesis: assessing the speed of a P300-
based brain-computer interface. IEEE Trans Rehabil Eng. 2000;8(2):174–9.
16. Furdea A, Halder S, Krusienski DJ, et al. An auditory oddball (P300) spelling system for
brain-computer interfaces. Psychophysiology. 2009;46(3):617–25.
17. Kubler A, Furdea A, Halder S, et al. A brain-computer interface controlled auditory event-
related potential (p300) spelling system for locked-in patients. Ann N Y Acad Sci. 2009;
1157:90–100.
18. Schreuder M, Blankertz B, Tangermann M. A new auditory multi-class brain-computer inter-
face paradigm: spatial hearing as an informative cue. PLoS One. 2010;5(4):e9813.
19. Halder S, Rea M, Andreoni R, et al. An auditory oddball brain-computer interface for binary
choices. Clin Neurophysiol. 2010;121(4):516–23.
20. Regan D. Some characteristics of average steady-state and transient responses evoked by
modulated light. Electroencephalogr Clin Neurophysiol. 1966;20(3):238–48.
21. Vialatte FB, Maurice M, Dauwels J, Cichocki A. Steady-state visually evoked potentials: focus
on essential paradigms and future perspectives. Prog Neurobiol. 2010;90(4):418–38.
22. Cecotti H. A self-paced and calibration-less SSVEP-based brain-computer interface speller.
IEEE Trans Neural Syst Rehabil Eng. 2010;18(2):127–33.
23. Pfurtscheller G, Lopes da Silva FH. Event-related EEG/MEG synchronization and desynchro-
nization: basic principles. Clin Neurophysiol. 1999;110(11):1842–57.
24. Pfurtscheller G, Neuper C, Flotzinger D, Pregenzer M. EEG-based discrimination between
imagination of right and left hand movement. Electroencephalogr Clin Neurophysiol. 1997;
103(6):642–51.
25. Scherer R, Muller GR, Neuper C, et al. An asynchronously controlled EEG-based virtual key-
board: improvement of the spelling rate. IEEE Trans Biomed Eng. 2004;51(6):979–84.
26. Nijboer F, Furdea A, Gunst I, et al. An auditory brain-computer interface (BCI). J Neurosci
Methods. 2008;167(1):43–50.
27. Birbaumer N. Slow cortical potentials: their origin, meaning, and clinical use. In: van Boxtel
GJM, Bocker KBE, editors. Brain and behavior past, present, and future. Tilburg: Tilburg
University Press; 1997. p. 25–39.
28. Elbert T, Rockstroh B, Lutzenberger W, Birbaumer N. Biofeedback of slow cortical potentials.
I. Electroencephalogr Clin Neurophysiol. 1980;48(3):293–301.
29. Vidal JJ. Toward direct brain-computer communication. Annu Rev Biophys Bioeng.
1973;2:157–80.
30. Thut G, Nietzel A, Brandt SA, Pascual-Leone A. Alpha-band electroencephalographic activity
over occipital cortex indexes visuospatial attention bias and predicts visual target detection.
J Neurosci. 2006;26(37):9494–502.
31. Kelly SP, Lalor EC, Reilly RB, Foxe JJ. Increases in alpha oscillatory power reflect an active
retinotopic mechanism for distracter suppression during sustained visuospatial attention.
J Neurophysiol. 2006;95(6):3844–51.
32. van Gerven M, Jensen O. Attention modulations of posterior alpha as a control signal for two-
dimensional brain-computer interfaces. J Neurosci Methods. 2009;179(1):78–84.
33. Kwong KK, Belliveau JW, Chesler DA, et al. Dynamic magnetic resonance imaging of human
brain activity during primary sensory stimulation. Proc Natl Acad Sci USA. 1992;89(12):5675–9.
34. Boly M, Coleman MR, Davis MH, et al. When thoughts become action: an fMRI paradigm to
study volitional brain activity in non-communicative brain injured patients. Neuroimage.
2007;36(3):979–92.
35. Monti M, Colemand MR, Owen AM. “Brain-reading” with real-time fMRI: communication
via detection of brain states in the absence of motor response. In: 14th annual meeting of the
organization for the human brain mapping. Melbourne: Elsevier; 2008. p. 133.
36. Irani F, Platek SM, Bunce S, et al. Functional near infrared spectroscopy (fNIRS): an emerging
neuroimaging technology with important applications for the study of brain disorders. Clin
Neuropsychol. 2007;21(1):9–37.
78 C. Chatelle et al.

37. Coyle SM, Ward TE, Markham CM. Brain-computer interface using a simplified functional
near-infrared spectroscopy system. J Neural Eng. 2007;4(3):219–26.
38. Luu S, Chau T. Decoding subjective preference from single-trial near-infrared spectroscopy
signals. J Neural Eng. 2009;6(1):016003.
39. Sitaram R, Zhang H, Guan C, et al. Temporal classification of multichannel near-infrared
spectroscopy signals of motor imagery for developing a brain-computer interface. Neuroimage.
2007;34(4):1416–27.
40. Nijboer F, Sellers EW, Mellinger J, et al. A P300-based brain-computer interface for people
with amyotrophic lateral sclerosis. Clin Neurophysiol. 2008;119(8):1909–16.
41. Lulé D, Noirhomme Q, Kleih S, et al. Probing command following in patients with disorders
of consciousness using a brain-computer interface. Clin Neurophysiol (Accepted).
42. Kubler A, Kotchoubey B, Hinterberger T, et al. The thought translation device: a neurophysi-
ological approach to communication in total motor paralysis. Exp Brain Res. 1999;124(2):
223–32.
43. Neuper C, Muller GR, Kubler A, et al. Clinical application of an EEG-based brain-computer
interface: a case study in a patient with severe motor impairment. Clin Neurophysiol. 2003;
114(3):399–409.
44. Perelmouter J, Kotchoubey B, Kübler A, et al. Language support program for thought transla-
tion devices. Automedica. 1999;18:67–84.
45. Pfurtscheller G, Muller-Putz GR, Schlogl A, et al. 15 years of BCI research at Graz University
of Technology: current projects. IEEE Trans Neural Syst Rehabil Eng. 2006;14(2):205–10.
46. Cruse D, Bekinschtein TA, Monti M, Owen AM. Detecting awareness in the vegetative state
with EEG. In: 16th annual meeting of the organization for human brain mapping. Barcelona:
Elsevier; 2010.
47. Birbaumer N, Ghanayim N, Hinterberger T, et al. A spelling device for the paralysed. Nature.
1999;398(6725):297–8.
48. Birbaumer N, Kubler A, Ghanayim N, et al. The thought translation device (TTD) for com-
pletely paralyzed patients. IEEE Trans Rehabil Eng. 2000;8(2):190–3.
49. Monti MM, Vanhaudenhuyse A, Coleman MR, et al. Willful modulation of brain activity in
disorders of consciousness. N Engl J Med. 2010;362:579–89.
50. Naito M, Michioka Y, Ozawa K, et al. A communication means for totally locked-in ALS
patients based on changes in cerebral blood volume measured with near-infrared light. IEICE
Trans Inf Syst. 2007;E90-D(7):1028–37.
51. Kennedy PR, Bakay RA. Restoration of neural output from a paralyzed patient by a direct
brain connection. Neuroreport. 1998;9(8):1707–11.
52. Kennedy PR, Bakay RA, Moore MM, et al. Direct control of a computer from the human
central nervous system. IEEE Trans Rehabil Eng. 2000;8(2):198–202.
53. Brumberg JS, Nieto-Castanon A, Kennedy PR, Guenther FH. Brain-computer interfaces for
speech communication. Speech Commun. 2010;52(4):367–79.
54. Hinterberger T, Widman G, Lal TN, et al. Voluntary brain regulation and communication with
electrocorticogram signals. Epilepsy Behav. 2008;13(2):300–6.
55. Leuthardt EC, Schalk G, Wolpaw JR, et al. A brain-computer interface using electrocortico-
graphic signals in humans. J Neural Eng. 2004;1(2):63–71.
56. Hill NJ, Lal TN, Schroder M, et al. Classifying EEG and ECoG signals without subject training
for fast BCI implementation: comparison of nonparalyzed and completely paralyzed subjects.
IEEE Trans Neural Syst Rehabil Eng. 2006;14(2):183–6.
57. Blankertz B, Sannelli C, Halder S, et al. Neurophysiological predictor of SMR-based BCI
performance. Neuroimage. 2010;51(4):1303–9.
58. Giacino J, Ashwal S, Childs N, et al. The minimally conscious state: definition and diagnostic
criteria. Neurology. 2002;58(3):349–53.
Chapter 8
Using Transcranial Magnetic Stimulation
to Measure Cerebral Connectivity in Patients
with Disorders of Consciousness

Melanie Boly, Olivia Gosseries, Marie-Aurélie Bruno, Mario Rosanova,

and Marcello Massimini

Assessment of the Level of Consciousness in Noncommunicative

Brain Damaged Patients

The clinical assessment of patients with disorders of consciousness is based on

repeated behavioral observations performed by trained clinicians. During the exam,
spontaneous movements of the patient, or her/his responses to multisensory stimu-
lations are studied and assessed according to specialized behavioral scales [1–4]. In
this context, the clinician searches typically for (1) signs of consciousness of self or
the environment; (2) sustainable, reproducible, or voluntary responses to tactile,
auditory, or noxious stimuli; and (3) signs of language comprehension or expression
[5]. If none of these behavioral characteristics can be detected during cautious and
repeated examinations, the subject is considered unconscious [6], while patients
showing nonreflexive behaviors but remaining unable to communicate are considered
to be in minimally conscious state [7]. Thus, according to clinical definition of
consciousness, the subjects are conscious if they can show they are. However, as
motor functions are often impaired in severely brain damaged patients, it can also
happen that a subject is conscious but unable to move or speak [8]. For this reason,
while the detection of voluntary signs at the patient’s bedside is considered sufficient
to infer that the subject is conscious, their absence does not necessarily imply the
absence of conscious perception [5, 9, 10].

M. Boly, M.D., Ph.D. (*) • M.-A. Bruno, Ph.D. • O. Gosseries, M.Sc.

Coma Science Group, Research Cyclotron Center, University of Liège,
Allée du 6 Août, Sart Tilman B30, 4000 Liège, Belgium
e-mail: [email protected]; [email protected]; [email protected]
M. Rosanova, Ph.D. • M. Massimini, Ph.D.
Physiology Department, University of Milan,
Via G.B. Grassi, 74, 20157 Milan, Italy
e-mail: [email protected]; [email protected]

C. Schnakers, S. Laureys (eds.), Coma and Disorders of Consciousness, 79

DOI 10.1007/978-1-4471-2440-5_8, © Springer-Verlag London 2012
80 M. Boly et al.

Recently, the development of new functional neuroimaging paradigms allowed

detecting signs of consciousness in patients completely unable to move [9]. For
example, in a recent study [11], a patient in clinical vegetative state, seemingly not
conscious, was scanned with the instruction to imagine playing tennis or wandering
around her home. Remarkably, the patient showed a functional MRI activation simi-
lar to that found in healthy volunteers while instructed to perform these tasks. This
case demonstrates that this patient was able to follow commands using thoughts,
even if a motor response was absent. However, there are many cases where a patient
could not show positive response to this paradigm, and nevertheless, be conscious.
For example, the presence of aphasia, akinetic mutism, catatonic depression, or dif-
fuse dopaminergic lesions could compromise this type of response to command,
despite the presence of cognition [9]. Moreover, because of frequent movement
artifacts and of possible alterations of neurovascular coupling, the acquisition and
interpretation of functional MRI data in patients with disorders of consciousness are
particularly challenging [12].
For these reasons, the absence of voluntary cerebral activity in the functional MRI
scanner, as the absence of clinical response to command, does not imply an absence
of consciousness. The behavioral approach and the active functional imaging para-
digms represent two levels where a communication can be established with patients
with disorders of consciousness and where a diagnosis of preserved consciousness
can be performed. However, in the case of a negative result, these two methods leave
the question of the presence or absence of consciousness unanswered.

Using Transcranial Magnetic Stimulation and EEG to Assess

the Brain’s Capacity to Generate Consciousness

We recently proposed an additional level where consciousness can be studied even

if the subject is unable to communicate, clinically or via neural activity [13]. This
paradigm aims at directly assessing the ability of the subject’s brain to generate
conscious perception, by assessing some properties of cerebral activity, which are
theoretically pertinent in this context. This possibility requires (1) starting from a
theory that suggests which properties are fundamental in the organization of a sys-
tem to generate conscious experience (2) identifying and implementing a measure
which practically allows detecting these properties in the brain. Here, we will take
the example of integrated information theory [14], which stipulates that consciousness
is equivalent to integrated information and that a physical system should be capable
of generating conscious perception to the extent that it can enter in a large number
of different states (information), without being breakable into a collection of inde-
pendent subsystems (integration). Then, a concrete method is set up, in order to
evaluate the brain’s ability to integrate information. To this aim, a combination of
transcranial magnetic stimulation and EEG is used (TMS-EEG). This technique
allows to directly stimulate different assemblies of cortical neurons and record the
immediate response of the rest of the brain. This method has been suggested to be
8 Using Transcranial Magnetic Stimulation to Measure Cerebral Connectivity in Patients 81

an efficient way to appreciate, in a general level, in which extent different regions of

thalamocortical system can interact globally (integration) and produce specific
responses (information). Though this approach does not aim at asserting if a subject
is really conscious or not, it may represent an objective way to evaluate brain ability
to generate conscious perception.

Validation Studies Using This Technique

Before being applied to patients, TMS-EEG technique requires some validation

works in different altered states of consciousness, in order to check its predictive
value for detecting the level of consciousness of the subjects. Promising studies
were performed during sleep and anesthesia, testing the theoretical predictions
underlying this technique. These studies showed that during these diminished con-
sciousness states, brain responses indeed lose the characteristics commonly
observed during wakefulness. The first study was performed during slow-wave
sleep, a state where the conscious content of the subjects is known to be greatly
reduced. In contrary to the rich connectivity observed during wakefulness, brain
responses to TMS are typically local, and a loss of integration is observed [15]
(Fig. 8.1). If a more widespread response is observed, it will be stereotypical, prop-
agating homogeneously in the brain, and losing all the complexity observed during
wakefulness [16].
A second study [17] was recently performed during deep sedation obtained using
the pharmacological agent midazolam. This study confirmed that when subjects
lose consciousness, the response to TMS is very different from that observed during
wakefulness, with typically local and stereotypical response, mimicking a local
slow wave, which does not propagate to the rest of the cortex. Brain connectivity is
then dramatically reduced, particularly in long-distance connections, and the
response does not persist more than a few tens of milliseconds – in contrast to the
more complex and long lasting response observed during wakefulness. This result
suggests that loss of consciousness during anesthesia may also be linked to a loss of
brain integration. Moreover, the stereotypical aspect of the response suggests a loss
of brain areas specialization, i.e., a loss of information [18].
During sleep as under anesthesia, the homogenous and stereotypical response
observed after TMS administration is thought to be generated not only due to a loss
of cerebral integration, but also to a disappearance of the specialization of the
different brain areas (loss of information), due to the presence of bistable dynamics.
Bistable dynamics are generated in a system when the neurons reach a certain
threshold of hyperpolarization. In contrary to the rich diversity of responses that can
be obtained during wakefulness, the occurrence of this type of dynamics will induce
a stereotypical behavior, with only all or nonresponses following a perturbation of
the system. This type of dynamics has been suggested to be particularly unfavorable
to the generation of integrated information in the brain, and to the generation of
conscious perception [14, 19, 20].
82 M. Boly et al.

Wakefulness Slow wave sleep

a BA 8 c
BA 8

BA 6 BA 6

BA 19 BA 19

0.1 µA/mm2

0.1 µA/mm2
0 4 x 10-4 0 4 x 10−4
µA / mm2 100 ms µA/mm2 100 ms

b d
BA 8 BA 8

BA 6 BA 6

BA 19 BA 19

Fig. 8.1 Loss of integration and differentiation during slow-wave sleep. TMS is applied to frontal
and occipital cortex during wakefulness (a, b) and slow-wave sleep (c, d). After source reconstruc-
tion, nonparametric statistics are applied in order to detect currents significantly evoked by the
response to TMS. For each condition, significant currents recorded during the whole poststimula-
tion period are displayed on cortical surface. On the right of each cortical surface, the temporal
profiles of signals recorded in three brain areas (Brodmann areas 8, 6, and 19) are displayed (the
stimulation time being shown as a red mark). During transition from wakefulness to slow-wave
sleep, distant cortical areas cease to be influenced by the stimulation, suggesting a loss of brain
integration. Moreover, brain responses to TMS become stereotypical, suggesting a loss of brain
activity differentiation (Reproduced with permission from Massimini et al. [13])

Perspectives

The TMS-EEG could in principle be applied to brain damaged patients to evaluate

their residual capacity to present a conscious experience. Beforehand, however, the
acquisition of normative data should ideally be realized in order to define the normal
or abnormal character of a given brain response, and to obtain quantitative measures
related to these normality criteria. Some work has already been performed in this
direction [21, 22], but it should be extended to a larger population and a larger num-
ber of brain areas in order to be more generally predictive. Second, a standardized
analysis procedure should be set up in order to extract from TMS-EEG data some
indexes, capturing in a synthetic manner the brain’s ability to generate integrated
information. For example, indexes using the entropy of the TMS-evoked response,
or its algorithmic complexity could be useful tools in order to summarize the brain’s
ability to produce responses that are both integrated and specific (information). This
8 Using Transcranial Magnetic Stimulation to Measure Cerebral Connectivity in Patients 83

type of measure would be low for a modular system, because in this case, the activ-
ity would remain local; it would also be low for networks with widespread but
homogeneous connectivity, because all the elements would respond in the same
way. In theory, only the networks that are both integrated and differentiated could
react to TMS with a response presenting a high complexity. Before applying this
technique for the diagnosis of brain damaged patients, other studies should be per-
formed using other anesthetic agents (e.g., propofol, ketamine, xenon, sevoflurane)
and different levels of sedation, in order to extend previously obtained results.
Indeed, a reliable marker of consciousness level should behave in the same way in
various altered states of consciousness, the etiology of which being physiological
(sleep), pharmacological (anesthesia), or pathological (coma, vegetative state and
related disorders). Even if the preliminary studies show very encouraging results,
these findings, thus, should be validated on a larger number of subjects and condi-
tions before the application of this technique for clinical diagnosis.

References

1. Giacino JT, Kalmar K, Whyte J. The JFK coma recovery scale-revised: measurement character-
istics and diagnostic utility. Arch Phys Med Rehabil. 2004;85(12):2020–9.
2. Gill-Thwaites H, Munday R. The sensory modality assessment and rehabilitation technique
(SMART): a valid and reliable assessment for vegetative state and minimally conscious state
patients. Brain Inj. 2004;18(12):1255–69.
3. Kalmar K, Giacino JT. The JFK coma recovery scale–revised. Neuropsychol Rehabil.
2005;15(3–4):454–60.
4. Shiel A, Horn SA, Wilson BA, et al. The wessex head injury matrix (WHIM) main scale: a
preliminary report on a scale to assess and monitor patient recovery after severe head injury.
Clin Rehabil. 2000;14(4):408–16.
5. Laureys S, Owen AM, Schiff ND. Brain function in coma, vegetative state, and related disor-
ders. Lancet Neurol. 2004;3(9):537–46.
6. The Multi-Society Task Force on PVS. Medical aspects of the persistent vegetative state (1). N
Engl J Med. 1994;330(21):1499–508.
7. Giacino JT, Ashwal S, Childs N, et al. The minimally conscious state: definition and diagnostic
criteria. Neurology. 2002;58(3):349–53.
8. Schnakers C, Perrin F, Schabus M, et al. Detecting consciousness in a total locked-in syndrome:
an active event-related paradigm. Neurocase. 2009;15(4):271–7.
9. Boly M, Coleman MR, Davis MH, et al. When thoughts become action: an fMRI paradigm to
study volitional brain activity in non-communicative brain injured patients. Neuroimage.
2007;36(3):979–92.
10. Monti MM, Vanhaudenhuyse A, Coleman MR, et al. Willful modulation of brain activity in
disorders of consciousness. N Engl J Med. 2010;362(7):579–89.
11. Owen AM, Coleman MR, Boly M, et al. Detecting awareness in the vegetative state. Science.
2006;313(5792):1402.
12. Giacino JT, Hirsch J, Schiff N, Laureys S. Functional neuroimaging applications for assess-
ment and rehabilitation planning in patients with disorders of consciousness. Arch Phys Med
Rehabil. 2006;87(12):67–76.
13. Massimini M, Boly M, Casali A, et al. A perturbational approach for evaluating the brain’s
capacity for consciousness. Prog Brain Res. 2009;177:201–14.
14. Tononi G. Consciousness as integrated information: a provisional manifesto. Biol Bull.
2008;215(3):216–42.
84 M. Boly et al.

15. Massimini M, Ferrarelli F, Huber R, et al. Breakdown of cortical effective connectivity during
sleep. Science. 2005;309(5744):2228–32.
16. Massimini M, Ferrarelli F, Esser SK, et al. Triggering sleep slow waves by transcranial mag-
netic stimulation. Proc Natl Acad Sci USA. 2007;104(20):8496–501.
17. Ferrarelli F, Massimini M, Sarasso S, et al. Breakdown in cortical effective connectivity during
midazolam-induced loss of consciousness. Proc Natl Acad Sci USA. 2010;107(6):2681–6.
18. Alkire MT, Hudetz AG, Tononi G. Consciousness and anesthesia. Science. 2008;322(5903):
876–80.
19. Tononi G, Massimini M. Why does consciousness fade in early sleep? Ann N Y Acad Sci.
2008;1129:330–4.
20. Boly M, Massimini M, Tononi G. Theoretical approaches to the diagnosis of altered states of
consciousness. Prog Brain Res. 2009;177:383–98.
21. Casarotto S, Romero Lauro LJ, Bellina V, et al. EEG responses to TMS are sensitive to changes
in the perturbation parameters and repeatable over time. PLoS One. 2010;5(4):10281.
22. Rosanova M, Casali A, Bellina V, et al. Natural frequencies of human corticothalamic circuits.
J Neurosci. 2009;29(24):7679–85.
Chapter 9
Sleep in Disorders of Consciousness

Victor Cologan and Manuel Schabus

Introduction

Today consciousness is considered as multifaceted concept having two major compo-

nents: arousal and awareness. Arousal refers to the level of consciousness (i.e., vigi-
lance) and is supported by brainstem neuronal populations projecting to both thalamic
and cortical areas. Awareness refers to the content of consciousness (i.e., awareness
of the self and of the environment) and relies on the functional integrity of the cere-
bral cortex and its subcortical connections. Both components progressively decrease
as sleep deepens. In some pathological situations, however, arousal and awareness are
dissociated. This, for instance, is the case of patients in coma, vegetative state (VS)
(today preferentially referred to as unresponsive wakefulness syndrome, UWS), or
minimally conscious state (MCS). The question thus arises whether and in which
ways sleep in severely brain-injured patients is being altered. It has been proposed
that besides “offline” neuronal plasticity, one of the major functions of sleep is to
restore general homeostasis (including, more specifically, subcortical brain struc-
tures) and to stabilize neuronal connections. Accordingly, alterations in sleep in DOC
(Disorders of Consciousness) might be associated to a lower probability of recovery.

V. Cologan, M.Sc.
Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du 6 Août, Sart Tilman B30, 4000 Liège, Belgium
e-mail: [email protected]
M. Schabus, Ph.D. (*)
Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du 6 Août, Sart Tilman B30, 4000 Liège, Belgium
Laboratory for Sleep and Consciousness Research, Department of Psychology,
University of Salzburg, Hellbrunnerstr. 34, Salzburg 5020 Austria
e-mail: [email protected]

C. Schnakers, S. Laureys (eds.), Coma and Disorders of Consciousness, 85

DOI 10.1007/978-1-4471-2440-5_9, © Springer-Verlag London 2012
86 V. Cologan and M. Schabus

Polysomnography (PSG), which is the only valid sleep recording method, can be
used to perform a clinical assessment of that kind. In particular, it provides useful
information for the characterization of altered brain activity and prognosis. Most of
the present research has been conducted on the acute phase of coma (i.e., immedi-
ately following brain injury) or on patients having already recovered consciousness.
Little is known about the transitional states leading to recovery (VS/UWS, MCS)
and other clinical entities such as locked-in syndrome (LIS) or brain death.
Sleep abnormalities are common in critically ill patients. In brain-injured patients,
about half of the total sleep time occurs during the daytime and the circadian rhythm
is often significantly diminished or even lost. Generally, these patients exhibit more
frequent arousals and awakenings than healthy volunteers, and decreases in rapid
eye movement (REM) sleep and deep non-REM sleep are common. Additionally to
brain injury, considerable sleep disruption also arises from hospital environments
throughout day and night. This is worth mentioning as sleep deprivation is known
to have major and devastating impacts on mood, daytime fatigue, residual cognitive
functioning, and the immune system, especially critical for brain-injured patients.

Studying Sleep in DOC

Although it is well-known that sleep abnormalities are common in critically ill

patients, their mechanisms remain poorly understood and their fine-grained charac-
terizations are still unknown [1]. First of all, one has to characterize the differences
found in DOC patients as compared to those in healthy individuals. From a behav-
ioral point of view, normal sleep is usually preceded by the search for a safe place, a
progressive but reversible decrease of responsiveness to external stimuli and a
decrease of motor output. In DOC, assessing these behavioral criteria is challenging
and rarely conclusive. From a physiological point of view, normal sleep is associated
with well-described cycles, stages, arousals, and microstructures (e.g., K-complexes,
spindles). Yet the traditional scoring manual from Rechtschaffen and Kales [2] is
often not appropriate for clinical contexts as summarized elsewhere [3]. In DOC, the
classification of sleep and the existence of polysomnographic elements (of “unusual”
shape or frequency characteristic), therefore, remains a matter of scientific debate.
Together with evidence of eye opening, the presence of a sleep–wake cycle defines
the threshold for the progression from a comatose state to VS/UWS or MCS.
However, despite the importance of a sleep–wake cycle for differential diagnosis,
there is little empirical evidence that DOC patients actually exhibit classical sleep
phenomena or display a circadian rhythm. Sleep–wake cycles are typically indirectly
inferred by behavioral observations with prolonged periods of opened or closed eyes.
Yet, as circadian rhythms are endogenously generated by a biological “Zeitgeber”
located in the hypothalamic suprachiasmatic nucleus (SCN), it would be needed that
more direct measures, such as the SCN controlled pineal hormone melatonin, are
adopted. Note that a handful of studies conducted in intensive care units indicate the
loss of circadian secretion of melatonin in sedated unresponsive patients [4].
9 Sleep in Disorders of Consciousness 87

Past studies reported day–night brain state differences in some but not other VS
patients using continuous EEG [5]. Furthermore, it was shown that heart rate and
blood pressure do not show the typical nocturnal decrease during night hours in
traumatic VS [6]. However, an inherent problem to studies of that kind is daytime
sleep, which decreases the amount of sleep during the nightly recording times.
Another study in persistent VS patients showed significant circadian changes in
body temperature and urinary excretion of hormones and sodium but revealed no
changes in blood pressure or pulse rate [7]. In a more recent study by Bekinschtein
and colleagues [8] it was reported that while VS/UWS patients with traumatic brain
injury exhibit well-formed circadian temperature rhythms, patients with anoxic-
hypoxic origin demonstrate no cycles or rhythmic behavior.
These findings suggest that the preservation of circadian rhythm may not only
vary with the patients’ state of consciousness, but also the extent of their individual
brain damage.
It appears that the existence of sleep and its characterization in DOC is a most
challenging issue as these patients do not show the normal behavioral, physiologi-
cal, and regulatory signs of sleep. Circadian and homeostatic factors are potent
modulators of sleep and it is, therefore, surprising that previous studies did not
investigate in greater detail their existence and possible diagnostic and prognostic
value before focusing on macro- or microstructural sleep changes. To this end, the
detection of changes in hormonal plasma levels (e.g., melatonin, cortisol), tempera-
ture, along with those of blood pressure, heart rate, and actimetry monitoring should
be performed first to accurately outline circadian variations in DOC.
The remaining question is the assessment of residual homeostasis in DOC.
Unlike sleep patterns, homeostasis is rarely addressed in DOC. Support for the pres-
ence of homeostatic sleep regulation in DOC could be provided by simple wake
extension protocols and by comparing the EEG before and after a period during
which patients are prevented from sleep, e.g., behaviorally by stimulating externally
during prolonged eye closure. A recent study by Landsness and colleagues [9] dem-
onstrated, besides regular alternating NREM/REM sleep patterns, a clear homeo-
static decline of EEG slow-wave activity across the night in all tested MCS but not
VS/UWS patients. The study of homeostatic regulation of slow-wave activity might,
therefore, provide a complementary tool of brain function for the near future.
Finally, previous studies on coma suggested that the presence of certain EEG
patterns may be reliable markers for favorable outcome [10, 11]. It was reported that
sleep patterns continue to become more “complex” (i.e., NREM–REM alterations,
reappearance of sleep spindles, etc.) during rehabilitation therapy, sometimes even
in parallel to patients’ cognitive recovery [12]. Some authors have used Rechtschaffen
and Kales sleep criteria to analyze the polysomnographic data in DOC [11, 13].
However, as many forms of brain damage may result in a relatively similar clinical
state of unconsciousness and as modification of brain activity in DOC may substan-
tially differ from physiological sleep patterns described by Rechtschaffen and Kales,
those criteria are hardly applicable for scoring sleep in severely brain damaged
patients. These scoring criteria do need update and revision in order to become fully
suitable for sleep–wake assessment in DOC.
88 V. Cologan and M. Schabus

In conclusion, the study of sleep is of particular interest in DOC with various

different etiologies as it might allow to identify relationships between specific PSG
graphoelements and underlying functional neuroanatomy. Specifically, NREM
sleep spindles and slow-wave sleep may reflect the preserved functional integrity
of the thalamocortical network; present REM sleep may reflect residual function-
ing of brainstem nuclei including pons and adjacent portions of the midbrain; and
last but not least the integrity of circadian organization of sleep–wake patterns
might be informative of residual hypothalamic and suprachiasmatic nucleus (SCN)
functioning.

Methodological Issues

In intensive care units, environmental factors are often not optimal for good sleep
[1]. For example, not perfectly adjusted mechanical ventilation may cause significant
sleep disruption. Further factors contributing to sleep abnormalities in intensive
care units include acute pain, discomfort, increased alertness induced by continuous
exposure to light or noise, as well as nutrition and nursing activities. In this context,
important to note is the fact that sleep deprivation is known to have negative impact
on immune and endocrine functions. It also induces sympathetic activation and
elevation of blood pressure, which in turn can increase morbidity. It is, therefore,
suggested that in order to support optimal recovery and promote good neurobehav-
ioral functioning, clinicians should be aware of these basic facts.
PSG is the only tool allowing reliable evaluation of sleep. Yet, recording good
quality signals in DOC is very challenging due to artifacts caused by strong perspi-
ration, thermal dysregulation, skin and skull lesions, or electrical artifacts from life-
supporting devices. When looking at the basic EEG structure, the first examination
to make is to test for reactivity, i.e., to check for a change in EEG caused by external
sensory stimulation. Painful or auditory stimuli are usually used for this purpose
and the presence of EEG reactivity often indicates a less severe disorder [14].
Another evaluation, although rarely used in comatose patients, concerns arousal
organization during sleep. Arousals can be defined as simultaneous changes of EEG
along with autonomic system or somatic activity [15]. Compared to healthy indi-
viduals, DOC patients do not exhibit normal arousal alternations or cyclic alternat-
ing patterns (CAPs) [16]. In these patients, changes in brain activation may be
considerably slowed, lasting a number of seconds or even minutes, and are often not
rhythmic. Arousal alternations are often more extreme than normal and may even
become life-threatening, especially those occurring in the vegetative system (e.g.,
involving cerebrospinal fluid pressure increase) [17]. These extreme sleep micro-
structure changes seen in DOC indicate that arousal control mechanisms have been
injured or severely altered.
The advantage of EEG recordings over functional neuroimaging includes its
high temporal resolution and its noninvasive nature, along with its portability and
low cost. Its main limitation is the spatial resolution and current lack of specificity
9 Sleep in Disorders of Consciousness 89

(diffuse slowing of background rhythms is, e.g., seen in various encephalopathies

regardless of the etiology). High-density EEG is one current development that
will allow better spatial source reconstruction, yet the spatial resolution of func-
tional magnetic resonance imaging (fMRI) including the assessment of deep
brain structures remains unattainable. It is also to be noted that EEG is often
heavily contaminated by uncontrolled eye movements and muscle activation
occurring in DOC. Sophisticated correction methods, e.g., various independent
component analysis (ICA) algorithms, may be necessary in order to obtain clean
EEG data exploitable for in-depth analysis and scientific as well as clinical
interpretation.

Coma

Prevalent alpha or (alpha-coma, AC), spindle-like activity (spindle-coma, SC),

and triphasic waves are the most typical EEG patterns observed in coma patients
[14]. In AC, alpha activity between 8 and 12 Hz dominates the frequency spec-
trum and presents an usual widespread topography with often high amplitude at
frontal electrodes and no reactivity to eye opening as usually observed (i.e., no
posterior block of alpha oscillations). Alpha coma patterns have been described
after hypoxia as well as drug intoxications. SC is a chronic electrophysiological
pattern in which physiological sleep components, specifically sleep spindles
(9–14 Hz) are superimposed on a background of delta (0.9–4 Hz) and theta
(4–8 Hz) activity. Triphasic waves consist in bursts of moderate to high amplitude
(100–300 mV) activity, usually of 1.5–2.5 Hz, and frequently predominating in
frontal regions. The initial negative component is the sharpest, whereas the fol-
lowing positive portion of the complex is the largest and is followed by another
negative wave. Triphasic waves are usually bisynchronous but may show shifting
asymmetries. Persistent asymmetry suggests an underlying structural lesion on
the side of the lower amplitude.
In humans, the pathophysiological mechanism of SC has been related to the
pontine raphe nuclei and thalamocortical circuits together with the impairment of
the ascending reticular activating system located in the midbrain, which usually
maintains arousal [18–20]. SC is considered to be a more “benign” condition than
AC, and is largely related to lower cerebral damages. Indeed, the overall mortality
in case of SC does not exceed 25% [21]. This quite good prognosis might be
explained by the preservation of thalamocortical functioning, which underlies spin-
dle generation [22].
Generally, EEG sleep patterns have been considered as favorable prognostic
markers in brain-injured patients [10, 11] and it has been shown that sleep patterns
continue to normalize during rehabilitation together with the recovery of cognitive
functioning [12]. The presence of organized sleep patterns is also predictive of
good outcome [23, 24]. In contrast, patients showing only monophasic EEG
(i.e., continuous low voltage theta-delta activity) or CAPs indicating arousal
90 V. Cologan and M. Schabus

a C4-A1 c C4-A1
C3-A2 C3-A2
P4-O2 P4-O2
P3-O1 P3-O1
fz-Cz Fz-Cz
ROC-A1 ROC-A1
LOC-AI LOC-AI
100 µV 100 µV
MYLO 1 sec CHIN 1 sec

b d C4-A1
C4-A1
C3-A2 C3-A2
P4-O2 P4-O2
P3-O1 P3-O1
Fz-Cz
ROC-A1
ROC-A1
LOC-AI
LOC-AI
CHIN
100 µV
5 sec MYLO 100 µV
CHIN
1 sec

Fig. 9.1 Polysomnographic recordings in subacute (7–14 days post onset) comatose patients. (a)
Panels depict a monophasic EEG with continuous low voltage theta-delta activity as well as (b)
recurrent abnormal (CAP) activity both indicative of bad outcome. The recovery of (c) NREM
sleep with well-structure spindles as well as the reappearance of REM sleep (d), on the other hand,
has been related to good outcome (Adapted from [24])

instability during NREM sleep, in the absence of classical sleep elements, have
been associated with bad outcome (death or severe disabilities). Especially REM
sleep elements (rapid ocular movements, muscle twitches, saw-tooth waves), alter-
nating regularly with NREM sleep elements as well as certain NREM sleep ele-
ments such as K-complexes and sleep spindles themselves, have been related to
good outcome (full recovery or only mild disability). Figure 9.1 illustrates some of
the key patterns discussed above (modified from [24]).
The reappearance of regular sleep patterns and specifically sleep spindles is of
good prognosis in comatose patients. While the absence of spindles in coma pre-
sumably results from the interruption of either the ascending reticulothalamocorti-
cal pathway or thalamocortical loops, the absence of sleep–wake cycles, which is
also typical in comatose patients, is associated with brainstem dysfunction.
These sleep patterns are associated to functional outcome as they allow a glimpse
into the residual connectivity of the lesioned brain. For example, it has been sug-
gested that the amount or intensity of sleep spindles reflects intact and efficiently
connected thalamocortical networks, as supported by recent data associating high
general cognitive and memory abilities to high spindle activity during the night [25,
26]. In conclusion, the magnitude of observed sleep alterations seems to be related
to injury severity.
9 Sleep in Disorders of Consciousness 91

“Vegetative State” / “Unresponsive Wakefulness Syndrome”

Patients in VS (today increasingly termed UWS) are unresponsive to external stim-

uli and are usually considered to be unconscious and unaware. Patients in this state
have a total loss of forebrain function but preservation of brainstem functions includ-
ing breathing, swallowing, or cranial nerve reflexes. Occasionally, VS patients
exhibit electrocerebral inactivity (i.e., no cerebral activity over 2 mV), although the
majority show some EEG activity. Unlike comatose patients, VS patients open and
close their eyes and thereby exhibit–at least at a behavioral level–behavior that
resembles normal sleep–wake cycling.
Patients in the full VS (formerly called apallic syndrome) often show severe
sleep fragmentation, which is likely caused by structural changes in brain areas
responsible for sleep maintenance, as well as absent sleep spindling [27, 28].
However, patients close to remission often show reappearance of NREM and REM
sleep stages, together with an increase in total sleep time, as well as sleep spindle
activity [29]. With regards to REM sleep, these patients exhibit significantly less
phasic REM events such as rapid eye movements, chin and leg muscle twitches
[13]. Yet, preserved sleep-related erections similar to those in normal subjects [30]
and nystagmus (which is an involuntary and jerky oscillatory movement of the eyes
different from that observed in REM sleep) [31] are reported in the literature.
It is discussed that this reduction of phasic rapid eye movements and muscles
twitches during REM sleep may reflect damage to the cholinergic mechanisms of
the pedunculopontine tegmentum (PPT).

Recovered State

It appears that also brain-injured patients who recover consciousness (exit-MCS or

recovered state) still suffer from various sleep disturbances. Insomnia-like sleep
complaints (i.e., sleeplessness) are reported as much as hypersomnia-like com-
plaints (i.e., excessive amount of sleepiness) and are probably physiologically linked
to structural changes in brain arousal systems [32–36]. Frequently, as seen in nor-
mal insomnia patients, DOC patients do overestimate the degree of their sleep dis-
turbance as revealed when comparing subjective (sleep diaries) and objective
(actigraphy, polysomnography) sleep data [37]. Beside these alterations, sleep pat-
tern and sleep onset latency modifications are common, with deep NREM sleep
being most preserved [38–41].
Fatigue, mood disturbances, and exacerbated cognitive deficits as well as
weakening of immune function are among the consequences of chronic sleep
alterations in brain injured patients. Treatment attempts in this population appear
rather scarce. Although for short-term treatment of sleep disorders the prescrip-
tion of hypnotic medications may be an effective approach, long-term use of these
medications is problematic, probably even more so than in non-brain-injured indi-
viduals. Nonpharmacological treatments like psychotherapeutic intervention [42]
92 V. Cologan and M. Schabus

or instrumental conditioning of specific brain oscillations (neurofeedback) [43]

may offer an alternative, yet longer than normal intervention durations have to
be expected.

Locked-in Syndrome

The literature about sleep in LIS patients is especially scarce. Brain lesions in these
patients affect the ventropontine area sparing the pontine tegmentum. A concomi-
tant cerebellum lesion is frequent as there is a rupture of the basilar artery. The
observed spectrum of altered sleep in LIS can vary from almost normal sleep pat-
terns [44] to hyposomnia [45], disorganized NREM sleep [46–50], or complete
REM absence [46, 47, 50, 51]. This wide spectrum is most likely explained by dif-
ferences in the exact location and extent of causative lesions. Thus, LIS patients can
present in rare cases no major sleep abnormalities although severe neurological
deficits are evident. Specifically, it appears that the more extended the pontine
lesion, the more pronounced the sleep disturbance, especially affecting REM sleep.
Lesion severity increases in case of bilateral or dorsal extensions, and of pontine
tegmentum involvement, especially if the serotonin-releasing raphe nuclei of the
brainstem (medial portion of the reticular formation) are included.
As mentioned previously, careful attention must be given to the environmental
factors (light, noise, nightly nursing activities) in which sleep disorders emerge in
these patients. This is specifically important for LIS patients who are completely
aware of their environment and who likewise suffer from the consequences of inad-
equate sleep, yet they are (paralyzed and mute) “locked-in” their bodies.

Conclusion

In summary, DOCs are often linked to alterations in sleep architecture and its micro-
structure. In the comatose state, the patient by definition shows no eye-opening and no
sleep–wake cycling. However, certain NREM and REM graphoelements may emerge,
yet in an often unusual generalized form. In the VS/UWS, patients by definition show
transient periods of eye-opening and sleep–wake cycles. However, very few studies
provide empirical evidence of the residual sleep architecture. Most literature exists for
the recovery from coma and deals with the reappearance of NREM graphoelements
such as sleep spindles. In these patients hypersomnia and insomnia appear to be com-
mon phenomena, which ask to be appropriately addressed and treated in the future. It
has to be noted that older studies might have mixed the two clinical entities VS and
MCS as the latter entity was only defined in 2002 by Giacino.
In the LIS (as a special case outside the classical DOC spectrum), sleep altera-
tions are very variable and depend mainly on the extent and location of the brain-
stem lesion.
9 Sleep in Disorders of Consciousness 93

To date there are no clearly defined polysomnographic patterns that reliably

define each of the reviewed DOC states. In addition, there is very little evidence
indicating whether sleep in DOC is still under normal circadian and homeostatic
control. Future studies should, therefore, aim for long-term PSG recordings together
with circadian measures in well-documented patients. These tools might in the long
run help to further differentiate the various DOC states, and especially complement
the often difficult distinction between VS/UWS and MCS patients (for review also
see [52]). In addition, certain sleep patterns such as spindles seem predictive of later
outcome, yet longitudinal sleep studies are still needed in order to come up with
reliable outcome markers relevant for individual patients.
From a methodological point of view, it is true that studying sleep in DOC is a
challenging endeavor. Long-term PSG of DOC patients in clinical or rehabilitation
centers is difficult because of various artifacts arising from environmental and nurs-
ing activities or from clinical instability or mandatory medication conditions.
Another major problem is that PSG signals in DOC are often so much altered that
standard sleep criteria can hardly be used for sleep scoring.
The characterization of sleep and its relationships with arousal, homeostasis, and
circadian rhythmicity is essential in order to further understand DOC conditions and
might offer access to highly needed complementary diagnostic and prognostic
information in this specific patient population.

References

1. Parthasarathy S, Tobin MJ. Sleep in the intensive care unit. Intensive Care Med. 2004;
30(2):197–206.
2. Rechtschaffen A, Kales A. A manual of standardized terminology, techniques and scoring
system for sleep stages of human subjects. Bethesda: U.S. Department of Health, Education,
and Welfare; 1968. p. 12.
3. Himanen SL, Hasan J. Limitations of Rechtschaffen and Kales. Sleep Med Rev. 2000;4(2):
149–67.
4. Mundigler G, et al. Impaired circadian rhythm of melatonin secretion in sedated critically ill
patients with severe sepsis. Crit Care Med. 2002;30(3):536–40.
5. Isono M, et al. Sleep cycle in patients in a state of permanent unconsciousness. Brain Inj.
2002;16(8):705–12.
6. Pattoneri P, et al. Circadian blood pressure and heart rate changes in patients in a persistent
vegetative state after traumatic brain injury. J Clin Hypertens (Greenwich). 2005;7(12):734–9.
7. Fukudome Y, et al. Circadian blood pressure in patients in a persistent vegetative state. Am J
Physiol. 1996;270(5 Pt 2):R1109–14.
8. Bekinschtein TA, et al. Circadian rhythms in the vegetative state. Brain Inj. 2009;23(11):
915–9.
9. Landsness E, et al. Electrophysiological correlates of behavioural changes in vigilance in veg-
etative state and minimally conscious state. Brain. 2011;134(8):2222–32.
10. Bergamasco B, et al. The sleep cycle in coma: prognostic value. Electroencephalogr Clin
Neurophysiol. 1968;25(1):87.
11. Chatrian GE, White Jr LE, Daly D. Electroencephalographic patterns resembling those of
sleep in certain comatose states after injuries to the head. Electroencephalogr Clin Neurophysiol.
1963;15:272–80.
94 V. Cologan and M. Schabus

12. Ron S, et al. Time-related changes in the distribution of sleep stages in brain injured patients.
Electroencephalogr Clin Neurophysiol. 1980;48(4):432–41.
13. Oksenberg A, et al. Phasic activities of rapid eye movement sleep in vegetative state patients.
Sleep. 2001;24(6):703–6.
14. Brenner RP. The interpretation of the EEG in stupor and coma. Neurologist. 2005;11(5):
271–84.
15. Halasz P, et al. The nature of arousal in sleep. J Sleep Res. 2004;13(1):1–23.
16. Freedman NS, et al. Abnormal sleep/wake cycles and the effect of environmental noise on
sleep disruption in the intensive care unit. Am J Respir Crit Care Med. 2001;163(2):451–7.
17. Evans BM. What does brain damage tell us about the mechanisms of sleep? J R Soc Med.
2002;95(12):591–7.
18. Britt Jr CW, Raso E, Gerson LP. Spindle coma, secondary to primary traumatic midbrain hem-
orrhage. Electroencephalogr Clin Neurophysiol. 1980;49(3–4):406–8.
19. Britt Jr CW. Nontraumatic “spindle coma”: clinical, EEG, and prognostic features. Neurology.
1981;31(4):393–7.
20. Seet RC, Lim EC, Wilder-Smith EP. Spindle coma from acute midbrain infarction. Neurology.
2005;64(12):2159–60.
21. Kaplan PW, et al. Clinical correlates and prognosis in early spindle coma. Clin Neurophysiol.
2000;111(4):584–90.
22. Schabus M, et al. Hemodynamic cerebral correlates of sleep spindles during human non- REM
sleep. Proc Natl Acad Sci USA. 2007;104(32):13164–9.
23. Evans BM, Bartlett JR. Prediction of outcome in severe head injury based on recognition of
sleep related activity in the polygraphic electroencephalogram. J Neurol Neurosurg Psychiatry.
1995;59(1):17–25.
24. Valente M, et al. Sleep organization pattern as a prognostic marker at the subacute stage of
post-traumatic coma. Clin Neurophysiol. 2002;113(11):1798–805.
25. Schabus M, et al. Sleep spindle-related activity in the human EEG and its relation to general
cognitive and learning abilities. Eur J Neurosci. 2006;23(7):1738–46.
26. Bodizs R, et al. Prediction of general mental ability based on neural oscillation measures of
sleep. J Sleep Res. 2005;14(3):285–92.
27. Giubilei F, et al. Sleep abnormalities in traumatic apallic syndrome. J Neurol Neurosurg
Psychiatry. 1995;58(4):484–6.
28. D’Aleo G, et al. Sleep spindles in the initial stages of the vegetative state. Ital J Neurol Sci.
1994;15(7):347–51.
29. D’Aleo G, et al. Sleep in the last remission stages of vegetative state of traumatic nature. Funct
Neurol. 1994;9(4):189–92.
30. Oksenberg A, et al. Sleep-related erections in vegetative state patients. Sleep. 2000;23(7):
953–7.
31. Gordon CR, Oksenberg A. Spontaneous nystagmus across the sleep-wake cycle in vegetative
state patients. Electroencephalogr Clin Neurophysiol. 1993;86(2):132–7.
32. George B, Landau-Ferey J. Twelve months’ follow-up by night sleep EEG after recovery from
severe head trauma. Neurochirurgia (Stuttg). 1986;29(2):45–7.
33. Keshavan MS, Channabasavanna SM, Reddy GN. Post-traumatic psychiatric disturbances:
patterns and predictors of outcome. Br J Psychiatry. 1981;138:157–60.
34. Cohen M, et al. Temporally related changes of sleep complaints in traumatic brain injured
patients. J Neurol Neurosurg Psychiatry. 1992;55(4):313–5.
35. Clinchot DM, et al. Defining sleep disturbance after brain injury. Am J Phys Med Rehabil.
1998;77(4):291–5.
36. Fichtenberg NL, et al. Insomnia in a post-acute brain injury sample. Brain Inj. 2002;16(3):
197–206.
37. Ouellet MC, Savard J, Morin CM. Insomnia following traumatic brain injury: a review.
Neurorehabil Neural Repair. 2004;18(4):187–98.
38. Guilleminault C, et al. Hypersomnia after head-neck trauma: a medicolegal dilemma.
Neurology. 2000;54(3):653–9.
9 Sleep in Disorders of Consciousness 95

39. Masel BE, et al. Excessive daytime sleepiness in adults with brain injuries. Arch Phys Med
Rehabil. 2001;82(11):1526–32.
40. Ouellet MC, Beaulieu-Bonneau S, Morin CM. Insomnia in patients with traumatic brain
injury: frequency, characteristics, and risk factors. J Head Trauma Rehabil. 2006;21(3):
199–212.
41. Ouellet MC, Morin CM. Subjective and objective measures of insomnia in the context of
traumatic brain injury: a preliminary study. Sleep Med. 2006;7(6):486–97.
42. Ouellet MC, Morin CM. Cognitive behavioral therapy for insomnia associated with traumatic
brain injury: a single-case study. Arch Phys Med Rehabil. 2004;85(8):1298–302.
43. Hoedlmoser K, et al. Non-pharmacological alternatives for the treatment of insomnia—
instrumental EEG conditioning, a new alternative? In: Soriento YE, editor. Melatonin, sleep
and insomnia. New York: Nova; 2011. p. 69–101.
44. Oksenberg A, et al. Polysomnography in locked-in syndrome. Electroencephalogr Clin
Neurophysiol. 1991;78(4):314–7.
45. Guilleminault C, Cathala JP, Castaigne P. Effects of 5-hydroxytryptophan on sleep of a patient
with a brain-stem lesion. Electroencephalogr Clin Neurophysiol. 1973;34(2):177–84.
46. Markand ON, Dyken ML. Sleep abnormalities in patients with brain stem lesions. Neurology.
1976;26(8):769–76.
47. Cummings JL, Greenberg R. Sleep patterns in the “locked-in” syndrome. Electroencephalogr
Clin Neurophysiol. 1977;43(2):270–1.
48. Freemon FR, Salinas-Garcia RF, Ward JW. Sleep patterns in a patient with a brain stem infarc-
tion involving the raphe nucleus. Electroencephalogr Clin Neurophysiol. 1974;36(6):657–60.
49. Autret A, et al. A syndrome of REM and non-REM sleep reduction and lateral gaze paresis
after medial tegmental pontine stroke. Computed tomographic scans and anatomical correla-
tions in four patients. Arch Neurol. 1988;45(11):1236–42.
50. Tamura K, et al. Disturbances of the sleep-waking cycle in patients with vascular brain stem
lesions. Clin Electroencephalogr. 1983;14(1):35–46.
51. Lavie P, et al. Localized pontine lesion: nearly total absence of REM sleep. Neurology.
1984;34(1):118–20.
52. Cologan V, et al. Sleep in disorders of consciousness. Sleep Med Rev. 2010;14:97–105.
Chapter 10
Sensory Stimulation Program

Haibo Di and Caroline Schnakers

Introduction

Progress in intensive care has led to an increase in the number of patients who sur-
vive a severe brain lesion. Although the main part recovers in the first days follow-
ing the coma, some of these patients stay in a disorder of consciousness. The medical
care provided to these patients in neurology and neurorehabilitation units is usually
mainly palliative and minimizes complications rather than improving the patients’
recovery. Until now, few treatments were validated in patients with disorders of
consciousness. Sensory stimulation programs constitute the treatment most fre-
quently applied during the recovery of these patients. These programs are based on
the principles of brain plasticity and of sensory deprivation [1].
The idea according to which an injured brain has the capacity to reorganize itself
to compensate for the affected regions is broadly accepted for several years [2]. The
most famous case illustrating this phenomenon is the case of Terry Wallis. Indeed,
this patient remained for 19 years in a minimally conscious state after a traumatic
brain injury. His spectacular recovery may be explained by this principle of brain
plasticity. A study of Voss et al. revealed a neural change (mainly involving the
cuneus and the precuneus) in this patient having recovered a functional verbal com-
munication and motor activity [3]. These results stress the importance to develop
therapeutics to optimize the brain plasticity and, hence, the full consciousness

H. Di, Ph.D. (*)

Physiology Department, International Vegetative State and Consciousness Science Institute,
Hangzhou Normal University, Xuelin, 310036 Hangzhou, Zhejiang, China
e-mail: [email protected]
C. Schnakers, Ph.D.
Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du 6 Août, Sart Tilman B30, 4000 Liège, Belgium
e-mail: [email protected]

C. Schnakers, S. Laureys (eds.), Coma and Disorders of Consciousness, 97

DOI 10.1007/978-1-4471-2440-5_10, © Springer-Verlag London 2012
98 H. Di and C. Schnakers

recovery in severely brain-injured adults. Providing sensory stimulations (auditory,

verbal, visual, olfactory, tactile, and gustatory) to the patient could potentially stim-
ulate affected neural networks, accelerate the brain plasticity, and avoid a sensory
deprivation which could slow down the patient’s consciousness recovery [1]. In this
chapter, we will introduce the theoretical principles underlying these programs and
the studies assessing their clinical interest. We will also discuss the limitations of
this treatment and consider future directions for clinical research.

Theoretical Principles Related to Sensory Stimulation Programs

Sensory stimulation programs usually consist in presenting different types of environ-

mental stimuli to the patient in order to optimize the consciousness level. These pro-
grams tend to avoid a sensory deprivation and to improve behavioral responses [4].
According to Ansell and Keenan [5], these programs allow the patients to follow a struc-
tured stimulation, optimizing their aptitude to react and to respond to their environment.
These would constitute enriched environments that would potentially have an impact on
the structure and the functions of the brain and, hence, on the brain plasticity.

Brain Plasticity

Years ago, neurologists considered that, after a brain lesion, neurons cannot recover
anymore and the brain cannot become richer of new neurons. However, more recently,
an increasing number of studies have revealed the remarkable capacity of young as
well as old brain to adapt and transform itself according to the environment, to restruc-
ture connections between neurons in order to develop networks, which optimize the
circulation of information provided by the environment. The adult nervous system
would be malleable and, hence, sensitive to modifications and adaptations [6, 7].

Effects of an Enriched Environment on the Behavior,

the Structure, and the Functions of the Brain

Sensory stimulation programs can be considered as enriched environments. The

principle of “enriched environment” was used by researchers studying animal behav-
ior. These experiments consisted in using ample animal boxes containing various
objects stimulating curiosity and manipulation. It is one of the oldest approaches
used to show that experiences can change the brain on a neuroanatomical and bio-
chemical level [8, 9]. On a behavioral level, studies on animals allowed showing
memory improvements in enriched environment in the course of adult life and during
10 Sensory Stimulation Program 99

aging. At a neurobiological level, studies revealed that this environment leads to an

increase in the thickness of the cortex and to cell modifications. Several studies per-
formed in rats showed that their stay in enriched environments increases their capac-
ity of learning, memory as well as their capacity of adaptation in comparison to rats
staying in impoverished environments [10, 11]. Sensory stimulations would allow
changes in the structure and the functioning of the nervous system but also in the
behaviors that the individual can demonstrate to interact with his surroundings.

Previous Studies on Sensory Stimulation Programs

Considering these principles, the Institutes for the Achievement of Human Potential
(IAHP) have introduced sensory stimulation programs in the neurorehabilitation
field and, despite the lack of scientific evidence, have supported the idea, according
to which these programs could enhance the synaptic reinnervation and, therefore,
the recovery from coma in severely brain-injured patients [12].
Numerous studies investigated the interest of these sensory stimulations in
patients with disorders of consciousness. The most part seems to show a positive
effect of these programs on consciousness recovery [4, 12–21]. These results are,
however, often affected by methodological biases and must be considered with cau-
tion. Indeed, Lombardi et al. assessed the efficacy of sensory stimulation programs
in patients diagnosed as being in a coma or in a vegetative state [22] in performing
a systematic review of randomized and nonrandomized controlled trials published
from 1966 to 2002. This meta-analysis reported only three studies corresponding to
the authors’ criteria, the other ones being, for the most part, noncontrolled studies
or descriptive case reports.
In the first study, Mitchell (1990) divided 24 severely brain-injured patients in
two groups of 12 patients hospitalized in the same neurological unit (the control
group and the experimental group) [17]. The control group was matched to the
experimental group according to the age, the gender, the brain lesion, and the
Glasgow Coma Scale (GCS) score at the admission. The experimental treatment
involved visual, auditory, olfactory, tactile, gustatory, kinesthetic, and vestibular
stimulations. The sessions of stimulations started 4–12 days after the brain lesion.
The treatment lasted 1 h and was applied 1–2 times a day. The coma duration and
the GCS score were considered. In the second study, Kater (1989) stimulated 30
(traumatic) severely brain-injured patients at least 2 weeks after the brain lesion
[23]. The control group was also matched to the experimental group according to
the age, the gender, the type of lesion, and the score on the GCS [24]. In both
groups, the patients were divided in three subgroups based on the GCS score: deep
(GCS 3–6), moderate (GCS 7–10), and light (GCS 11–14). The experimental treat-
ment included the stimulation of six sensory modalities (i.e., visual, auditory, olfac-
tory, tactile, kinesthetic, and verbal). The treatment lasted 45 min and was applied
twice a day, 6 days a week during a period of 1–3 months, according to the patient’s
recovery. A behavioral assessment was performed before and after treatment. These
100 H. Di and C. Schnakers

two studies showed a higher recovery in the experimental group than in the control
group, suggesting a positive impact of the sensory stimulation programs on the
recovery of severely brain-injured patients. The third study conducted by Johnson
et al. (1993) included 14 traumatic patients (GCS <8). The experimental group (7
subjects) and the controls group (7 subjects) were matched according to the age and
the GCS score [25]. The treatment consisted in stimulating the five senses 20 min a
day during the entire stay in the intensive care. GCS scores, brainstem reflexes (e.g.,
oculo-cephalic and oculo-vestibular reflexes) as well as several physiological mea-
surements (e.g., skin conductance, respiratory and heart rates) were assessed 20 min
pre- and posttreatment. However, the authors did not find any significant change.
According to Lombardi et al., these studies do not allow confirming the efficacy
of this type of treatment for patients recovering from coma [22]. Indeed, beside an
insufficient description of the sensory stimulation program applied, their results are
contradictory and their methodology is different (e.g., type and duration of treat-
ment, assessment, and outcome measures), which does not allow the comparison of
studies. The authors conclude that there is no reliable evidence supporting the inter-
est of sensory stimulation programs in comatose or vegetative patients. Another bias
the authors did not consider is the part of improvements due to spontaneous recov-
ery. Indeed, these studies were mainly performed in the acute or subacute stage,
period during which a spontaneous recovery has the highest probability to appear.
None of these studies allows dissociating improvements due to the treatment from
improvements due to spontaneous recovery.
Since 2002, several studies have been published. They, nevertheless, present the
same methodological issues as described above [26–28]. Interestingly, Oh and Seo
[28] decided to investigate whether the improvements observed after a sensory stim-
ulation program exceeded spontaneous recovery. The sensory stimulation program
involved auditory, visual, olfactory, gustatory, tactile, and kinesthetic stimulations.
The consciousness level was assessed by using the GCS scores. An ABA paradigm
was used as the treatment was applied during 4 weeks (phase B) and compared to
the baselines (phase A). Results showed a significant fluctuation of the behavioral
responses according to the presence or the absence of the treatment—more complex
responses being observed in the presence of treatment [28]. This study suggests that
sensory stimulation programs would have an impact superior to spontaneous recov-
ery on the improvement of consciousness in patients recovering from coma.
However, this study included a small number of patients (n = 5) and was, therefore,
mainly descriptive.

Limitations and Perspectives

As regards methodological biases of previous studies, we cannot conclude the effec-

tive impact of sensory stimulation programs on the consciousness recovery. It
would, therefore, be needed to increase the quality of future studies, e.g., by consid-
ering the following points:
10 Sensory Stimulation Program 101

The Use of Valid and Sensitive Behavioral Scales

Some of the studies led on the efficacy of sensory stimulation programs do not use
valid behavioral scales, but physiological and electrophysiological measurements
[22]. However, the detection of physiological and electrophysiological changes is
pointless without the observation of clinical and behavioral improvements. The use
of a standardized and validated scale detecting such improvements is, therefore,
recommended, but is not sufficient as this scale also has to be sensitive. Previous
studies have shown that a scale such as the Coma Recovery Scale-Revised [29]
allows a better detection of signs of consciousness than the GCS, which is usually
used in studies on sensory stimulation programs [30, 31]. The choice of the behav-
ioral scale that will objectify patients’ improvements is nevertheless a critical point
as it constitutes an indicator of the effectiveness of the treatment.

Studies on Chronic Patients

Future studies should recruit patients being in stable clinical conditions, and there-
fore beyond the acute stage. Indeed, about 50% of brain-injured patients present a
spontaneous recovery during this period [32]. Besides, it is difficult to administer
rehabilitation programs in intensive care units where patients are often sedated.

An ABAB Design

This design is one of the oldest and most widely used time-series method developed
in behavioral psychology. It typically includes at least one phase of treatment (phase
B) alternating with baselines (“non-treatment” phase or phase A). In other words, it
allows measuring the effects of treatment as compared to what happens when the
treatment is absent or removed, and hence allows comforting that improvements
observed are due to the treatment and not to spontaneous recovery. Using such a
design may be even more interesting than using a controlled study, as matching
patients can be highly problematic as regards the diversity of ages, etiologies, and
brain lesions. As mentioned before, this design was successfully used in one study
including five patients [28].

Multicentric Studies

The sample size should be sufficient to demonstrate the clinical significance of the
treatment. It is, therefore, reasonable to suggest the implementation of multicentric
studies in future clinical researches.
102 H. Di and C. Schnakers

Conclusion

Taking care of patients recovering from coma is uneasy, current therapeutics being
neither well-developed nor well-validated. Sensory stimulations programs is the
most widely known treatment applied to severely brain-injured patients. Until now,
no study permitted to demonstrate its clinical interest. Demonstrating the efficacy of
treatments is, nevertheless, a necessary step to the improvement of the quality of life
of such patients. Further investigations addressing this question are, therefore,
needed in the future.

References

1. Wood R. Critical analysis of the concept of sensory stimulation for patients in vegetative states.
Brain Inj. 1991;5(4):401–9.
2. Hummel F, Cohen L. Drivers of brain plasticity. Curr Opin Neurol. 2005;18(6):667–74.
3. Voss H, Uluc A, Dyke J, et al. Possible axonal regrowth in late recovery from the minimally
conscious state. J Clin Invest. 2006;116(7):2005–11.
4. Jones R, Hux K, Morton-Anderson K, Knepper L. Auditory stimulation effect on a comatose
survivor of traumatic brain injury. Arch Phys Med Rehabil. 1994;75(2):164–71.
5. Ansell B, Keenan J. The Western Neuro Sensory Stimulation Profile: a tool for assessing slow-
to-recover head-injured patients. Arch Phys Med Rehabil. 1989;70:104–8.
6. Merabet L, Pascual-Leone A. Neural reorganization following sensory loss: the opportunity of
change. Nat Rev Neurosci. 2010;11(1):44–52.
7. Barnes S, Finnerty G. Sensory experience and cortical rewiring. Neuroscientist. 2010;16(2):
186–98.
8. Diamond M. Enriching heredity. New York: The Free Press; 1988.
9. Renner M, Rosenzweig M. Enriched and improverished environments: effects on brain and
behaviour. New York: Springer; 1987.
10. Nithianantharajah J, Hannan J. Enriched environments, experience-dependent plasticity and
disorders of the nervous system. Nat Rev Neurosci. 2006;7(9):697–709.
11. Sale A, Berardi N, Maffei L. Enrich the environment to empower the brain. Trends Neurosci.
2009;32(4):233–9.
12. LeWinn E, Dimancescu M. Environmental deprivation and enrichment in coma. Lancet.
1978;2:156–7.
13. Blackerly WF. Intensity of rehabilitation and length of stay. Brain Inj. 1990;4(2):167–73.
14. Doman G, Wilkinson R, Dimancescu M, Pelligra R. The effects of intense multi-sensory stim-
ulation on coma arousal and recovery. Neuropsych Rehabil. 1993;3(2):203–12.
15. Hall M, MacDonald S, Young G. The effectiveness of directed multisensory stimulation versus
non directed stimulation in comatose closed head injured patients: pilot study of a single sub-
ject design. Brain Inj. 1992;6(5):435–45.
16. Lippert-Grüner M, Terhaag D. Multimodal early onset stimulation (MEOS) in rehabilitation
after brain injury. Brain Inj. 2000;14(6):585–94.
17. Mitchell S, Bradley V, Welch J, Britton P. Coma arousal procedure: a therapeutic intervention
in the treatment of head injury. Brain Inj. 1990;4(3):273–9.
18. Sisson R. Effects of auditory stimuli on comatose patients with head injury. Heart Lung.
1990;19(4):373–8.
19. Tablot L, Whitaker H. Brain-injured persons in an altered state of consciousness: measures and
intervention strategies. Brain Inj. 1994;8(8):689–99.
10 Sensory Stimulation Program 103

20. Wilson S, Powell G, Elliott K, Thwaites H. Sensory stimulation in prolonged coma: four single
case studies. Brain Inj. 1991;5(4):393–400.
21. Wilson S, Powell G, Brock D, Thwaites H. Behavioural differences between patients who
emerged from vegetative state and those who did not. Brain Inj. 1996;10(7):509–16.
22. Lombardi F. Sensory stimulation of brain-injured individuals in coma or vegetative state:
results of a Cochrane systematic review. Clin Rehabil. 2002;16:465–73.
23. Kater K. Response of head-injured patients to sensory stimulation. West J Nurs Res.
1989;11:20–33.
24. Teasdale G, Jennett B. Assessment of coma and impaired consciousness. A practical scale.
Lancet. 1974;2(7872):81–4.
25. Johnson D, Roethig Johnston K, Richards D. Biochemical and physiological parameters of
recovery in acute severe head injury: responses to multisensory stimulation. Brain Inj.
1993;7(6):491–9.
26. Davis A, Gimenez A. Cognitive-behavioral recovery in comatose patients following auditory
sensory stimulation. J Neurosci Nurs. 2003;35(4):202–9.
27. Urbenjaphol P, Jitpanya C, Khaoropthu S. Effects of the sensory stimulation program on
recovery in unconscious patients with traumatic brain injury. J Neurosci Nurs. 2009;41:10–6.
28. Oh H, Seo W. Sensory stimulation programme to improve recovery in comatose patients.
J Clin Nurs. 2003;12:394–404.
29. Giacino J, Kalmar K, Whyte J. The JFK Coma Recovery Scale-Revised: measurement charac-
teristics and diagnostic utility. Arch Phys Med Rehabil. 2004;85(12):2020–9.
30. Schnakers C, Majerus S, Giacino J, et al. A French validation study of the Coma Recovery
Scale-Revised (CRS-R). Brain Inj. 2008;22(10):786–92.
31. Schnakers C, Giacino J, Kalmar K, et al. Does the FOUR score correctly diagnose the vegeta-
tive and minimally conscious states? Ann Neurol. 2006;60(6):744–5.
32. Groswasser Z, Cohen M, Blankstein E. Polytrauma associated with traumatic brain injury:
incidence, nature and impact on rehabilitation outcome. Brain Inj. 1990;4:161–6.
Chapter 11
Feasibility of Oral Feeding in Patients
with Disorders of Consciousness

Audrey Maudoux, Ingrid Breuskin, Olivia Gosseries, Caroline Schnakers,

and Audrey Vanhaudenhuyse

Introduction

Eating and drinking are basic pleasures of life, considered as obvious to most of us.
However, the ease with which we perform these actions masks the complexity of the
underlying neuronal control. Various regions of the central nervous system, from
brainstem to cortex, are involved in the realization of this complex sensorimotor
sequence. Swallowing involves muscles of the face, tongue, pharynx, larynx, and
esophagus, with a total of 26 pairs of muscles, to which must be added the unique

A. Maudoux, M.D. (*)

Oto-Rhino-Laryngology Head and Neck Surgery Department, University Hospital of Liège,
Sart-Tilman-B35, 4000 Liège, Belgium
Coma Science Group, Cyclotron Research Center, University of Liège,
Sart-Tilman-B30, 4000 Liège, Belgium
e-mail: [email protected]
I. Breuskin, M.D., Ph.D.
Oto-Rhino-Laryngology Head and Neck Surgery Department, University Hospital of Liège,
Sart-Tilman-B35, 4000 Liège, Belgium
e-mail: [email protected]
O. Gosseries, M.Sc.
Coma Science Group, Cyclotron Research Center, University of Liège,
Sart-Tilman-B30, 4000 Liège, Belgium
e-mail: [email protected]
C. Schnakers, Ph.D. • A. Vanhaudenhuyse, Ph.D.
Coma Science Group, Cyclotron Research Center, University of Liège,
Sart-Tilman-B30, 4000 Liège, Belgium
e-mail: [email protected]; [email protected]

C. Schnakers, S. Laureys (eds.), Coma and Disorders of Consciousness, 105

DOI 10.1007/978-1-4471-2440-5_11, © Springer-Verlag London 2012
106 A. Maudoux et al.

upper lingual muscle, and five pairs of cranial nerves. The complexity of the underlying
neural mechanism as well as the number of muscles and cranial nerves involved made
the study of swallowing in animals and humans difficult. Although diagnosis meth-
ods, assessment, and rehabilitation of swallowing disorders have been extensively
studied, no major breakthrough in the understanding of the neurophysiological basis
supporting the phenomenon has been provided. Nevertheless, interest in the study of
swallowing has increased in the past few years. Initially, groups of patients with neu-
rological disorders (e.g., stroke) were studied. A precise anatomical study of brain
lesions using CT scanner or magnetic resonance imaging (MRI) was correlated with
the clinical evaluation of swallowing in order to establish a relationship between neu-
rological injury and the presence and type of dysphagia. Subsequently, the develop-
ment of functional imaging techniques allowed the study of swallowing in healthy
volunteers. According to several studies, the frequency of dysphagia among subjects
with severe brain injury varies between 25% and 61% [1–3]. These swallowing
difficulties are mainly due to physiological deficits (affecting the swallowing mecha-
nism) and impaired cognition. Faced with the difficult management of patients with
altered state of consciousness, the use of gustatory stimuli, as well as the rehabilitation
of swallowing could constitute a supplementary therapy which is currently rarely con-
sidered [4]. Furthermore, one may ask whether the clinical assessment of swallowing
could be used as a tool for neurological diagnosis or prognosis of consciousness
recovery. The review presented here aims to summarize our current knowledge regard-
ing the neural control of swallowing (whether initiated voluntarily or not), to assess
the role of awareness and willingness on the control of swallowing, and finally, to
establish the feasibility of oral feeding in patients with disorders of consciousness.

Neurophysiology of Swallowing

Since Magendie in 1825, swallowing is usually divided into three parts: oral, pha-
ryngeal, and esophageal. The oral phase is usually described as a voluntary phase,
the pharyngeal phase and the esophageal phase (which is mainly under the control
of the autonomic nervous system) as reflex (Fig. 11.1).

The Oral Phase

The oral phase is the only voluntary stage of swallowing. The different steps of this
first phase are the ingestion, mastication, and food insalivation. These steps lead to the
formation of the bolus, which will then be transported to the pharynx. During the oral
phase, the oral cavity should remain closed in order to contain food. At the front, the
contraction of the orbicularis oris muscle acts as the anterior sphincter. At the back,
the velum, lowered against the tongue, acts as the posterior sphincter. It prevents the
early passage of food in the pharynx. During mastication, the tongue carries food from
the teeth to crush, cut, and tear the aliments. The insalivation occurs during chewing.
This step has both a digestive role, due to the action of amylase, and a mechanical role
11 Feasibility of Oral Feeding in Patients with Disorders of Consciousness 107

Fig. 11.1 The three phases of

swallowing: A Oral phase, B
Pharyngeal phase, C Esophageal
phase. 1 nasal cavity, 2 soft
palate, 3 mouth, 4 tongue,
5 mandible, 6 nasopharynx, 7
1
oropharynx, 8 laryngopharynx,
9 epiglottis, 10 larynx, 11 vocal 2
cords, 12 esophagus (Adapted
from [5])

A
4 B

5 9

10 C
6
7
8

11
12

by providing lubrication and cohesion to the bolus. During the propulsion step, the
bolus is moved to the back of the oral cavity. When food reaches the Wassilieff area
(a mucous membrane covering the soft palate, the root of the tongue, the vallecula,
and the posterior pharyngeal wall), the pharyngeal phase of swallowing begins. Thus,
the Wassilieff area delimits the transition from the oral phase to the second stage of
swallowing—the pharyngeal phase. The oral phase fluctuates in time depending on
taste, food consistency, environment, hunger, motivation, and patient’s level of con-
sciousness. It relies on good tongue mobility and on perfect function of numerous
muscles: the suprahyoid muscles, the jaw closing muscles (temporalis muscle,
masseter muscle), the pterygoid and the infrahyoid muscles (allowing stabilization of
the hyoid bone), the orbicularis oris muscle, and the palatoglossus muscles.

The Pharyngeal Phase

When the food reaches the pharyngeal space, the swallowing reflex itself starts.
This is followed by a series of events leading to the transportation of the bolus to the
esophagus and to the protection of the airway. The fine and synchronized movements
of the different muscles involved are controlled by the brainstem via the central pat-
tern generator (CPG). This phase includes three stages.
108 A. Maudoux et al.

1. Protection of the nasopharynx and larynx

The occlusion of the velopharyngeal sphincter aims to isolate the oropharynx from
the nasopharynx. Elevation of the soft palate together with the approximation
of the lateral pharyngeal walls avoid nasal reflux of food and liquid. Protection
of lower airways is provided by the occlusion of the laryngeal sphincter (eleva-
tion of the larynx, closing of the vocal cord plane and of the ventricular folds,
forward tilt of the arytenoids, and lowering of the epiglottis). The ascension of the
larynx is a critical swallowing step. It protects the airway by positioning the larynx
under the root of the tongue and by closing the ventricles of the larynx. This upward
movement also promotes the opening of the upper esophageal sphincter (UES).
2. Propulsion of the bolus through the pharynx
The propulsion of the bolus is mainly provided by the posterior movement of the
root of the tongue, pushing the bolus into the esophagus. To complete the hypo-
pharyngeal draining, the pharyngeal constrictor muscles contract sequentially,
achieving a peristaltic wave and pushing the food bolus from top to bottom.
3. Upper esophageal sphincter opening
During swallowing, the cricopharyngeal muscle relaxes and the sphincter is pulled
forward by contraction of suprahyoid muscles, allowing its opening. Finally, the
pharyngeal phase ends, and the UES closes until the next swallowing.

The Esophageal Phase

The esophageal phase ensures the transfer of the bolus from the UES to the stomach
through a peristaltic wave. By working in sync, the two muscle layers of the esopha-
gus (external longitudinal and inner circular layers) create the movements of
peristalsis.

The Neurological Control

Swallowing requires perfect coordination between the different muscles mentioned

above. It is achieved through fine control of the central nervous system. Indeed, as
the respiratory and digestive tracts cross in the pharynx, the slightest misstep in the
sequence of swallowing may cause the passage of part of the bolus into the airway.
The main structure in-charge of the neural control is the swallowing CPG, located
in the brainstem. The CPG receives both sensory inputs from the periphery (oral
cavity, pharynx, and larynx) and cortical afferents. It integrates these informations,
develops a motor program, and transmits this program to the motor neuron nuclei.
It organizes all the sequential activation of motor neurons involved in swallowing
(Fig. 11.2).
The sensory inputs come primarily from the glossopharyngeal (IX) and the supe-
rior laryngeal nerves, a branch of the vagus nerve (X), both crucial for airway protec-
tion. Stimulation of areas innervated by the glossopharyngeal nerve appears to
11 Feasibility of Oral Feeding in Patients with Disorders of Consciousness 109

Cortical Dorsal group Peripheral sensory

inputs Nucleus tractus solitarius afferents inputs

IX−X−V

Ventral group XII

Nucleus ambiguus
V
C1−C3

VII
IX X

Pons Cervical cord

Medulla oblongata

Fig. 11.2 Schematic representation of the central pattern generator (CPG). Sensory inputs from
the periphery (oral cavity, pharynx, and larynx) and cortical afferents project around the nucleus of
the solitary tract (dorsal swallowing group). The dorsal group integrates these informations, devel-
ops the motor swallowing program and transmits this program to the premotor neurons of the
ventral group, located next to the nucleus ambiguus (Adapted from [6] and [7])

facilitate swallowing in humans [8]. Superior laryngeal nerve stimulation is also

efficient to trigger swallowing in most mammals [9]. Informations from the
glossopharyngeal and superior laryngeal nerves are transmitted to the swallowing
center via the solitary tract. To the sensory inputs coming from the superior laryngeal
and glossopharyngeal nerves must be added the influence of sensory afferents from
the trigeminal nerve (V), and particularly from one of its branches, the mandibular
nerve (V3). Mechanoreceptors located in the oral cavity and at the level of the tem-
poromandibular joints provide informations on the consistency and volume of the
bolus. It modulates the motor response by changing its amplitude and/or duration.
In the medulla oblongata, there is a network of interneurons responsible for the
development of the motor program of swallowing [6]. This network is divided into
two groups: a dorsal group organized at the level of the dorsal solitary nucleus and
the reticular formation and a ventral group organized around the nucleus ambiguus.
The dorsal cluster receives sensory inputs from the periphery and cortical afferents.
It develops the motor program sent to the swallowing motor nuclei by the ventral
cluster. The trigeminal motor nuclei are located in the middle part of the pons. They
innervate the tensor veli palatini, the anterior belly of digastric, mylohyoid muscles,
masseter muscles, temporalis muscles, and medial and lateral pterygoid muscles.
The facial motor nuclei are located at the bottom of the pons. They control the
muscles of facial expression, the posterior belly of digastric, and stylohyoid mus-
cles. The nucleus ambiguus, which occupies the entire height of the medulla oblon-
gata, controls the innervation of the esophagus and the velopharynx (motor fibers of
IX and X) via its rostral part, whereas its caudal part innervates the larynx (the
110 A. Maudoux et al.

Table 11.1 Cranial nerves involved in the process of swallowing

Cranial nerves Role
V Trigeminal Sensory: face, mouth, cheek, chin, lips, palate, teeth, nasal cavity,
Three branches: mandible, anterior two-thirds of the tongue
Ophthalmic nerve (V1) Motor: masticator muscles, tensors of the soft palate, mylohyoid
Maxillary nerve (V2) muscles, anterior belly of digastric
Mandibular nerve (V3)
VII Facial Taste: two-thirds of the tongue
Secretory: lacrimal, submandibular, and sublingual glands,
mucous membranes of the palate and nose
Motor: facial expression muscles, nasal muscles, posterior belly
of digastric, stylohyoid muscles, stapedius muscles
IX Glossopharyngeal Taste: posterior third of the tongue
Secretory: parotid glands
Sensory: posterior third of the tongue, pharynx, and soft palate
Motor: stylopharyngeus muscle, pharyngeal constrictors
X Vagus Taste: epiglottic region and root of the tongue
Sensory: epiglottis, laryngeal mucosa
Motor: esophageal muscles, laryngeal muscles (except stylopha-
ryngeus) with the help of the accessory nerve (XI), muscles of
the soft palate (except tensor veli palatini)
XII Hypoglossal Motor: all the muscles of the tongue (except palatoglossus),
geniohyoid muscles, thyrohyoid muscles
Adapted from [5, 10, 11]

bulbar accessory nerve [XI]). The hypoglossal nuclei (XII) associated with the C1
roots innervates the muscles of the tongue. The C2 and C3 cervical roots, together
with the hypoglossal nerves, control the infrahyoid muscles.
In summary, sensory informations are dependent on three pairs of cranial nerves:
the trigeminal (V), the glossopharyngeal (IX), and the vagus nerves (X). Yet, affer-
ents from the superior laryngeal nerve are the most powerful to trigger swallowing.
The innervation of the different muscles involved in swallowing is, in turn, provided
by five pairs of cranial nerves: the trigeminal (V), the facial (VII), the glossopharyn-
geal (IX), the vagus (X), and the hypoglossal nerves (XII) (Table 11.1).

Swallowing and Consciousness

Animal Studies

At the beginning of the last century, Sherrington observed swallowing in decere-

brate cats, in which regions located above the pons were disconnected from the
brainstem [12]. He described the effects of different stimuli (electrical, mechanical,
11 Feasibility of Oral Feeding in Patients with Disorders of Consciousness 111

liquid, or chemical) which, in contact with a region innervated by the superior laryn-
geal nerve, could trigger swallowing. This swallowing, neither involving oral prepa-
ration nor propulsion of the bolus, was regarded as reflex. The swallowing movement
observed in Sherrington’s cats can be divided into three stages: the closure of the
larynx, the propulsion of the bolus into the esophagus, and the opening of the UES
[13]. These three steps involve structures innervated by cranial nerves V, VII, X, and
the cervical plexus, C1–C3, whose nuclei are distributed in the lower brainstem and
spinal cord. Swallowing reflex has also been observed in decerebrate and anesthe-
tized goats. Swallowing was triggered by electrical stimulation of the superior
laryngeal nerve [14]. When trying to reproduce the experiment with glossopharyn-
geal nerve electrical stimulation, no swallowing reflex was observed. However, it
seemed to have a facilitating role on the swallowing movements induced by the
stimulation of the superior laryngeal nerve. These experiments suggest that trigger-
ing swallowing movements in decerebrate mammals are possible, even in the
absence of a cortical control.

Human Studies

Ultrasound studies on oral sensorimotor function and swallowing in human fetuses

showed early onset of swallowing mechanisms [15]. Indeed, the swallowing is
important for regulating the volume and composition of the amniotic fluid. The
pharyngeal phase of swallowing is one of the first pharyngeal motor responses
observed in fetuses and appears between the 10th and 14th week of gestation. The
growing network of interneurons in the brainstem controlling the pharyngeal
phase of swallowing achieved a functional level during fetal development as early
as the 11th week. From the 22nd to 24th weeks of gestation, swallowing move-
ments are constantly observed. The swallowing reflex seems also present even in
the absence of cortex in the human fetus. Indeed, several ultrasound studies show
the presence of swallowing movements in anencephalic fetuses [16]. These obser-
vations demonstrate that human fetus can achieve swallowing movements even
before a complete development of cortical and subcortical structures. One might
conclude that in humans, as observed in other mammals, swallowing seems to be
mainly controlled by the brainstem. However, many electrophysiological studies,
neuroimaging, and clinical observations show that the cerebral cortex also plays
an important role in the process of swallowing [17–19], even when the swallow-
ing is considered as reflex or automatic [20, 21]. This cortical involvement
becomes evident when looking at the incidence of dysphagia among subjects suf-
fering from stroke. In these patients, with anoxic brain damage, rate of dysphagia
ranges from <30% [22] to more than 50% [23]. Thus, lesions of the cerebral cor-
tex have important implications on human swallowing. Since the early studies of
cortical electrical stimulation done by Penfield, the involvement of the cerebral
cortex in the process of swallowing has been confirmed. Through the use of
112 A. Maudoux et al.

transcranial magnetic stimulation (TMS), Hamdy and colleagues demonstrated

the existence of an asymmetrical somatotopic representation of the different mus-
cles involved in swallowing [24], which was not correlated to the laterality of the
subject. If during a stroke the hemisphere with the largest representation of the
corticospinal tract involved in swallowing is damage, it increases the risk of dys-
phagia [24]. Recovery of an effective swallowing function, in the case of unilat-
eral involvement of the corticospinal tract, depends on the presence and
development of the same tract in the contralateral hemisphere [25–27]. The devel-
opment of functional imaging techniques (Positron Emission Tomography: PET,
functional magnetic resonance imaging: fMRI) has revolutionized the study of the
brain. These techniques allowed for a better understanding of cortical mecha-
nisms involved in swallowing and confirm the important role of the cerebral cor-
tex. The main cortical and subcortical areas involved are the cerebellum, the basal
ganglia and the sensorimotor, prefrontal, anterior cingulate, insula, and temporo-
parietal regions [28–32].

Precentral Gyrus

The precentral gyrus is the most consistently cited region in functional imaging
studies exploring swallowing. It encompasses the premotor cortex (including the
supplementary motor area, SMA) and the primary motor cortex. This region con-
trols the muscles of the oral cavity, pharynx, and larynx [24] and is mainly involved
in planned and voluntary movements. However, the precentral gyrus seems also
active during the production of automatic swallowing and its role in the realization
automatic movement has to be clarified [20]. The SMA is located in front of the
primary motor cortex. This region is related to the planning of complex movements
and especially sequential movements. It seems to have a modulatory action on the
motor sequence and to be involved in the initiation of the pharygo-oesophageal
phase of swallowing [28].

The Prefrontal Cortex

The prefrontal cortex is the anterior part of the frontal lobe and is located in front of
the premotor areas. This region is associated with the planning of complex cognitive
tasks. It is involved in the genesis of emotional states and in the regulation of auto-
nomic changes accompanying those states. The prefrontal cortex is also implicated
in the analysis of olfactory information.

The Anterior Cingulate Cortex

The anterior cingulate cortex is important for conscious awareness and for the pro-
cessing of stimuli related to emotion. Activation of the anterior cingulate cortex
11 Feasibility of Oral Feeding in Patients with Disorders of Consciousness 113

could account for the affective and attentional aspect of swallowing. Others consider
this region (which is involved in cognitive and attentional processes) as an interface
between intention and execution of the motor sequences involved in swallowing
[33]. In addition, the anterior cingulate cortex also plays a role in mediating the
visceral motor responses of the digestive tract.

The Insula

Several functional neuroimaging studies report activation of the insula during swal-
lowing [20, 28, 34]. Damage to this isolated region may induce dysphagia [34]. The
insula is involved, among other regions, in monitoring and analyzing informations
concerning the body’s homeostatic state. It is also involved in visceral motor control
(automatic) [35], somatic sensations of the orofacial region in primates, voluntary
control of orofacial movements, and control of speech. Activation of the insula
seems to play a role in integrating the sensory and motor aspects of the digestive
tract.

Parietal Regions and the Postcentral Gyrus

The sensory cortex is involved in the processing of facial as well as gustatory stim-
uli [36]. Its activation during swallowing reflects the large number of sensory infor-
mations coming from the oropharyngeal region. As already discussed, the sensory
inputs are essential for the regulation of swallowing [20].

Temporal Regions

In their work, Martin and colleagues proposed to link the temporal cortex activation,
more exactly the primary auditory cortex, to the analysis of acoustic stimuli pro-
duced by chewing and swallowing [20]. Indeed, we receive, via bone conduction,
sounds associated with swallowing. Furthermore, a study using PET suggested the
involvement of the anteromedial portion of the temporal lobe in the recognition of
gustatory stimuli [37].

The Basal Ganglia

The basal ganglia encompass several subcortical structures such as the caudate
nucleus, putamen or globus pallidus. These structures receive information from
various brain areas (frontal, prefrontal, and parietal) and transmit them to the SMA.
In doing so, the basal ganglia exert a facilitator effect on movement. Clinical studies
show that lesions of the basal ganglia, as found in Parkinson’s disease, lead to
difficulty in coordinating the oropharyngeal phase of swallowing.
114 A. Maudoux et al.

The Cerebellum

The cerebellum is known to be a regulator of the motor sequence [38–40]. It inte-

grates the sensory inputs and organizes, based on these inputs, the efferent motor
response [41]. It ensures a correct pharyngeal and laryngeal synergy and chronom-
etry such as elevating the larynx, closing the glottis, or triggering the swallowing
reflex.
The fact that many different cortical areas are activated during swallowing is not
surprising. Indeed, most studies are based on observation of the mechanisms
involved during voluntary swallowing (i.e., the swallowing of food or saliva on
command). Unlike voluntary swallowing, one might think that spontaneous swal-
lowing might not require cortical control. However, observation of patients with
swallowing apraxia suggests the opposite. This disorder—characterized by the
inability to perform reflex swallowing while voluntary swallowing (i.e., swallowing
on command) is preserved—was observed in some patients with left frontolateral
cortical lesions [42]. In addition, several studies have shown that reflex swallowing
is associated with activation of several brain areas [20, 21]. The network activated
during this particular type of swallowing is different from that observed during vol-
untary swallowing. According to Kern et al., reflex swallowing activates the pri-
mary sensorimotor cortex in contrast to voluntary swallowing, which activates, in
addition to this region, a multitude of other cortical regions such as the insula, pre-
frontal and cingulate gyri, cuneus, and precuneus [21]. These additional regions
could reflect the preparation and the voluntary realization of the movement. Another
study of Martin and colleagues showed a higher activity of the posterior part (caudal
part) of the anterior cingulate gyrus during voluntary swallowing as opposed to
reflex swallowing [20]. Although studies on decerebrated animals show that stimu-
lation of laryngeal or pharyngeal mucosa may trigger a swallowing reflex move-
ment without any cortical control, studies in adult humans show that cortical activity
is needed both in voluntary and swallowing reflex. A certain level of cortical integ-
rity seems necessary for the realization of voluntary or swallowing reflex move-
ments. However, the impact of the level of consciousness on a patient’s swallowing
ability has to be determined. The observation of swallowing in different physiologi-
cal alteration of consciousness states (e.g., during sleep), or under medication (e.g.,
during anesthesia) could give us interesting clue. During sleep, swallowing move-
ments are less frequent and long periods without swallowing are observed. When
swallowing occurs, it is associated with signs of awakenings on electroencephalog-
raphy (EEG), both during REM (rapid eye movement) and non-REM sleep.
Moreover, the frequency of swallowing movements seems related to the stage of
sleep. When sleep deepens, less swallowing movements are observed [43, 44]. In
patients receiving general anesthesia, changes in swallowing reflex are often
observed during the perioperative period. Several reasons can explain this swallow-
ing alteration: traumatic reasons (i.e., mucosal lesions as a consequence of endotra-
cheal intubation) and pharmacological reasons associated to the own
pharmacodynamic effects of each anesthetic agent used, including alteration of con-
sciousness [45].
11 Feasibility of Oral Feeding in Patients with Disorders of Consciousness 115

Oral Feeding in Patients with Disorder of Consciousness

Although approaches aiming at improving the well-being of patients with disorder

of consciousness increase, and although the use of sensory stimuli are used more
and more often in rehabilitation programs, the use of oral and gustatory stimuli
appears to have been shelved. In this challenging population, the utility of gustatory
stimuli is difficult to address. Indeed, the frequency of swallowing disorders associ-
ated with traumatic or anoxic brain injury varies between 25% and 61%
[1–3]. In addition, repeated inhalation can lead to complications (e.g., aspiration
pneumonia) with dramatic consequences in this fragile population. Unresponsive
wakefulness syndrome (UWS) patients are characterized by a lack of wakefulness
and awareness. They only show reflexive movements. From an anatomical point of
view, their brainstem is relatively preserved, while the white and gray matters are
severely affected. They also show an impairment of their associative cortices
involved in the realization of complex information processing such as memory,
attention, planning, or language. Unlike UWS patients, minimally conscious state
(MCS) patients show reproducible but fluctuating signs of consciousness (e.g.,
visual tracking or response to command). These MCS patients are, however, not
able to communicate [46]. Their brain metabolic activity is higher than in UWS
patients [47] and their associative areas are less affected. Based on our current
knowledge, we know that relative preservation of the brainstem, as observed in
UWS patients, cannot guarantee the absence of a swallowing disorder. We also
know that the cerebral cortex plays an important role in either voluntary or reflex
swallowing [20]. Accordingly, given the better overall brain activity in MCS
patients, the probability of observing a preserved swallowing function with a low
aspiration risk (i.e., with a low risk of penetration of ingested material in the air-
way), is theoretically more important than in UWS patients. However, any cortical
lesion, even if limited, is likely to cause major swallowing disorders.
According to a study of MacKay and colleagues conducted in patients with trau-
matic brain injury, in whom 41% showed signs of aspiration, the factor influencing
the most the recovery of a full oral feeding was the Ranchos Los Amigos (RLA)
score [48]. This scale was developed at Ranchos Los Amigos Hospital in California
to assess cognitive functioning [49]. A score lower than III on the RLA scale refers
to an UWS patient, a score between III and V to a MCS patient, and a score of VI to
a patient who emerged from the minimally conscious state. In the MacKay study,
the minimal RLA score required to be able to begin an oral therapy was IV. To
achieve a complete diet of three meals a day, a score of VI or higher was required.
According to this study, the patient’s ability to eat seemed related to the level of
consciousness. For these authors, the severity of the swallowing disorders is related
to a low score on the Glasgow coma scale (GCS), a low RLA score, the presence of
a tracheotomy, and at least 2 weeks of mechanical ventilation [1]. Similarly, Ward
and colleagues observed that patients with a GCS score between 3 and 8 showed a
longer delay before the resumption of oral feeding compared to patients with a GCS
score above 8 [50]. Other studies also show that low level of consciousness is
116 A. Maudoux et al.

associated with low oral feeding capacity [51, 52]. To assess the feasibility of an
effective and objective evaluation of swallowing in patients with severe disorders of
consciousness, Pirozzi and colleagues evaluated 12 tracheotomized patients with
severe disorder of consciousness following a traumatic brain injury [53]. In this
study, one of the secondary objectives was to evaluate the feasibility of using liquid
or solid food in a sensory stimulation program at the acute phase. Out of the 12
subjects, 7 were diagnosed as UWS (RLA score of II) and 5 as MCS (RLA score of
III). First, for 92% of their patients, a comprehensive swallowing assessment includ-
ing video-fluoroscopy (using four substances of different consistencies) could be
achieved. Video-fluoroscopy involves the ingestion of a barium-based contrast. The
progression and distribution of this product in the oral cavity, pharynx, and esopha-
gus are then followed using an X-ray system, allowing the observation and evalua-
tion of the three phases of swallowing. Thus, no matter the level of consciousness,
it seems possible to objectively assess the patient’s swallowing abilities. It should be
noted that the authors have reported signs of aspiration in 25% of their subjects
(n = 3), which is less than the aspiration rate reported by MacKay and colleagues
[1]. In these three patients, two were UWS and presented lesions in the brainstem
and cortex. The third was MCS and had only cortical lesions. All the aspirations
observed in these three patients were silent (i.e., occurring without visible or audible
sign such as a cough reflex). In consequence, traditional bedside swallowing assess-
ments would have missed aspirations in these three patients. This corroborates pre-
vious publications showing that traditional clinical bedside swallowing assessments
are only 66% accurate in screening aspiration risks [54, 55]. Making a decision on
whether oral feeding can be safely resumed cannot be done based on bedside evalu-
ation alone. Besides swallowing assessment, the authors performed various exer-
cises with the patients. Their rehabilitation technique for oral feeding included oral
range of motion exercises, thermal and taste stimulations with small amounts of
specified food, and/or drink consistencies, therapy to reduce bite reflex and care-
giver education. Following this therapy, none of the patients developed aspiration
pneumonia. When discharged from the acute care, all subjects were receiving some
amount of food and/or drink. Moreover, gastrostomy could be withdrawn in five
subjects as oral feeding was sufficient to meet their daily needs. An improvement in
cognitive-behavioral functions was also observed in 10 out of 12 subjects (6 UWS
and 4 MCS). They had an RLA score of III–VII at the time of discharge. Again, the
patient’s swallowing abilities seem to be related to changes in the patient’s level of
consciousness, at least during the acute phase. However, it is impossible to deter-
mine whether the increase of oral feeding could be related to the rehabilitation effect
or to the spontaneous evolution of consciousness. Similarly, Brady and colleagues
[4] have evaluated 25 patients with altered state of consciousness with an RLA
score varying between II and III (UWS and MCS). These patients were divided into
two groups: one group of 10 subjects with an RLA score of III who received oral
feeding rapidly and a second group of 15 subjects for whom the initiation of oral
feeding was delayed. In this second group, feeding was started as soon as an RLA
score of III or more was reached. For all these patients, an objective swallowing
assessment was performed either by using video-fluoroscopy or fiberoptic
11 Feasibility of Oral Feeding in Patients with Disorders of Consciousness 117

endoscopic evaluation. None of the patient showed evidence of inhalation during

the evaluation and none of the patient with an RLA score lower than III received
oral feeding. At discharge, when considering the number of patients receiving three
oral meals a day, there was no difference between the two groups. Again, this study
highlights the fact that patients with an altered state of consciousness can fully par-
ticipate in objective swallowing assessments. For the authors, oral feeding is related
to consciousness. On the contrary, there seems to be no relationship between early
therapeutic oral feeding and improvement of consciousness. They concluded that it
is safe to provide therapeutic oral feeding (i.e., small amounts of food given by a
speech therapist) to patients with an RLA score of III (MCS) under specific condi-
tions. These conditions are: no demonstration of aspiration or elimination of aspira-
tion by means of volume or consistency modification (on a baseline instrumental
swallowing examination) and close supervision during oral feeding.

Conclusions

Swallowing is a complex sensorimotor function divided into three phases: oral, pha-
ryngeal, and oesophageal. The achievement of the complete swallowing sequence
involves the muscles of the face, tongue, pharynx, larynx, and esophagus, with a
total of 26 pairs of muscles, to which must be added the unique upper lingual mus-
cle and five pairs of cranial nerves. Various regions of the central nervous system,
from brainstem to cortex, are involved in the realization of this complex sensorimo-
tor sequence. Because the respiratory and digestive tracts cross in the pharynx, the
slightest misstep in the sequence of swallowing may lead to complications some-
times life-threatening. The study of the neurophysiology of swallowing has evolved
since the late nineteenth century and Penfield’s studies, using direct electrical brain
stimulation, to the most recent studies using TMS, PET, and fMRI. Initially,
researches focused on the study of peripheral nerves, the organization of the CPG,
and the importance of the brainstem. Later, they focused on the involvement of
cortical and subcortical regions in reflex and voluntary swallowing movement, and
the interactions between these cortical regions and the CPG.
Dysphagia concerns more than half of the patients with severe brain injury [1]
and only a few UWS and MCS patients receive oral feeding. We know that swallow-
ing is controlled by a region located in the brainstem, the CPG, which incorporates
the afferents from the periphery in order to select the best motor response. We also
know that for both voluntary and reflex swallowing, cortical and subcortical brain
structures are involved. Only a few studies focused on swallowing disorders in
patients with impaired consciousness. However, all tend to state that the level of
consciousness during the acute phase determines the feasibility of resuming oral
feeding [1, 4, 48, 50–52]. Similarly, rapid resumption of oral feeding in patients
with altered state of consciousness is considered as a sign of better prognosis [56].
This is true in the acute phase and it would be interesting to evaluate whether a
relationship between consciousness and oral feeding still exists in the chronic phase.
118 A. Maudoux et al.

Conversely, the use of therapeutic oral feeding does not appear to affect the recovery
of consciousness [4]. Thus, rehabilitation of swallowing should not be considered
as a therapeutic tool but rather as a management option for these patients. Including
the rehabilitation of swallowing as part of a more global sensory stimulation pro-
gram could be achieved in this particular population and improve patient’s quality
of life. Indeed, administration of a therapeutic oral feeding in selected patients (with
RLA score III [MCS]), seems safe [4]. Because of the limited number of studies that
have addressed this issue and because those studies included only few patients, it is
difficult to give a clear and definitive answer to our initial questions. Further studies
have to be performed in a larger sample of patients and have to evaluate the relation-
ship between consciousness recovery, lesions of the central nervous system, and
swallowing abilities, both during acute and chronic phase. In conclusion, the evalu-
ation of swallowing abilities is important and should be a part of the evaluation of
all patients with altered state of consciousness. Indeed, whatever the underlying
neurological disease, dysphagia, because of its potential respiratory and nutritional
complications, is a marker of poor prognosis.

References

1. Mackay LE, Morgan AS, Bernstein BA. Factors affecting oral feeding with severe traumatic
brain injury. J Head Trauma Rehabil. 1999;14:435–47.
2. Winstein CJ. Neurogenic dysphagia. Frequency, progression, and outcome in adults following
head injury. Phys Ther. 1983;63:1992–7.
3. Mackay LE, Morgan AS, Bernstein BA. Swallowing disorders in severe brain injury: risk fac-
tors affecting return to oral intake. Arch Phys Med Rehabil. 1999;80:365–71.
4. Brady SL, Darragh M, Escobar NG, et al. Persons with disorders of consciousness: are oral
feedings safe/effective? Brain Inj. 2006;20:1329–34.
5. Bleeckx, D. Dysphagie: évaluation et rééducation des troubles de la déglutition. De Boeck
2002, 1st edn.
6. Jean A. Brain stem control of swallowing: neuronal network and cellular mechanisms. Physiol
Rev. 2001;81:929–69.
7. Ertekin C, Aydogdu I. Neurophysiology of swallowing. Clin Neurophysiol. 2003;114:2226–44.
8. Sinclair WJ. Initiation of reflex swallowing from the naso- and oropharynx. Am J Physiol.
1970;218:956–60.
9. Doty R. Neural organisation of deglutition. In: Code CF, editor. Handbook of physiology.
Washington, DC: American Physiological Society; 1968. p. 1861–902.
10. McFarland. L’anatomie en orthophonie : Parole, voix et déglutition. Elsevier Masson 2006.
11. Mistry S, Hamdy S. Neural control of feeding and swallowing. Phys Med Rehabil Clin N Am.
2008;19:709–28, vii-viii.
12. Miller FR, Sherrington CS. Some observations on the buccopharyngeal stage of reflex degluti-
tion in the cat. Q J Exp Physiol. 1916;9:147–86.
13. Hughes T. Neurology of swallowing and oral feeding disorders: assessment and management.
J Neurol Neurosurg Psychiatry. 2003;74 Suppl 3:iii48–52.
14. Ciampini G, Jean A. Role of glossopharyngeal and trigeminal afferents in the initiation and
propagation of swallowing. I–Glossopharyngeal afferents. J Physiol Paris. 1980;76:49–60.
15. Delaney AL, Arvedson JC. Development of swallowing and feeding: prenatal through first
year of life. Dev Disabil Res Rev. 2008;14:105–17.
11 Feasibility of Oral Feeding in Patients with Disorders of Consciousness 119

16. Peleg D, Goldman JA. Fetal deglutition: a study of the anencephalic fetus. Eur J Obstet Gynecol
Reprod Biol. 1978;8:133–6.
17. Martin RE, Sessle BJ. The role of the cerebral cortex in swallowing. Dysphagia.
1993;8:195–202.
18. Miller AJ. The neuroscientific principles of swallowing and dysphagia. San Diego/London:
Singular Publication Group; 1999.
19. Michou E, Hamdy S. Cortical input in control of swallowing. Curr Opin Otolaryngol Head
Neck Surg. 2009;17:166–71.
20. Martin RE, Goodyear BG, Gati JS, Menon RS. Cerebral cortical representation of automatic
and volitional swallowing in humans. J Neurophysiol. 2001;85:938–50.
21. Kern MK, Jaradeh S, Arndorfer RC, Shaker R. Cerebral cortical representation of reflexive and
volitional swallowing in humans. Am J Physiol Gastrointest Liver Physiol. 2001;
280:G354–60.
22. Young EC, Durant-Jones L. Developing a dysphagia program in an acute care hospital: a needs
assessment. Dysphagia. 1990;5:159–65.
23. Gordon C, Hewer RL, Wade DT. Dysphagia in acute stroke. Br Med J (Clin Res Ed).
1987;295:411–4.
24. Hamdy S, Aziz Q, Rothwell JC, et al. The cortical topography of human swallowing muscula-
ture in health and disease. Nat Med. 1996;2:1217–24.
25. Hamdy S, Aziz Q, Rothwell JC, et al. Explaining oropharyngeal dysphagia after unilateral
hemispheric stroke. Lancet. 1997;350:686–92.
26. Hamdy S, Rothwell JC, Aziz Q, et al. Long-term reorganization of human motor cortex driven
by short-term sensory stimulation. Nat Neurosci. 1998;1:64–8.
27. Hamdy S, Aziz Q, Rothwell JC, et al. Recovery of swallowing after dysphagic stroke relates
to functional reorganization in the intact motor cortex. Gastroenterology. 1998;115:1104–12.
28. Hamdy S, Mikulis DJ, Crawley A, et al. Cortical activation during human volitional swallow-
ing: an event-related fMRI study. Am J Physiol. 1999;277:G219–25.
29. Hamdy S, Rothwell JC, Brooks DJ, et al. Identification of the cerebral loci processing human
swallowing with H2(15)O PET activation. J Neurophysiol. 1999;81:1917–26.
30. Mosier K, Patel R, Liu WC, et al. Cortical representation of swallowing in normal adults:
functional implications. Laryngoscope. 1999;109:1417–23.
31. Zald DH, Pardo JV. The functional neuroanatomy of voluntary swallowing. Ann Neurol.
1999;46:281–6.
32. Mosier KM, Liu WC, Maldjian JA, Shah R, Modi B. Lateralization of cortical function in
swallowing: a functional MR imaging study. AJNR Am J Neuroradiol. 1999;20:1520–6.
33. Paus T. Primate anterior cingulate cortex: where motor control, drive and cognition interface.
Nat Rev Neurosci. 2001;2:417–24.
34. Daniels SK, Foundas AL. The role of the insular cortex in dysphagia. Dysphagia.
1997;12:146–56.
35. Penfield W, Rasmussen T. The cerebral cortex of man. New York: Macmillan; 1950.
36. Burton H, Benjamin RM. Central projections of the gustatory system. In: Beidler LM, editor.
Handbook of sensory physiology, Chemical Senses Sect 2, Taste, vol. 4. Berlin: Springer;
1971. p. 148–63.
37. Small DM, Jones-Gotman M, Zatorre RJ, et al. A role for the right anterior temporal lobe in
taste quality recognition. J Neurosci. 1997;17:5136–42.
38. Miall RC. The cerebellum, predictive control and motor coordination. Novartis Found Symp.
1998;218:272–84; discussion 84–90.
39. Ivry R. Cerebellar timing systems. Int Rev Neurobiol. 1997;41:555–73.
40. Doya K. Complementary roles of basal ganglia and cerebellum in learning and motor control.
Curr Opin Neurobiol. 2000;10:732–9.
41. Hikosaka O, Nakahara H, Rand MK, et al. Parallel neural networks for learning sequential
procedures. Trends Neurosci. 1999;22:464–71.
42. Robbins J, Levin RL. Swallowing after unilateral stroke of the cerebral cortex: preliminary
experience. Dysphagia. 1988;3:11–7.
120 A. Maudoux et al.

43. Lichter I, Muir RC. The pattern of swallowing during sleep. Electroencephalogr Clin
Neurophysiol. 1975;38:427–32.
44. Sato K, Nakashima T. Human adult deglutition during sleep. Ann Otol Rhinol Laryngol.
2006;115:334–9.
45. de Larminat V, Dureuil B. Changes in the deglutition reflex during the perioperative period.
Ann Fr Anesth Reanim. 1994;13:49–56.
46. Giacino JT, Ashwal S, Childs N, et al. The minimally conscious state: definition and diagnostic
criteria. Neurology. 2002;58:349–53.
47. Laureys S, Owen AM, Schiff ND. Brain function in coma, vegetative state, and related disor-
ders. Lancet Neurol. 2004;3:537–46.
48. MacKay LaM AS. Early swallowing disorders with severe head injuries: relationships between
the RLA and the progression of oral intake. Dysphagia. 1993;8:161.
49. Hagen C. The Rancho Levels of Cognitive Functioning. The revised levels - 3rd edition. 1998.
Rancho Los Amigos Medical Center.
50. Ward EC, Green K, Morton AL. Patterns and predictors of swallowing resolution following
adult traumatic brain injury. J Head Trauma Rehabil. 2007;22:184–91.
51. Cherney LR, Halper AS. Swallowing problems in adults with traumatic brain injury. Semin
Neurol. 1996;16:349–53.
52. Morgan A, Ward E, Murdoch B, et al. Incidence, characteristics, and predictive factors for
Dysphagia after pediatric traumatic brain injury. J Head Trauma Rehabil. 2003;18:239–51.
53. O’Neil-Pirozzi TM, Momose KJ, Mello J, et al. Feasibility of swallowing interventions for
tracheostomized individuals with severely disordered consciousness following traumatic brain
injury. Brain Inj. 2003;17:389–99.
54. Splaingard ML, Hutchins B, Sulton LD, Chaudhuri G. Aspiration in rehabilitation patients:
videofluoroscopy vs bedside clinical assessment. Arch Phys Med Rehabil. 1988;69:637–40.
55. Linden P, Kuhlemeier KV, Patterson C. The probability of correctly predicting subglottic pen-
etration from clinical observations. Dysphagia. 1993;8:170–9.
56. Formisano R, Voogt RD, Buzzi MG, et al. Time interval of oral feeding recovery as a prognos-
tic factor in severe traumatic brain injury. Brain Inj. 2004;18:103–9.
Chapter 12
Pharmacological Treatments

Olivia Gosseries, Marie Thonnard, and Steven Laureys

Introduction

Disorders of consciousness (DOC) resulting from a severe brain injury include

coma [1], the vegetative state [2], and the minimally conscious state [3]. Coma is
characterized by a complete absence of arousal and awareness whereas the vegeta-
tive state is characterized by an absence of clinical signs of consciousness in spite
of a preserved vigilance (opening of the eyes). Patients in a minimally conscious
state (MCS) show some signs of consciousness (such as nonreflexive movements
without language comprehension, MCS−; or such as command following, MCS+)
[4] but remain unable to communicate. If consciousness disorders are currently
more clearly defined, no pharmacological treatment is known to improve the level
of consciousness of these patients as demonstrated by placebo controlled random-
ized studies. Evidence based recommendations cannot be provided because there
are only a few uncontrolled studies that are often methodologically suboptimal [5].
Pilot studies and case reports showed that some severely brain-damaged patients

O. Gosseries, M.Sc. (*)

Coma Science Group, Cyclotron Research Centre, University of Liège,
Allée du 6 août, Sart Tilman B30, 4000 Liège, Belgium
Department of Neurology, University Hospital of Liège,
Liège, Belgium
National Fund for Scientific Research,
Liège, Belgium
e-mail: [email protected]
M. Thonnard, M.Sc. • S. Laureys, M.D., Ph.D.
Coma Science Group, Cyclotron Research Centre, University of Liège,
Allée du 6 août, Sart Tilman B30, 4000 Liège, Belgium
Department of Neurology, University Hospital of Liège,
Liège, Belgium
e-mail: [email protected]; [email protected]

C. Schnakers, S. Laureys (eds.), Coma and Disorders of Consciousness, 121

DOI 10.1007/978-1-4471-2440-5_12, © Springer-Verlag London 2012
122 O. Gosseries et al.

could benefit from pharmacological, rehabilitation, or sensory stimulation treat-

ments; but, in general, responses to treatment still remain unsatisfactory.
Several pharmacological agents can contribute to the recovery of consciousness
in some patients with an altered state of consciousness by targeting various path-
ways of the central nervous system. Sensory perception is controlled by a complex
neural network, which includes reticulothalamic cholinergic projections, thalamo-
cortical, and reticulocortical glutaminergic projections. Lesions in the white matter
connections of these networks may affect consciousness and cognition, and several
drugs including dopaminergic agents can act on this network by supporting a recov-
ery of consciousness.
This chapter summarizes the current state of knowledge on potentially useful
drugs acting on the recovery of consciousness in severely brain-damaged patients
and provides indications on the underlying mechanisms possibly explaining the
effects of these drugs on the awakening and recovery of consciousness. Exploratory
and retrospective studies as well as case reports concerning sporadic cases of spec-
tacular recovery will be discussed regarding drugs as amantadine, levodopa, bro-
mocriptine, apomorphine, methyphenidate, zolpidem, baclofen, sertaline,
amitriptyline, desipramine, and lamotrigine (Table 12.1).

Potential Pharmacological Treatments

Amantadine

Amantadine is an old dopaminergic agent initially used in the treatment of

Parkinson’s disease. For its antiviral properties, amantadine was also employed
against influenza. It increases the availability of dopamine in the striatum both at
the pre- and postsynaptic levels, facilitates the release of dopamine, and delays
its recapture, resulting in an increase of synaptic dopamine concentration. At the
postsynaptic level, amantadine increases the number of dopaminergic recep-
tors [6]. It is also a dose-dependent antagonist of the N-methyl-d-aspartate
receptors.
The use of amantadine is correlated with a better outcome among severe trau-
matic brain-injured patients. A retrospective study has shown, e.g., that in 74
acute traumatic patients diagnosed in a vegetative state, the group treated with
amantadine (n = 41) obtained higher scores on the Glasgow Coma Scale (GCS)
[7] when discharged from the intensive care unit than the group who did not
receive the drug [8]. Mortality was also lower in the treatment group (6%) than
in the nontreatment group (51%) [8]. Another study on 35 patients showed
improvement of behavioral scores (mini-mental state examination [MMSE] [9],
Glasgow outcome scale [10], and disability rating scale [DRS] [11]) during a
treatment over 6 weeks in the acute phase of severe traumatic brain injury [12].
Note that these two studies took place when the patients were in acute stage and,
thus, presented a bias of spontaneous recovery, which may have interfered with
 12

Table 12.1 Pharmacological treatment studies aiming to improve the level of consciousness in patients with severe brain damage with chronic disorders of
consciousness
Pharmacological Study (first Interval lesion-
agent author, year) N/etiol. Diagnostic treatment Treatment Effect Results Type of study Ref
Dopaminergic agent
Amantadine Schnakers 1anox MCS 2 years 200 mg/day for 6 weeks + ↑ CRS-R score, response to Prospective, [16]
(2008) command, automatic ABAB
motor behavior paradigm
Patrick 10TBI VS/MCS >1 month 50 mg/day for 3 weeks, + ↑ CNC, WNSSP, DRS Prospective, [19]
(2006) 100 mg/day for 1 week, scores double-blind
50 mg/day for 2 weeks randomized
Pharmacological Treatments

Hughes 123TBI Severe lesion ±6 weeks 2 × 100 mg for 1 week, / No improvement Retrospective [20]
(2005) 200 mg 2×/day for
2 weeks
Saniova 41TBI VS <3 months 200 mg 2×/day for + ↑ GCS score, ↓ mortality Retrospective [8]
(2004) 3 days
Meythaler 35TBI MCS 4 days to 200 mg/day for 6 weeks + ↑ MMSE, GOS, DRS scores Prospective, [12]
(2002) 6 weeks placebo
controlled
double-blind
randomized
Zafonte 1TBI MCS 5 months 100 mg/day, ↑100 mg/5 days + Emergence of MCS, Case-study, [14]
(1998) with maximum 400 mg/ ↑ CNC score, active ABAB
day. After 35 days, participation paradigm
diminution
100 mg/5 days, when
100 mg/day, ↑ again
Gualtieri 30TBI Moderate to 2–144 months 50–100 mg/day, ↑ 50–100 mg/ + ↑ awakening, ↓ tiredness, Case-study [17]
(1989) severe week during 6 weeks up to ↓ distractedness and
lesion 1 year. If no response, stop ↓ aggressiveness
treatment after 6 weeks
(continued)
123
Table 12.1 (continued)
124

Pharmacological Study (first Interval

agent author, year) N/etiol. Diagnostic lesion-treatment Treatment Effect Results Type of study Ref
Levodopa Matsuda 5TBI VS/MCS 3–22 months Up to 450 mg/day + Emergence from VS and Case-study [24]
(2005) MCS, response to
command, verbal and
written production,
functional communication
Krimchansky 8TBI VS 1–7 months 5/250 mg (carbidopa), 25% + Response to command within Prospective [25]
(2004) of a tablet 2 weeks, functional
3×/day, ↑ of a quarter of communication within
tablet every week. 1 month, one patient
Maximum three tablets/ remains MCS
day
Matsuda 3TBI VS 3–12 months 100/25 mg (carbidopa) + Response to command, visual Case-study [23]
(2003) 2–3×/day pursuit, “yes-no”
communication
Haig 1TBI VS 6 months 100/10 mg (carbidopa) + Response to command, verbal Case-study [22]
(1990) 2×/day communication
Bromocriptine Passler 5TBI VS 33–50 days 1.25 mg 2×/day then 2.5 mg + Rate of transition VS-MCS Case-study [26]
(2001) 2×/week for 2–6 months larger than normal,
↑ DRS and CRS-R
scores, functional
communication
Apomorphine Fridman 8TBI VS/MCS 1–4 months ±5 mgmL−1, begins with + ↑ awakening, ↑ CNC and Case-study [28]
(2010) ratio of 2 mgh−1 for DRS scores, response to
12 h/day up to 8 mgh−1 command, functional
for 12–16 h/day for communication for all
6 months except one patient
Fridman 1TBI MCS 104 days ±5 mgmL−1, begins with + Response to command, Prospective [27]
(2009) ratio of 2 mgh−1 for “yes-no” communica- case-study
12 h/day up to 8 mgh−1 tion, ↑ CNC, DRS, GOS
for 12–16 h/day for scores
6 months
O. Gosseries et al.
 12

Methylphenidate Martin (2007) 22 (17 VS/MCS 3 months to Between 7.5 and 25 mg 2×/ / No difference in the rate and Retrospective [31]
TBI) 8 years day accuracy of responses to meta-
command analysis
Moein (2006) 40TBI Severe lesion <3 weeks Between 0.3 and 20 mg/kg/ + ↓ duration of hospital stay Prospective, [30]
dose 2×/day during (23%) double-blind
hospitalization placebo
controlled
randomized
Drug Patrick (2003) 10 (7 VS/MCS 22 days to Amantadine 100–200 mg/ + ↑ WNSSP scores, ↑ Case-study, [34]
combina- TBI) 4 months day, bromocriptine responses to structured ABA
tions 0.125–400 mg/day, stimuli, ↑ of the paradigm
methylphenidate recuperation ratio
Pharmacological Treatments

10–80 mg/day, levodopa

25/100 mg/day or
pramipexole 0.125–
0.5 mg/day for
7–132 days
GABAergic agent
Zolpidem Whyte (2009) 15 (8TBI) VS/MCS 3 months to 10 mg + and Only one responder to treatment Multicentric, [41]
23 years / (VS to MCS), ↑ score double-blind
CRS-R, visual pursuit, randomized
response to command
Cohen (2008) 1anox MCS 8 months 5 mg/day then 10 mg/day + ↑ arousal, ↑ interactions, Case-study [39]
for 3 weeks then 20 mg/ ↑ verbal and social
day responses
Shames 1anox MCS 18 months 10 mg + ↑ arousal, ↑ RLAS score, Case-study [40]
(2008) ↑ verbal communication,
reading, counting, verbal
production, auto-
alimentation
Singh (2008) 1TBI MCS 4 years 10 mg/day for 1 week / No effect on tests with Single-blind [43]
instructions which ↑ in case-study
complexity
(continued)
125
Table 12.1 (continued)
126

Pharmacological Study (first Interval

agent author, year) N/etiol. Diagnostic lesion-treatment Treatment Effect Results Type of study Ref
Lo 2anox VS ±1 month 10 mg for 2 days, repeated / No arousal improvement, no Double-blind [42]
(2008) once 2 days later recovering of conscious- case-study
ness, no ↑ GCS score
Brefel- 1anox MCS 2 years 20 mg/day for 1 week + ↑ arousal, ↑ motor and Prospective [38]
Courbon neuropsychological double-
(2007) performance (functional blind,
communication, eating, placebo
walking, reading and controlled
repeating words) randomized
Clauss 2TBI and VS 3–5 years 10 mg/day for 3, 5 and + ↑ arousal, ↑ GCS and Case-study [36]
(2006) 1 anox 6 years RLAS scores
Clauss 1TBI Semi-comatose 3 years 10 mg + Verbal response after 15 min, Case-study [35]
(2000) talking, answering simple
questions, spontaneous
interaction, counting, writing
Baclofen Sara 5(2TBI) VS 6–10 months 100 mg/day continuously, + ↑ arousal, ↑ CRS-R score, Case-study [51]
(2009) up to 200 mg/day response to command
Sara 1non-TBI VS 19 months 100 mg/day continuously, + VS to MCS, verbal Case-study [49]
(2007) up to 200 mg/day communication,
“yes-non” communica-
tion, eating, emotional
response
Taira 2TBI VS 1.5– 50 mg/day for 3 day then + ↑ arousal, response to Case-study [50]
(2007) 2.5 months 75 mg/day for l0 days command
and 100 mg/day for
l0 days then 50 mg/day
for 5 days then 100 mg/
day for 23 days
Kawecki 1TBI VS 18 days 100 mg/day + Spasticity, motor and Case-study [48]
(2007) sensory aphasia
improvement
O. Gosseries et al.
 12

Serotoninergic agent
Sertaline Meythaler 11TBI Severe lesion <2 weeks 100 mg/day for 2 weeks / No improvement Prospective, [53]
(2001) placebo
controlled
randomized
blinded
Amitriptyline Reinhard 3TBI Severe lesion 2–19 months 50–75 mg/day + ↑ arousal and initiation, Case-study [54]
and (1996) response to command,
desipramine “yes-no” communica-
tion, verbal
communication
Pharmacological Treatments

Anti-convulsive agent
Lamotrigine Showalter 13 (6TBI) Severe lesion 20 days to 150–400 mg + ↑ of consciousness and Case-study [55]
(2000) 10 months cognition, discharge
from hospital sooner
than expected
N number of patients, etiol etiology, TBI traumatic brain injury, anox anoxia, VS vegetative state, MCS minimally conscious state, d day, w week, mg milligram,
↑ improvement, ↓ decrease, + positive, / no effect, MMSE Mini Mental State Examination, GOS Glasgow Outcome Scale, DRS Disability Rating Scale, CRS-
R Coma Recovery Scale-Revised, RLAS Rancho Los Amigos Scale, GCS Glasgow Coma Scale, CNC Coma Near Coma Scale, WNSSP Western Neuro Sensory
Program Stimulation
127
128 O. Gosseries et al.

the results obtained. A spontaneous recovery may, however, also occur very late
in patients with DOC.
Zafonte et al. reported a dose-dependent response to amantadine in a
minimally conscious patient examined 5 months after a cranial trauma. During
the treatment, the patient recovered his communication abilities and the score
on the coma/near coma (CNC) scale [13] increased. This effect was reversible
when the treatment was stopped; and during its reintroduction, the patient could
communicate again [14]. Schnakers et al. have also shown the benefits of aman-
tadine in a patient who had been in a post-anoxic minimally conscious state for
2 years and who only presented visual pursuit. Following the treatment, he was
able to answer commands and the scores of the coma recovery scale-revised
(CRS-R) increased [15, 16]. Amantadine has been shown to improve conscious-
ness but also to reduce fatigue, distraction, and aggressiveness among 30 patients
evaluated between 2 months and 12 years after a traumatic brain injury [17].
Finally, in a pediatric population, an improvement has been observed on
behavioral scales (CNC, DRS, and Western NeuroSensory Stimulation Profile:
WSSNP) [18] after taking amantadine and pramipexol (a dopamine antago-
nist) in ten children with severe traumatic injury [19]. On the other hand, in a
group of 123 patients who suffered a severe cranial trauma, the rate of recov-
ery of consciousness was not different between the treatment and the control
group [20].
All the studies quoted up to now used only behavioral scales to objectify the
effects of the amantadine treatment. Only one study by Schnakers et al. used posi-
tron emission tomography (PET) to show an increase in the cerebral metabolic
activity in the frontoparietal network, which is known to be critical for the emer-
gence of consciousness [21], and the sensitivomotor areas in a minimally conscious
patient after the administration of amantadine [16].

Levodopa

As amantadine, levodopa is a dopaminergic agent initially indicated in the treatment

of Parkinson’s disease. A remarkable recovery has been observed in a 24-year-old
man with a cranial trauma diagnosed in a vegetative state for 6 months and who was
able to speak a few days after the administration of levodopa [22]. Note that no
standardized validated diagnostic behavioral assessment was performed and this
study like any other of the presented reports has also been published before the
introduction, in 2002, of the criteria of the minimally conscious state, and thus the
initial diagnosis of vegetative state might have been inaccurate. Five other DOC
patients with traumatic lesions also became more responsive after taking levodopa,
which was initially given to treat extrapyramidal signs [23, 24]. Finally, in a last
uncontrolled unblinded study, eight vegetative patients recovered signs of con-
sciousness after the administration of progressive amounts of levodopa. All patients
could respond to commands within the firsts 2 weeks of treatment and seven of
12 Pharmacological Treatments 129

them, including two assessed more than 9 months postinjury, were able to interact
in a functional way [25].

Bromocriptine

Bromocriptine is another dopamine agonist used primarily to treat Parkinson’s dis-

ease. This agent, less studied, is mainly an agonist of the postsynaptic dopamine D2
receptors. It has been associated with a higher rate of patients recovering from a
posttraumatic persistent vegetative state to a minimally conscious state in a retro-
spective study [26].

Apomorphine

Apomorphine is a nonselective dopaminergic agonist, which activates D1 and D2

receptors with a preference for the latter. This therapy was initially indicated to treat
Parkinson’s disease and erectile impotence but it has also shown positive effects in
a few severe brain-injured patients. A minimally conscious patient treated with apo-
morphine 104 days after a brain trauma suddenly recovered consciousness after one
day of treatment. He was able to move his legs upon request and to answer yes–no
questions, which was not the case before [27]. The patient recovered full conscious-
ness and a considerable functional recovery was maintained even after stopping the
treatment. Diffusion tensor imaging showed a reduction in thalamocortical and cor-
ticothalamic projections, as expected in these patients. Another uncontrolled case-
study on eight vegetative and minimally conscious patients with traumatic etiology
who were treated continuously with apomorphine showed a recovery of conscious-
ness for all patients except one, with an increase in the CNC and DRS scores [28].
These improvements lasted for at least 1 year, even after stopping the treatment.

Methylphenidate

This agent was initially used for children presenting attention deficit hyperactivity
disorders and was also prescribed for narcoleptic patients. This neurostimulant
increases the release of dopamine and noradrenalin while blocking the reuptake and
inhibiting monoamine oxydase [29].
Few studies have been conducted and no conclusive evidence supports the use of
a psycho-stimulant to improve the level of consciousness in DOC patients. Only one
study suggests that the early use of methylphenidate in intensive care is associated
with shorter hospital stays after severe trauma [30]. On the other hand, a recent
meta-analysis on 22 chronic patients with severe brain damage (17 of traumatic
130 O. Gosseries et al.

etiology) did not show any clinical improvement on the percentage of responses to
command after the administration of methylphenidate [31].
This pharmacological agent could rather have an effect on attention in the acute
and subacute phases of recovery. Indeed, a subjective improvement of mood, of
work-output, and vigilance was reported among 14 brain-traumatized patients hav-
ing attention and memory deficits and being treated with methylphenidate [32].
Other studies showed similar positive effects of methylphenidate on attention and
memory in patients suffering from moderate brain injury [33].
Finally, ten children and teenagers in a vegetative state and minimally conscious
state were treated with a combination of dopaminergic drugs (amantadine, meth-
ylphenidate, bromocriptine, levodopa, pramipexole) and presented an improve-
ment of responses to structured stimuli in an uncontrolled, unblinded prospective
study [34].

Zolpidem

Zolpidem is an imidazopyridine which acts like an agonist on subtype 1 of the

inhibiting receptors of the gamma-aminobutyric acid (GABAA). This agent was ini-
tially recommended in the treatment of insomnia and presents sedative, anticonvul-
sive, anxiolytic, and myorelaxant effects.
A number of studies have reported the use of zolpidem as an “awakening” agent
among vegetative and minimally conscious patients. This drug produces, occa-
sionally, a clear paradoxical temporary effect on the level of consciousness in
patients with severe brain damage. The effect of zolpidem has been described for
the first time in 2000 after the fortuitous discovery in an allegedly vegetative patient
with cranial trauma for more than 3 years and who started to speak and communi-
cate with his family 20 min after the administration of the medication [35]. Clauss
et al. reported impressive effects of this drug in four vegetative patients who had
suffered a traumatic or anoxic cerebral lesion 3–5 years before [35, 36]. The
patients were able to answer questions, speak, and feed themselves shortly after
taking a single dose of zolpidem (10 mg). Improvements were also observed for
the GCS scale (increasing from 5–9 to 10–15) and the Rancho Los Amigos scale
(from III to V–VII) [37]. The level of consciousness of these patients returned to
its initial state 4 h after the administration of the drug but an improvement was
observed again at the time of readministration. Similar transitory effects were also
reported among patients in a minimally conscious state resulting from cerebral
anoxia [38–40].
Only one study investigated the percentage of responders among patients in a
vegetative state and in a minimally conscious state [39]. Among 15 patients, only
one indicated a significant clinical response passing from vegetative to minimally
conscious state. The remaining 14 patients did not show any improvement. The rate
of response to zolpidem would, thus, be close to 7%. Whyte et al. suggest that the
effect of zolpidem is bimodal rather than gradual since no tendency for improvement
12 Pharmacological Treatments 131

of performance was observed among patients who did not respond to the treatment.
Some case-studies also underlined the absence of improvement among patients suf-
fering from a postanoxic encephalopathy or a severe cerebral trauma [41–43].
Several studies were interested in the mechanisms able to explain the effect of
zolpidem. Single photon emission computed tomography studies showed that zolpi-
dem increases the cerebral metabolism of hypoactive areas following traumatic or
anoxic lesions [35, 36, 44]. In the same way, using PET in a minimally conscious
patient, improvement of neuropsychological performances was correlated with an
increase in cerebral metabolism in the frontal and postrolandic areas after taking
zolpidem. Activations were also observed in anterior cingular and orbitofrontal cor-
tex, areas known to be involved in motivational processes. A mechanism of cell
dormancy was introduced to explain the effect of zolpidem: certain nonspecific
areas of the brain, adjacent or distant to the initially damaged zones (e.g., the ipsi-
lateral, contralateral hemisphere, or the cerebellum), would be inhibited by the
lesion. These inactive parts of the brain would recover their function after taking
zolpidem [35, 38, 44, 45], generating a recovery of consciousness. A recent study
using magnetoencephalography also showed that zolpidem decreased the number of
pathological slow waves associated to dormant cerebral tissue in a patient having
undergone a cerebrovascular accident [46].
Zolpidem could, thus, interact with the limbic loops of the brain by modulating
subcortical connections, and more particularly the globus pallidus, which would
bring the thalamocortical activity back to normal and would allow a recovery of
consciousness (see the mesocircuit model) [47]. From a molecular point of view,
changes could take place at the level of glutamate and GABA neurotransmitters
close to the cerebral lesions. The release of glutamate produces an excitotoxicity
and an excess of inhibitory GABA neurotransmitters as well as a long-term over-
sensitiveness of the GABAA receptors [35]. The inhibitory neurotransmitters, while
binding to the receptors of the ionic channels, generate a reduction of metabolism
and blood flow in the adjacent cerebral areas, thus, causing a state of cells dor-
mancy. While binding to GABAA receptors of dormant cells, zolpidem provokes
the inversion of the abnormal state of the neurons and associated metabolic
inhibition.

Baclofen

Baclofen is an agonist agent of the GABAB receptors, which acts on the posterior
horn of the spinal cord and which is used mainly against spasticity. Spasticity, a
frequent symptom following central nervous system lesions, can limit voluntary
movements in patients with DOC. Baclofen, crossing with difficulty the hematoen-
cephalic barrier, the antispasmodic effect remains modest when it is administered
orally. Therefore, a direct and continuous perfusion of baclofen in low doses in the
cerebrospinal fluid can be an effective treatment. Intrathecal baclofen therapy can
be a useful treatment against severe spasticity among DOC patients improving the
132 O. Gosseries et al.

quality of life by reducing pain related spasms but also by controlling persistent
autonomic dysfunctions such as tachycardia, tachypnea, fever, and breathing
difficulties.
Among vegetative patients who were treated with baclofen, some impressive
cases of recovery were reported in uncontrolled case-studies [48–50]. An 8-year-old
boy, who sustained a severe brain injury following a car accident, received baclofen
to treat his spasticity 2 months and a half after the accident. Three days after the
treatment, the patient started to pronounce words, to answer verbal requests, and
subsequently showed good recovery. Similarly, an 18-year-old man diagnosed in a
vegetative state received intrathecal baclofen 1 month and a half after his brain
trauma. Five days later, he was able to answer verbal requests and he subsequently
recovered motor and cognitive capacities [50]. Another case-study also showed an
impressive recovery in an 11-year-old girl presenting a vegetative state on admis-
sion due to diffuse axonal injury following a car accident. After 18 days, she was
treated with intrathecal baclofen and she recovered from spasticity, motor, and sen-
sory aphasia [48]. Note, however, that the time postinsult was relatively short in
these three patients and spontaneous recovery might have occurred as well.
A positive effect of baclofen was also observed in five allegedly vegetative
patients treated for spasticity in the chronic stage (at least 19 months after the injury).
Two weeks after, all except one patient presented clinical improvement, which
remained stable until the end of the 6-months follow-up. Improvements went from a
high increase in vigilance to a recovery of consciousness, as revealed by the changes
in the CRS-R scores [51]. Finally, in one last case-study, following the introduction
of intrathecal baclofen, a 44-year-old man recovered consciousness 19 months after
a nontraumatic brain injury (i.e., subarachnoid hemorrhage due to a ruptured aneu-
rysm). He however remains dependent on others for his daily activities [49].
Several assumptions have been made to explain the effects of baclofen on recov-
ery of consciousness. Some authors suggest a modulation of motor impulses of the
spinal cord on the possible cortical reactivation [51]. Baclofen would improve ner-
vous conduction in the demyelinized axons and could possibly accelerate the resto-
ration of diffuse axonal injury [52]. A modulation of the sleep–wake cycle has also
been considered as a mechanism responsible for the effect of baclofen [51]. But all
these explanations at present remain hypothetical.

Serotonin-Reuptake Inhibitors

Sertalin is an antidepressant of the selective serotonin reuptake inhibitor class.

Acute brain lesions have been reported to increase cerebral serotonin but in the
chronic phase a downregulation of the serotonin system is thought to occur. In a
prospective placebo-controlled randomized study, sertalin treatment did not improve
the level of consciousness among 11 severely brain-injured patients [53]. On the
other hand, amitriptylin and desipramin treatment (tricyclic antidepressants also
blocking the recapture of serotonin) improved recovery in three patients with severe
12 Pharmacological Treatments 133

traumatic brain injury in an uncontrolled unblinded study [54]. Two of these patients
presented deterioration when the drug was stopped, but an improvement was again
observed at the time of reintroduction of the treatment. The third patient started to
talk after being mute for more than 1 year. The possible mechanisms of action in
this context remain largely unexplained.

Lamotrigin

Lamotrigin is an antiepileptic agent used in the treatment of epilepsy and bipolar

disorders. By inhibiting the voltage-dependent sodium channels, it stabilizes the
neuronal membrane and inhibits glutamate release, an excitatory neurotransmit-
ter. The effects on the sodium ion channels contribute to the antiepileptic effects
while the antiglutamatergic agents act more on the psychotropic effects with a
possible neuroprotective action. Functional improvement of DOC patients has
been observed after administration of lamotrigin, showing recovery of conscious-
ness and cognition combined with an earlier discharge from hospital [55]. This
uncontrolled unblinded study concluded a possible effect on functional recovery,
and particularly in patients who had already emerged from the minimally conscious
state.

Mechanisms Aiming to Explain the Possible Positive Effects

of Pharmacological Treatments

Each drug affects one or more neuronal pathways. Amantadine, levodopa, bro-
mocriptine, methylphenidate, and apomorphine act mainly on the dopaminergic
system; whereas, zolpidem and baclofen affect preferentially the gabaergic system
(albeit at different locations in the nervous system), while sertraline, amitriptyline,
and desipramine interfere rather in serotoninergic processes. The subjacent neuro-
logical mechanisms to the positive effects of these drugs are currently not well-
understood. As we have seen, amantadine and zolpidem would increase the
metabolism of hypoactive cerebral regions [16, 56]. Zolpidem would play a main
role in the gabaergic system of the limbic loops in the brain [57]; whereas, baclofen
would act more on the spinal cord and would support the regeneration of motor
neurons [51].
The model of the mesocircuit has been proposed recently to explain the various
pharmacological effects on the recovery of consciousness [47, 57] (Fig. 12.1). The
central thalamic nuclei (CTN) seem particularly important in the emergence of con-
sciousness. They receive ascending projections coming from the brainstem encom-
passing the arousal systems that control the activity of many cortical and thalamic
neurons during the sleep–wake cycle. The CTN are strongly nerved by cholinergic,
serotoninergic, noradrenergic afferents of the arousal systems in the brainstem.
134 O. Gosseries et al.

Frontal cortex Parietal/occipital/temporal cortex

Striatum

Glob Central Glutamate

Dopamine palli us thalamus (lamotrigine)
dus
(amantadine,
levodopa,
bromocriptine,...)

Zolpidem
Excitation
Inhibition

Fig. 12.1 Model of the mesocircuit that aims to explain the mechanisms of pharmacologically
induced recovery of consciousness (Adapted from [47])

These same neurons of the CTN are also innervated by the downward projections
coming from the areas of the frontoparietal cortex. Collectively, these ascending
and descending pathways seem to modulate the level of consciousness [57]. The
frontoparietal cortex (and its subcortical modulation via striatum, globus pallidus,
and thalamus) is also prevalent for the emergence of consciousness. Thalamocortical
projections coming from the CTN activate in normal conditions the neurons of the
cortex and striatum. Lesions at this level result in a reduction of cerebral metabo-
lism. Neurons of the striatum inhibit the internal globus pallidus but require a strong
basic synaptic activity and elevated levels of dopaminergic innervations in order to
maintain their state in activity. Without projections of the striatum to the globus pal-
lidus (e.g., by a lack of dopaminergic innervations), the globus pallidus itself will
inhibit the CTN, which in turn will inhibit the cortical structures, and this sequence
could, thus, generate consciousness disorders. Disturbances in this mesocircuit
influence the total dynamics of the dominating corticothalamic and frontoparietal
systems [58]. Dopaminergic drugs could, therefore, facilitate projections of the
striatum on the globus pallidus, which would modulate the frontoparietal cortical
neurons and would restore the corticosubcortical loops. Zolpidem is thought to act
directly on the globus pallidus and would make it possible to inhibit it (as it is usu-
ally the case due to the action of the striatum), which would also restore the activity
of the CTN, whereas the glutamatergic agents (e.g., lamotrigine) would intervene
directly on the CTN (Fig. 12.1).
12 Pharmacological Treatments 135

Conclusion

No strong evidence currently supports or disproves the specific use of a pharmaco-

logical agent in order to improve the level of consciousness in DOC patients. As
discussed, a number of small and uncontrolled case or cohort studies have reported
a positive clinical response. Indeed, transitory or permanent improvements were
observed among some vegetative or minimally conscious patients of various etiolo-
gies (traumatic brain injury, cerebrovascular accident, anoxia, etc.). Reported effects
were variable, ranging from increase in wakefulness, partial recovery of conscious-
ness, motor, verbal, or communication functions to full recovery of cognitive
functioning.
Some of the reviewed treatments seem to benefit patients with severe disorders
of consciousness (e.g., amantadine, zolpidem, baclofen), whereas others (e.g.,
methylphenidate, lamotrigine) seem to possibly be more beneficial for brain-dam-
aged but conscious patients improving their attention deficit disorder [59]. The
effects were observed in the short- and long-term, and this, generated by a single
dose (e.g., zolpidem) or by a continuous treatment (e.g., amantadine, baclofen,
levodopa).
To date, the absence of double-blind placebo-controlled randomized multicentric
studies does not permit to propose any evidence-based therapeutic guidelines for a
specific diagnostic entity of DOC (vegetative versus minimally conscious state),
taking into account etiology and duration of the brain damage. Indeed, the reported
data come from case or cohort reports presenting methodological limitations and
not permitting to disentangle the drug effect from the natural evolution of the dis-
ease. Studies including a broader number of patients are especially influenced by
the extreme heterogeneity of DOC, such as the site of neuropathological lesions,
time elapsed between injury and the introduction of the treatment, confounding
drugs received, and medical comorbidities. Moreover, it is difficult to compare
between studies since they lack homogeneity in methodology and differ in the dura-
tion of treatment, administered doses, and patients’ demographics and clinical sta-
tus (Table 12.1). Measurement tools and behavioral scales are also very different
across studies (e.g., CRS-R, DRS, GCS, CNC, WSSNP) and a standardization of
bedside assessment seems necessary. The most adapted scale in this context is the
coma recovery scale-revised which has been specifically developed to distinguish
between vegetative and minimally conscious patients [60]. Obviously, placebo-con-
trolled, double-blind randomized multicentric studies are necessary before drawing
any conclusion on the effectiveness of therapeutic treatments in these challenging
patients.
Even if several hypotheses have been proposed, such as the mesocircuit model,
there is still a need to better understand the mechanisms and involved neurotransmit-
ters systems underlying residual neural plasticity and recovery from severe brain
damage. It would also be necessary to develop biomarkers of consciousness recov-
ery using neuroimagery or electrophysiology and to better define therapeutic
responders (e.g., genetic markers). It could also be useful (albeit methodologically
136 O. Gosseries et al.

more complicated) to test the efficacy of a combination of different pharmatherapeutic

agents and/or combinations with nonpharmacological rehabilitation interventions
and this by refining the timing of coercive interventions (therapy based on amount
and intensity). The nonpharmacological interventions such as physical and cognitive
rehabilitation as well as the sensory stimulation programs are complementary and
essential in the therapy planning of DOC patients. Finally, deep brain stimulation is
currently being investigated as a therapeutic option in some selected patients with
posttraumatic minimally conscious state but still remains in the research field and
should be more investigated before being implemented in the clinical routine [61].

References

1. Brown EN, Lydic R, Schiff ND. General anesthesia, sleep and coma. N Engl J Med.
2010;363(27):2638–50.
2. The Multi-Society Task Force on PVS. Medical aspects of the persistent vegetative state (1).
N Engl J Med. 1994;330(21):1499–508.
3. Giacino JT, Ashwal S, Childs N, et al. The minimally conscious state: definition and diagnostic
criteria. Neurology. 2002;58(3):349–53.
4. Bruno MA, Vanhaudenhuyse A, Thibaut A, Moonen G, Laureys S. From unresponsive wake-
fulness to minimally conscious PLUS and functional locked-in syndromes: recent advances in
our understanding of disorders of consciousness. J Neurol. 2011;258(7):1373–84.
5. Lombardi F, Taricco M, De Tanti A, Telaro E, Liberati A. Sensory stimulation of brain-injured
individuals in coma or vegetative state: results of a Cochrane systematic review, Clin Rehabil.
2002;16(5):464–72.
6. Zafonte R, Lexell J, Cullen N. Possible applications for dopaminergic agents following trau-
matic brain injury: part 2. J Head Trauma Rehabil. 2001;16(1):112–6.
7. Born JD. The Glasgow-Liège Scale. Prognostic value and evaluation of motor response and
brain stem reflexes after severe head injury. Acta Neurochir. 1988;95:49–52.
8. Saniova B, Drobny M, Kneslova L, et al. The outcome of patients with severe head injuries
treated with amantadine sulphate. J Neural Transm. 2004;111(4):511–4.
9. Folstein M, Robins L, Helzer J. The mini-mental state examination. Arch Gen Psychiatry.
1983;40(7):812.
10. Jennett B, Bond M. Assessment of outcome after severe brain damage. Lancet. 1975;
1(7905):480–4.
11. Rappaport M, Hall KM, Hopkins K, et al. Disability rating scale for severe head trauma: coma
to community. Arch Phys Med Rehabil. 1982;63(3):118–23.
12. Meythaler JM, Brunner RC, Johnson A, et al. Amantadine to improve neurorecovery in
traumatic brain injury-associated diffuse axonal injury: a pilot double-blind randomized trial.
J Head Trauma Rehabil. 2002;17(4):300–13.
13. Rappaport M. The coma/near coma scale. 2000. Retrieved 21 Aug 21, 2006, from http://www.
tbims.org/combi/cnc
14. Zafonte R, Watanabe T, Mann N. Amantadine: a potential treatment for the minimally con-
scious state. Brain Inj. 1998;12(7):617–21.
15. Giacino JT, Kalmar K, Whyte J. The JFK Coma Recovery Scale-Revised: measurement char-
acteristics and diagnostic utility. Arch Phys Med Rehabil. 2004;85(12):2020–9.
16. Schnakers C, Hustinx R, Vandewalle G, et al. Measuring the effect of amantadine in chronic
anoxic minimally conscious state. J Neurol Neurosurg Psychiatry. 2008;79(2):225–7.
17. Gualtieri T, Chandler M, Coons T, et al. Amantadine: a new clinical profile for traumatic brain
injury. Clin Neuropharmacol. 1989;12(4):258–70.
12 Pharmacological Treatments 137

18. Ansell BJ, Keenan JE. The Western Neuro Sensory Stimulation Profile: a tool for assessing
slow-to-recover head-injured patients. Arch Phys Med Rehabil. 1989;70(2):104–8.
19. Patrick P, Blackman J, Mabry J, et al. Dopamine agonist therapy in low-response children fol-
lowing traumatic brain injury. J Child Neurol. 2006;21(10):879–85.
20. Hughes S, Colantonio A, Santaguida P, et al. Amantadine to enhance readiness for rehabilita-
tion following severe traumatic brain injury. Brain Inj. 2005;19(14):1197–206.
21. Laureys S, Piret S, Ledoux D. Quantifying consciousness. Lancet Neurol. 2005;4(12):
789–90.
22. Haig A, Ruess J. Recovery from vegetative state of six months’ duration associated with
Sinemet (levodopa/carbidopa). Arch Phys Med Rehabil. 1990;71(13):1081–3.
23. Matsuda W, Matsumura A, Komatsu Y, et al. Awakenings from persistent vegetative state:
report of three cases with parkinsonism and brain stem lesions on MRI. J Neurol Neurosurg
Psychiatry. 2003;74(11):1571–3.
24. Matsuda W, Komatsu Y, Yanaka K, et al. Levodopa treatment for patients in persistent vegeta-
tive or minimally conscious states. Neuropsychol Rehabil. 2005;15(3–4):414–27.
25. Krimchansky B, Keren O, Sazbon L, et al. Differential time and related appearance of signs,
indicating improvement in the state of consciousness in vegetative state traumatic brain injury
(VS-TBI) patients after initiation of dopamine treatment. Brain Inj. 2004;18(11):1099–105.
26. Passler MA, Riggs RV. Positive outcomes in traumatic brain injury-vegetative state: patients
treated with bromocriptine. Arch Phys Med Rehabil. 2001;82(3):311–5.
27. Fridman E, Calvar J, Bonetto M, et al. Fast awakening from minimally conscious state with
apomorphine. Brain Inj. 2009;23(2):172–7.
28. Fridman E, Krimchansky B, Bonetto M, et al. Continuous subcutaneous apomorphine for
severe disorders of consciousness after traumatic brain injury. Brain Inj. 2010;24(4):636–41.
29. Chew E, Zafonte R. Pharmacological management of neurobehavioral disorders following
traumatic brain injury–a state-of-the-art review. J Rehabil Res Dev. 2009;46(6):851–79.
30. Moein H, Khalili H, Keramatian K. Effect of methylphenidate on ICU and hospital length of
stay in patients with severe and moderate traumatic brain injury. Clin Neurol Neurosurg.
2006;108(6):539–42.
31. Martin R, Whyte J. The effects of methylphenidate on command following and yes/no com-
munication in persons with severe disorders of consciousness: a meta-analysis of n-of-1 stud-
ies. Am J Phys Med Rehabil. 2007;86(8):613–20.
32. Gualtieri C, Evans R. Stimulant treatment for the neurobehavioural sequelae of traumatic brain
injury. Brain Inj. 1988;2(4):273–90.
33. Whyte J, Hart T, Vaccaro M, et al. Effects of methylphenidate on attention deficits after trau-
matic brain injury: a multidimensional, randomized, controlled trial. Am J Phys Med Rehabil.
2004;83(6):401–20.
34. Patrick P, Buck M, Conaway M, et al. The use of dopamine enhancing medications with chil-
dren in low response states following brain injury. Brain Inj. 2003;17(6):497–506.
35. Clauss RP, Guldenpfennig WM, Nel HW, et al. Extraordinary arousal from semi-comatose
state on zolpidem. A case report. S Afr Med J. 2000;90(1):68–72.
36. Clauss R, Nel W. Drug induced arousal from the permanent vegetative state. NeuroRehabilitation.
2006;21(1):23–8.
37. Hagen C, Malkmus D, Durham P. Levels of cognitive functioning. Downey: Rancho Los
Amigos Hospital Inc.; 1987.
38. Brefel-Courbon C, Payoux P, Ory F, et al. Clinical and imaging evidence of zolpidem effect in
hypoxic encephalopathy. Ann Neurol. 2007;62(1):102–5.
39. Cohen SI, Duong TT. Increased arousal in a patient with anoxic brain injury after administra-
tion of zolpidem. Am J Phys Med Rehabil. 2008;87(3):229–31.
40. Shames JL, Ring H. Transient reversal of anoxic brain injury-related minimally conscious state
after zolpidem administration: a case report. Arch Phys Med Rehabil. 2008;89(2):386–8.
41. Whyte J, Myers R. Incidence of clinically significant responses to zolpidem among patients
with disorders of consciousness: a preliminary placebo controlled trial. Am J Phys Med
Rehabil. 2009;88(5):410–8.
138 O. Gosseries et al.

42. Lo Y, Tan E, Ratnagopal P, et al. Zolpidem and its effects on hypoxic encephalopathy. Ann
Neurol. 2008;64(4):477–8.
43. Singh R, McDonald C, Dawson K, et al. Zolpidem in a minimally conscious state. Brain Inj.
2008;22(1):103–6.
44. Cohen L, Chaaban B, Habert MO. Transient improvement of aphasia with zolpidem. N Engl J
Med. 2004;350(9):949–50.
45. Clauss RP, Nel WH. Effect of zolpidem on brain injury and diaschisis as detected by 99mTc
HMPAO brain SPECT in humans. Arzneimittelforschung. 2004;54(10):641–6.
46. Hall S, Yamawaki N, Fisher A, et al. GABA(A) alpha-1 subunit mediated desynchronization
of elevated low frequency oscillations alleviates specific dysfunction in stroke–a case report.
Clin Neurophysiol. 2010;121(4):549–55.
47. Schiff N. Recovery of consciousness after brain injury: a mesocircuit hypothesis. Trends
Neurosci. 2010;33(1):1–9.
48. Kawecki Z, Kwiatkowski S, Grzegorzewski P, et al. Sudden improvement of all neurological
functions after general anesthesia and two-day intrathecal infusion of baclofen in a child with
primary brain-stem injury. Przegl Lek. 2007;64(2):13–4.
49. Sarà M, Sacco S, Cipolla F, et al. An unexpected recovery from permanent vegetative state.
Brain Inj. 2007;21(1):101–3.
50. Taira T, Hori T. Intrathecal baclofen in the treatment of post-stroke central pain, dystonia, and
persistent vegetative state. Acta Neurochir Suppl. 2007;97(Pt 1):227–9.
51. Sarà M, Pistoia F, Mura E, et al. Intrathecal baclofen in patients with persistent vegetative
state: 2 hypotheses. Arch Phys Med Rehabil. 2009;90(7):1245–9.
52. Taira T. Intrathecal administration of GABA agonists in the vegetative state. Prog Brain Res.
2009;177:317–28.
53. Meythaler J, Depalma L, Devivo M, et al. Sertraline to improve arousal and alertness in severe
traumatic brain injury secondary to motor vehicle crashes. Brain Inj. 2001;15(4):321–31.
54. Reinhard D, Whyte J, Sandel M. Improved arousal and initiation following tricyclic antide-
pressant use in severe brain injury. Arch Phys Med Rehabil. 1996;77(1):80–3.
55. Showalter P, Kimmel D. Stimulating consciousness and cognition following severe brain
injury: a new potential clinical use for lamotrigine. Brain Inj. 2000;14:997–1001.
56. Clauss R. Neurotransmitters in coma, vegetative and minimally conscious states, pharmaco-
logical interventions. Med Hypotheses. 2010;75(3):287–90.
57. Schiff N. Recovery of consciousness after severe brain injury: the role of arousal regulation
mechanisms and some speculation on the heart-brain interface. Cleve Clin J Med. 2010;77
Suppl 3:S27–33.
58. Laureys S, Schiff N. Coma and consciousness: paradigms (re)framed by neuroimaging.
Neuroimage. 2011 Dec 27. [Epub ahead of print]
59. Pistoia F, Mura E, Govoni S, et al. Awakenings and awareness recovery in disorders of con-
sciousness: is there a role for drugs? CNS Drugs. 2010;24(8):625–38.
60. American Congress of Rehabilitation Medicine, Brain Injury-Interdisciplinary Special Interest
Group, Disorders of Consciousness Task Force, Seel R, Sherer M, Whyte J, Katz D, Giacino
J, Rosenbaum A, Hammond F, Kalmar K, Pape T, Zafonte R, Biester R, Kaelin D, Kean J,
Zasler N. Assessment scales for disorders of consciousness: evidence-based recommendations
for clinical practice and research. Arch Phys Med Rehabil. 2010;91(12):1795–813.
61. Schiff ND, Giacino JT, Kalmar K, et al. Behavioural improvements with thalamic stimulation
after severe traumatic brain injury. Nature. 2007;448(7153):600–3.
Chapter 13
Deep Brain Stimulation

Marie Thonnard

Introduction

Chronic vegetative state/unresponsive wakefulness and minimally conscious state

are disorders of consciousness (DOC) that can follow severe brain lesions [1]. These
states are currently relatively well-defined but their therapeutic options remain rela-
tively limited. Reports on patients having received noninvasive (generally pharma-
cological) treatment are mainly uncontrolled. Often these are case reports and the
methodology is frequently suboptimal to provide empirical evidences for the treat-
ment efficiency. However, while invasive, one therapeutic technique is increasingly
studied: deep brain stimulation (DBS). This technique is used as a treatment in dif-
ferent neurological diseases but particularly in Parkinson’s disease, for which tech-
nological advances have allowed the treatment of a large number of patients. Thanks
to these advances, studies related to DBS as a treatment for DOC could be devel-
oped. Even though methodological difficulties are evident, results are promising.
We propose in this chapter a description of the DBS technique, an outline of previ-
ous clinical applications, review of studies using this technique in DOC patients
along with a summary of mechanisms explaining the efficacy of DBS for treating
these patients.

M. Thonnard, M.Sc.
Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du 6 août, 8, Sart-Tilman B30, 4000 Liège, Belgium
Department of Neurology, University Hospital of Liège,
Liège, Belgium
e-mail: [email protected]

C. Schnakers, S. Laureys (eds.), Coma and Disorders of Consciousness, 139

DOI 10.1007/978-1-4471-2440-5_13, © Springer-Verlag London 2012
140 M. Thonnard

History and Clinical Applications

DBS is a treatment aiming to stimulate target brain structures of one or both hemi-
sphere using surgically implanted electrodes and an electrical pulse generator, also
called “brain pacemaker” or “neurostimulator.” Bartholow [2] was the first to report
the use of this technique in humans. DBS was first used in order to localize and map
functions of particular cerebral regions [3], and, later, to investigate the effects of
stimulation on deep brain structures [4]. In the 1960s emerged the idea that it is pos-
sible to treat neurological disease with this technique, but it was only in the 1970s
that the first trials using DBS were reported, particularly in patients with chronic
pain [5], epilepsy [6], or movement disorders [7]. Thanks to technological advances
of neurostimulators that can be surgically implanted, first trials about the long-term
use of DBS were reported in the 1990s [8, 9]. Since then, this technique has been
used to treat several neurological and psychiatric disorders such as motor disorders
(e.g., essential tremor, dystonia, Parkinson’s disease), Tourette’s syndrome, chronic
pain, or obsessive-compulsive disorders.
Nowadays, DBS is used in the treatment of several pathologies. Notably, the
stimulation of ventral intermediate nucleus of the thalamus seems to be efficient in
the treatment of essential tremor [9–12]. Concerning the treatment of dystonia, the
stimulation of the globus pallidus and of the thalamus has shown some efficiency
[13–15]. Several target brain areas are used for the treatment of Parkinson’s disease,
with diverse effects according to the target regions stimulated. Whereas the stimula-
tion of the ventral intermediate nucleus significantly reduces tremor [16, 17], the
stimulation of the subthalamic nucleus and of the globus pallidus also provokes a
decrease of bradykinesia and hypertonia [18–22].

DBS and DOC Treatment

Previous Studies

In the literature, there are few studies on DBS in patients with DOC. Although some
methodological bias could be discussed, studies performed in this area have allowed
a better understanding of DBS mechanisms in DOC. Early animal studies by
Moruzzi and Magoun [23] have shown that the stimulation of the midbrain reticular
formation leads to electroencephalogram desynchronizations, similar to those
observed in wakefulness states. Later, other studies have shown the relation between
the midbrain reticular formation and the level of vigilance in both animals [24] and
humans [25]. The first DBS studies in vegetative patients focused on the reticular
formation, the basal ganglia, and especially, on the thalamus. Sturm et al. [26],
Hassler et al. [27], and McLardy et al. [28] were the first to report clinical results
using DBS in intralaminar and rostral-ventral thalamic nuclei in unconscious
patients. However, their clinical results and the long-term follow-up were incomplete
13 Deep Brain Stimulation 141

and suboptimal. In the 1980s, Medtronic Inc. (Minneapolis, Minnesota, USA)

initiated a multicentric study involving French, Japanese, and American teams.
Cohadon [29] have studied the effect of DBS on parafascicular intralaminar and
centromedian thalamic nuclei in 25 patients who were in a vegetative state since at
least 3 months. In this cohort, 12 patients showed no clinical or behavioral improve-
ments in a follow-up period of 1–12 years. The remaining 13 patients manifested a
significant improvement of the level of consciousness and of interpersonal interac-
tions after 1–3 weeks of treatment. In a study by Tsubokawa et al. [30], the authors
have stimulated the midbrain reticular formation (cuneiform nucleus) and unspecific
thalamic nuclei (median-parafascicular complex) in eight vegetative patients
(2–3 months post injury). Patients were assessed with behavioral and electrophysi-
ological measures (e.g., electroencephalogram, brainstem auditory evoked poten-
tials, and somatosensory evoked potentials) and behavioral improvements were
reported in three cases (i.e., response to command and vocalizations/verbalizations)
with a partial improvement noticed in a fourth patient (i.e., recovery of oral feeding
and emotional expressions). The remaining four patients did not show any improve-
ment after 1 year of DBS. However, even though some changes have been observed
directly after the activation of the DBS device (e.g., more frequent eye and mouth
opening), most behavioral improvements were recorded only after 3–4 months of
treatment.
Other studies by Katayama et al. [31] and Hosobuchi and Yingling [32] have
shown similar improvements in vegetative patients. However, in both studies, the
observed improvements could also be explained, entirely or partially, by spontane-
ous recovery. Indeed, the reported improvements appeared within a 1 year postin-
jury period, during which a spontaneous recovery is most likely to appear. In
addition, the authors did not use a placebo-controlled double-blind design or a
period of DBS stimulation withdrawal. Finally, criteria of minimally conscious state
having been defined only a few years ago [33], some patients could have met, a
posteriori, the criteria of MCS.
Yamamoto et al. [34] have studied the effect of DBS in stimulating the midbrain
reticular formation and the median-parafascicular complex of the thalamus in 21
vegetative patients and 5 minimally conscious patients. Eight vegetative patients
were reported to have emerged and were able to respond to simple commands. Four
out of five minimally conscious patients also had emerged (showing recovery of
functional communication) and could return living at home.
Most often, results in truly “vegetative” patients did not allow to unequivocally
concluding on the efficiency of DBS. In part, this could be explained by the massive
brain lesions presented by vegetative patients. Indeed, in nontraumatic etiology, dif-
fuse ischemic damages of the neocortex and severe thalamic injuries are classically
encountered. In posttraumatic patients, diffuse axonal injury as well as thalamic
lesions are frequently observed. In both etiologies, subcortical lesions of white mat-
ter with thalamic deafferentations can be found. Brain structures with possible
residual function may not be able to function because of corticocortical and thal-
amocortical disconnections [35]. Thus, even though DBS of the thalamus in vegeta-
tive patients are aimed to enhance the level of wakefulness and consciousness by
142 M. Thonnard

activating respective neuronal networks, it would be efficient only in rare cases.

Indeed, DBS cannot have a significant effect on a widespread network via the thala-
mus if the connections within this network or between the thalamus and this net-
work are disrupted. For traumatic lesions, often involving multifocal lesions (and
thus the preservation of a larger set of neuronal connections), DBS could be more
efficient. In 2007, Schiff et al. [36] have reported the case of a chronic posttraumatic
patient treated with DBS of thalamic intralaminar nuclei in a double-blind design
with recording of several baselines. Clinically, the patient was in a minimally con-
scious state and was 6 years postinjury, decreasing the possibility of a spontaneous
recovery. In addition, the patient did not show any improvement despite several
previous rehabilitation programs. Behavioral changes were assessed with the coma
recovery scale-revised [37] and with objects denomination, motor limb control, and
oral feeding tests. Improvements appeared after the placement of the neurostimula-
tor. Intelligible verbalizations and functional object use were already reported dur-
ing the titration phase. Later, responses to command, spontaneous limb movements,
oral feeding, and, especially, a recovery of functional communication were
observed.
The methodology of this last case report seems to compensate the bias observed
in previous studies. Indeed, the authors have used several baselines to be sure that
the observed changes were linked to the presence of the treatment and could not be
attributed to the ongoing rehabilitation program. Besides, for such a chronic patient,
the probability of changes due to a spontaneous recovery was very small (but not
inexistent) [38]. Importantly, brain imaging in this patient showed neither lesions in
language-related regions nor lesions in the thalamus.

Mechanisms Involved in DOC Treatment Using DBS

Consciousness seems to be underlined by a large network of cortical regions

(e.g., frontoparietal) but also by thalamo-cortical and cortico-cortical connectivity
[39, 40]. Damages within this network can lead to DOC. Indeed, in the minimally
conscious state, this network would be more preserved than in the vegetative state.
The thalamus plays an important role in consciousness processing as it has many
connections with this fronto-parietal “awareness-network”. Hence, a thalamic lesion
can be responsible for severe DOC. The aim of stimulating the thalamus would be
to reactivate connections within this network in compensating for the lack of thal-
amic activation. Thereby, DBS of the thalamus would have a higher efficacy in
patients with a relatively preserved thalamocortical axonal network.
Within the thalamus, intralaminar nuclei seem to be particularly connected with
higher level cortical areas. Previous studies have shown selective vulnerability of
neurons within central nuclei of the thalamus. Indeed, cellular loss in these particu-
lar nuclei would be strongly associated with the patients’ level of recovery [35, 41].
Whereas the level of functioning is higher in patients with lesions of anterior and
median central nuclei, an additional damage in intralaminar regions often results in
13 Deep Brain Stimulation 143

a severe inability or even a vegetative state. The central nuclei of the thalamus, and
particularly the intralaminar nuclei, seem therefore to play a central role in DOC
pathology. The central nuclei seem to play a key-role in wakefulness and vigilance
functions as regards their high number of connections with the brainstem and the
basal midbrain [42]. In addition, activity in central nuclei of the thalamus could vary
according to the level of vigilance and according to the cognitive workload required
to achieve a task [25], explaining performance fluctuations usually observed
in DOC.

Intralaminar Nuclei of the Thalamus as a Target for DBS

In the treatment of DOC, the target area for DBS that would offer the best results
seems to be the intralaminar nuclei of the thalamus and this for several reasons
[43]. First, the central thalamus is supposed to be altered as regards pathophysio-
logical mechanisms linked to traumatic brain injury, which are diffuse axonal dam-
age, downward displacement of the thalamus and the brainstem; and/or ischemic
lesions of the brainstem, the midbrain, the thalamus, and of the basal ganglia.
Indeed, diffuse axonal lesions lead to axonal disconnections within a large brain
network, generating an extensive neuronal death. As mentioned earlier, cellular
loss in central thalamus seems to be particularly associated with DOC patients’
level of recovery [35, 41]. Then, a traumatic brain injury generally leads to a
decrease in forebrain activity and that partly because of the role of central thalamic
nuclei. As mentioned earlier, central thalamic neurons are important in attention
and wakefulness regulations. These neurons could be monitored by the wakeful-
ness level generated by ascending brainstem connections [44], but also by medial
frontal regions that calibrate the level of cortical activation during cognitive tasks
[25]. Thereby, lesions of these regions could decrease the activity in the ipsilateral
hemisphere and cause bilateral metabolic dysfunctions. Finally, DBS seems to
generate action potentials in intact axonal fibers. Central and intralaminar nuclei
neurons release an excitatory neurotransmitter, glutamate, which provokes a depo-
larization of striatum and cortex. When the striatum is activated, this can produce
a large pattern of brain activations. Thus, in compensating for neuronal activity
in the central thalamus, the stimulation of these neurons using DBS could facilitate
the induction and maintaining the activity in a large network of neurons through
the entire brain and thus lead to the recovery of cognitive functions underlined
by these networks (e.g., attention, memory, language, executive functions, etc.).
Finally, a sustained DBS of intralaminar thalamic nuclei could have long-term
behavioral effects, such as those observed in the study of Schiff et al. [36]. Indeed,
in this patient initially diagnosed as being in a minimally conscious state, improve-
ments persisted during the stimulation withdrawal period. Similar changes have
also been reported in animals treated with repetitive DBS. These long-term effects
could reflect neuroplasticity phenomena that await further characterization and
understanding.
144 M. Thonnard

Conclusion

The controlled study of DBS as a potential treatment for DOC is relatively recent.
Evolving in parallel to technological advances and scientific discoveries, studies
about DBS in DOC tend to show promising results. However, older studies present
methodological bias considering present knowledge, particularly as regards patients’
selection, target area for DBS and methodological design. Previous studies have
nevertheless allowed highlighting several criteria for the DBS use in DOC patients:
(1) selection of posttraumatic patients, nontraumatic lesions generally affecting a
larger number of brain regions and connections; (2) selection of patients at least
1 year post injury, in order to decrease the effect of spontaneous recovery; (3) selec-
tion of patients in a minimally conscious state, vegetative patients usually showing
widespread axonal injury and larger brain disconnections; (4) selection of patients
showing no lesions in the thalamus or in regions crucial to consciousness recovery
(e.g., frontoparietal and language regions); (5) the use of a valid experimental design
(e.g., double-blind and baseline measurements); (6) a rehabilitation program main-
tained and unchanged during the DBS trial, in order to verify that improvements are
due to stimulation and not to the rehabilitation program. The study by Schiff et al.
[36] has notably allowed showing DBS-related improvements taking into account
all recommendations aforementioned. Nevertheless, it remains a case report that
does not allow generalizing results to a larger population of patients. Therefore, the
application of such an invasive methodology in a bigger sample is necessary to vali-
date the DBS use as a treatment for DOC, taking into account the challenging ethi-
cal issues involved [45].

References

1. Laureys S, Owen AM, Schiff ND. Brain function in coma, vegetative state, and related disor-
ders. Lancet Neurol. 2004;3:537–46.
2. Bartholow R. Experimental investigations into the functions of the human brain. Am J Med
Sci. 1874;67:305–16.
3. Penfield W, Boldrey E. Somatic motor and sensory representation in the cerebral cortex of man
as studied by electrical stimulation. Brain. 1937;60:389–443.
4. Spiegel EA, Wycis HT, Marks M, Lee AJ. Stereotaxic apparatus for operations on the human
brain. Science. 1947;106:349–50.
5. Hosobuchi Y, Adams JE, Rutkin B. Chronic thalamic stimulation for the control of facial anes-
thesia dolorosa. Arch Neurol. 1973;29:158–61.
6. Cooper IS. Effect of chronic stimulation of anterior cerebellum on neurological disease.
Lancet. 1973;1:206.
7. Brice J, McLellan L. Suppression of intention tremor by contingent deep-brain stimulation.
Lancet. 1980;1:1221.
8. Benabid AL, Pollak P, Gervason C, et al. Long-term suppression of tremor by chronic stimula-
tion of the ventral intermediate thalamic nucleus. Lancet. 1991;337:403–6.
9. Benabid AL, Pollak P, Gao D, et al. Chronic electrical stimulation of the ventralis intermedius
nucleus of the thalamus as a treatment of movement disorders. J Neurosurg. 1996;84:203–14.
13 Deep Brain Stimulation 145

10. Koller WC, Lyons KE, Wilkinson SB, Pahwa R. Efficacy of unilateral deep brain stimulation
of the VIM nucleus of the thalamus for essential head tremor. Mov Disord. 1999;14:
847–50.
11. Ondo W, Jankovic J, Schwartz K, et al. Unilateral thalamic deep brain stimulation for refrac-
tory essential tremor and Parkinson’s disease tremor. Neurology. 1998;51:1063–9.
12. Rehncrona S, Johnels B, Widner H, et al. Long-term efficacy of thalamic deep brain stimula-
tion for tremor: double-blind assessments. Mov Disord. 2003;18:163–70.
13. Kumar R, Dagher A, Hutchison WD, et al. Globus pallidus deep brain stimulation for general-
ized dystonia: clinical and PET investigation. Neurology. 1999;53:871–4.
14. Sellal F, Hirsch E, Barth P, et al. A case of symptomatic hemidystonia improved by ventropos-
terolateral thalamic electrostimulation. Mov Disord. 1993;8:515–8.
15. Lozano AM, Kumar R, Gross RE, et al. Globus pallidus internus pallidotomy for generalized
dystonia. Mov Disord. 1997;12:865–70.
16. Kumar R, Lozano AM, Sime E, Lang AE. Long-term follow-up of thalamic deep brain stimu-
lation for essential and parkinsonian tremor. Neurology. 2003;61:1601–4.
17. Putzke JD, Wharen Jr RE, Wszolek ZK, et al. Thalamic deep brain stimulation for tremor-
predominant Parkinson’s disease. Parkinsonism Relat Disord. 2003;10:81–8.
18. Anderson VC, Burchiel KJ, Hogarth P, et al. Pallidal vs subthalamic nucleus deep brain stimu-
lation in Parkinson disease. Arch Neurol. 2005;62:554–60.
19. Peppe A, Pierantozzi M, Altibrandi MG, et al. Bilateral GPi DBS is useful to reduce abnormal
involuntary movements in advanced Parkinson’s disease patients, but its action is related to
modality and site of stimulation. Eur J Neurol. 2001;8:579–86.
20. Loher TJ, Burgunder JM, Weber S, et al. Effect of chronic pallidal deep brain stimulation on
off period dystonia and sensory symptoms in advanced Parkinson’s disease. J Neurol Neurosurg
Psychiatry. 2002;73:395–9.
21. Burchiel KJ, Anderson VC, Favre J, Hammerstad JP. Comparison of pallidal and subthalamic
nucleus deep brain stimulation for advanced Parkinson’s disease: results of a randomized,
blinded pilot study. Neurosurgery. 1999;45:1375–82.
22. Kumar R, Lozano AM, Montgomery E, Lang AE. Pallidotomy and deep brain stimulation of
the pallidum and subthalamic nucleus in advanced Parkinson’s disease. Mov Disord. 1998;13
Suppl 1:73–82.
23. Moruzzi G, Magoun HW. Brain stem reticular formation and activation of the EEG.
Electroencephalogr Clin Neurophysiol. 1949;1:455–73.
24. Fuster JM. Effects of stimulation of brain stem on tachistoscopic perception. Science.
1958;127:150.
25. Kinomura S, Larsson J, Gulyas B, Roland PE. Activation by attention of the human reticular
formation and thalamic intralaminar nuclei. Science. 1996;271:512–5.
26. Sturm V, Kuhner A, Schmitt HP, et al. Chronic electrical stimulation of the thalamic unspecific
activating system in a patient with coma due to midbrain and upper brain stem infarction. Acta
Neurochir (Wien). 1979;47:235–44.
27. Hassler R, Ore GD, Dieckmann G, et al. Behavioural and EEG arousal induced by stimulation
of unspecific projection systems in a patient with post-traumatic apallic syndrome.
Electroencephalogr Clin Neurophysiol. 1969;27:306.
28. McLardy T, Ervin F, Mark V, et al. Attempted inset-electrodes-arousal from traumatic coma:
neuropathological findings. Trans Am Neurol Assoc. 1968;93:25–30.
29. Cohadon F. Deep brain stimulation in cases of prolonged traumatic unconsciousness. In:
Lazorthes Y, Upton ARM, editors. Neurostimulation: an overview. New York: Futura
Publishers; 1985.
30. Tsubokawa T, Yamamoto T, Katayama Y, et al. Deep-brain stimulation in a persistent vegeta-
tive state: follow-up results and criteria for selection of candidates. Brain Inj. 1990;4:315–27.
31. Katayama Y, Tsubokawa T, Yamamoto T, et al. Characterization and modification of brain
activity with deep brain stimulation in patients in a persistent vegetative state: pain-related late
positive component of cerebral evoked potential. Pacing Clin Electrophysiol. 1991;14:
116–21.
146 M. Thonnard

32. Hosobuchi Y, Yingling C. The treatment of prolonged coma with neurostimulation. In:
Devinsky O, Beric A, Dogali M, editors. Electrical and magnetic stimulation of the brain and
spinal cord. New York: Raven; 1993. p. 247–52.
33. Giacino JT, Ashwal S, Childs N, et al. The minimally conscious state: definition and diagnostic
criteria. Neurology. 2002;58:349–53.
34. Yamamoto T, Kobayashi K, Kasai M, et al. DBS therapy for the vegetative state and minimally
conscious state. Acta Neurochir Suppl. 2005;93:101–4.
35. Adams JH, Graham DI, Jennett B. The neuropathology of the vegetative state after an acute
brain insult. Brain. 2000;123(Pt 7):1327–38.
36. Schiff ND, Giacino JT, Kalmar K, et al. Behavioural improvements with thalamic stimulation
after severe traumatic brain injury. Nature. 2007;448(7153):600–3.
37. Giacino JT, Kalmar K, Whyte J. The JFK Coma Recovery Scale-Revised: measurement char-
acteristics and diagnostic utility. Arch Phys Med Rehabil. 2004;85:2020–9.
38. Lammi MH, Smith VH, Tate RL, Taylor CM. The minimally conscious state and recovery
potential: a follow-up study 2 to 5 years after traumatic brain injury. Arch Phys Med Rehabil.
2005;86:746–54.
39. Laureys S, Berré J, Goldman S. Cerebral function in coma, vegetative state, minimally con-
scious state, locked-in syndrome and brain death. In: Vincent JL, editor. Yearbook of intensive
care and emergency medicine. Berlin: Springer; 2001. p. 386–96.
40. Laureys S, Goldman S, Phillips C, et al. Impaired effective cortical connectivity in vegetative
state: preliminary investigation using PET. Neuroimage. 1999;9:377–82.
41. Maxwell WL, MacKinnon MA, Smith DH, et al. Thalamic nuclei after human blunt head
injury. J Neuropathol Exp Neurol. 2006;65:478–88.
42. Krout KE, Belzer RE, Loewy AD. Brainstem projections to midline and intralaminar thalamic
nuclei of the rat. J Comp Neurol. 2002;448:53–101.
43. Schiff ND. Central thalamic deep-brain stimulation in the severely injured brain: rationale and
proposed mechanisms of action. Ann N Y Acad Sci. 2009;1157:101–16.
44. Parvizi J, Damasio A. Consciousness and the brainstem. Cognition. 2001;79:135–60.
45. Fins JJ. Neuroethics and neuroimaging: moving toward transparency. Am J Bioeth.
2008;8(9):46–52.
Chapter 14
The Ethics of Managing Disorders
of Consciousness

Athena Demertzi, Olivia Gosseries, Marie-Aurélie Bruno,

Caroline Schnakers, Audrey Vanhaudenhuyse, Camille Chatelle,
Vanessa Charland-Verville, Aurore Thibaut, Marie Thonnard,
and Steven Laureys

Introduction

The introduction of the mechanical ventilator in the 1950s and the development of
intensive care in the 1960s made it possible to sustain many patients’ vegetative
functions and allowed them to survive their severe injuries. Despite such advance-
ment, in many cases, patients were found to suffer from altered states of conscious-
ness, which had never been encountered before as these patients would normally
die from apnea [1]. The imminent ethical impact of these profound states of
unconsciousness was reflected in the composition of the first bioethical committees

A. Demertzi, M.Sc. (*) • M.-A. Bruno, Ph.D. • C. Schnakers, Ph.D. • A. Vanhaudenhuyse, Ph.D.
• C. Chatelle, M.Sc. • V. Charland-Verville, M.Sc. • A. Thibaut, M.Sc. • M. Thonnard, M.Sc. •
S. Laureys, M.D., Ph.D.
Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du 6 Août, Sart Tilman B30, 4000 Liège, Belgium
Department of Neurology, CHU University Hospital of Liège, Liège, Belgium
e-mail: [email protected]; [email protected];
[email protected]; [email protected]; [email protected];
[email protected]; [email protected]; [email protected]; steven.
[email protected]
O. Gosseries, M.Sc.
Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du 6 Août, Sart Tilman B30, 4000 Liège, Belgium
Department of Neurology, CHU University Hospital of Liège, Liège, Belgium
National Fund for Scientific Research, Liège, Belgium
e-mail: [email protected]

C. Schnakers, S. Laureys (eds.), Coma and Disorders of Consciousness, 147

DOI 10.1007/978-1-4471-2440-5_14, © Springer-Verlag London 2012
148 A. Demertzi et al.

discussing the redefinition of life and the concept of therapeutic obstinacy. In 1968,
the Ad Hoc Committee of Harvard Medical School published a milestone paper for
the redefinition of death as irreversible coma and brain failure [2]. The committee
mainly included several physicians, a theologian, a lawyer, and a historian of sci-
ence, anticipating the medical, legal, and societal debates that were to follow. We
will here give a brief overview of some ethical issues related to the concept of con-
sciousness and the medical management of patients with disorders of consciousness
(DOC) such as comatose, vegetative, and minimally conscious states that may be
encountered in the intensive care setting. We will emphasize the problem of pain
management and end-of-life decision-making.

Ethical Issues on Clinical Management

Confusions and controversies are often related to the way we define things. One
such multifaceted term is consciousness, which has many divergent connotations
[3]. The way we define consciousness is crucial, as it may govern our attitudes
toward medical management of DOC. For example, in a wide survey (n = 1,858)
among medical and paramedical professionals, we recently found that although the
majority of healthcare workers denied the distinction between consciousness and
the brain, more than one-third of medical and paramedical professionals still
regarded mind and brain as separate entities. Such dualistic opinions may have
implications to the formulation of scientific questions about the nature of con-
sciousness, to clinical management of DOC, and to the reception of both by the
general public [4]. We here adopt a perspective where consciousness is clinically
defined as having two components, wakefulness and awareness [5]. Under this
definition, many variant altered states of consciousness may be included. The most
familiar of all is the transition from conscious wakefulness to deep sleep; the
drowsier we become, the less aware we get of our surroundings and of ourselves.
This implies that patients in coma and under anesthesia (i.e., pharmacological
coma) are unaware because they cannot be awakened, even after noxious stimula-
tion. The vegetative state (VS), now called unresponsive wakefulness syndrome
(UWS), is defined as “wakefulness without awareness” where patients may open
their eyes but never will exhibit nonreflex voluntary movements [6]. A patient in a
minimally conscious state (MCS) may show some signs stating awareness (e.g.,
visual pursuit, orientation to pain, or nonsystematic command following) but is
unable to communicate her or his thoughts or feelings [7]. Because these behav-
ioral signs of consciousness often are small and fluctuating in time, this condition
may be challenging to diagnose and differentiate from VS [8]. It has been sug-
gested that once conscious awareness has been detected in a noncommunicative
patient [9, 10], this may be a good reason to preserve life-sustaining aids [11].
However, the moral significance of preserved consciousness has been questioned
on the grounds that it may not always be in patients’ best interest to continue a
severely disabled life [12].
14 The Ethics of Managing Disorders of Consciousness 149

One challenging issue is the conscious perception of pain in these patients. As

defined by the Multi-Society Task Force on PVS, “pain and suffering refer to the
unpleasant experiences that occur in response to stimulation of peripheral nocicep-
tive receptors and their peripheral and central afferent pathways or that they may
emanate endogenously from the depths of human self-perception” [13]. Thus, pain
constitutes a conscious experience with a physical (nociception) and a psychological
counterpart (suffering), suggesting that nociception by itself is not sufficient to cause
suffering. The management of pain in patients with DOC is challenging because
patients in VS/UWS or MCS cannot verbally or nonverbally communicate their feel-
ings or experiences [1]. This is reflected on how clinicians perceive pain in these
patients. According to recently surveyed attitudes among healthcare professionals, a
majority of respondents supported that patients in MCS (96%) perceive pain whereas
opinions were less clear for the patients in VS/UWS (56%) [14]. Considering these
results on varying beliefs about pain perception in DOC, physicians and healthcare
workers’ views on analgesia and symptom management may also be affected. Since
nearly half of the surveyed doctors expressed that VS/UWS patients do not feel pain,
they could be expected to act accordingly, e.g., by not providing analgesic medica-
tion in these patients during care or during the dying process after withdrawal of
artificial hydration and nutrition [15], the latter on the grounds that these patients
may not experience suffering from hunger or thirst [16].
How are clinicians supposed to infer whether VS/UWS or MCS patients feel
pain or suffering? At the patient’s bedside, we are limited to evaluate the behavioral
responsiveness to pain: if patients show no signs of voluntary movement (i.e., local-
izing the source of pain) in response to a noxious stimulus, it may be concluded they
do not experience pain. Conscious but paralyzed “locked-in syndrome” patients,
classically showing no movements or decerebrate (i.e., stereotyped extension)/
decorticate (i.e., stereotyped flexion) posturing, illustrate it is not necessarily the
case. In response to noxious stimulation, DOC patients will frequently show
increased arousal levels (evidenced by opening or widening of the eyes), increased
respiratory/heart rate and blood pressure, or grimace-like/crying-like behavior. As
all these responses are also seen in infants with anencephaly [17], they are consid-
ered to be of subcortical origin and not necessarily reflecting conscious perception
of pain. However, the absence of one of these behavioral responses cannot be taken
as a proof of the absence of conscious perception [18], and the inference of pain and
suffering merely by observing behavioral responses may be misleading. Repeated
clinical examinations performed by experienced examiners using standardized tools
such as the recently proposed “Nociception Coma Scale” [19] are essential for the
behavioral assessment of pain. Additional information coming from functional
neuroimaging studies may help in the formulation of a clearer clinical picture.
For example, in a positron emission tomography (PET) study, it was shown that
VS/UWS patients may show cerebral processing of the noxious stimulus (activation
of primary somatosensory areas) but the observed neural activity was isolated and
disconnected from higher-order associative brain areas, which are considered nec-
essary for conscious perception of pain [20]. It is important to stress that very dif-
ferent results were obtained in MCS patients where functional neuroimaging studies
150 A. Demertzi et al.

have shown a cerebral activation in a widespread cerebral network as compared to

unresponsive patients, but similar to healthy controls, suggesting a potential pain
perception in these patients [21]. In light of an incomplete pain perception in
VS/UWS patients, the existing risk for misdiagnosis [8], the inconclusive drug-
related effects in DOC [22] and the limitations of interpreting neuroimaging
results [23], pain prophylaxis and drug treatment have been proposed for all patients
suffering from DOC [24].
In intensive care settings, medical doctors and assisting staff are daily confronted
with situations where clinical decisions are critical, such as continuing or withdraw-
ing life-sustaining treatment. Treatment limitations can be different, depending on
whether the decision is made preoperatively or after an intervention [25]. In the for-
mer case, it may come as a refusal of cardiopulmonary resuscitation (CPR) in case of
cardiopulmonary arrest; in the latter case, it most usually comes as a decision to
withdraw treatment such as the artificial respirator or artificial nutrition and hydra-
tion (ANH). CPR is almost automatically performed as an emergency therapy in
order to restore heart beat and ceased breathing, unless the patient or the legal repre-
sentative has refused it in advance as a do-not-resuscitate (DNR) order. Nevertheless,
it should be noted that DNR orders do not necessarily prohibit other therapies; they
rather authorize the physician to act on this specific therapy [26]. When the clinical
condition of a patient has been stabilized and denoted as irreversible, decisions about
ANH limitation may be considered. From a bioethical standpoint, withdrawing ANH
is comparable to withdrawing mechanical ventilation, even if emotionally they may
be perceived differently. In the ICU setting, the majority of deaths are the result of a
medical decision to withhold or withdraw treatment [27]. Such decisions are evi-
dence-based and rely on validated clinical or paraclinical markers of bad outcome
[28]. Despite the controversy as to whether ANH constitutes a medical treatment
[29], and thus should never be withdrawn from patients [30], most of the medical
community (especially Anglo-Saxon) agrees on considering it as a medical therapy
which can be refused by patients and surrogate decision makers [31]. Such decisions
in the VS/UWS are only justified when a case is considered as irreversible [32]. Up
to date, guidelines defining a definitive outcome in VS/UWS is that, if no recovery is
observed within 3 months after a nontraumatic injury or 12 months after a traumatic
injury, the condition of the patient can be considered as permanent [13].
The controversies around the clinical management at the end-of-life in DOC
patients were reflected in a recent European survey (n = 2,475), where the majority
of healthcare professionals (66%) agreed to withdraw treatment from chronic
VS/UWS patients whereas only 28% agreed so for the chronic MCS; additionally,
most clinicians wished not to be kept alive if they imagined themselves in a chronic
VS (82%) and similar proportion (67%) agreed so if they imagined themselves in a
chronic MCS [33]. Geographical region and religion were among the factors that
explained most of the variance in the responses [33]. The detected differences
between the two states could be due to the existing legal ambiguity around MCS,
which may have influenced the surveyed participants to differentiate between
expressing preferences for self versus others, by implicitly recognizing that the lat-
ter could be a step on the slippery slope to euthanasia.
14 The Ethics of Managing Disorders of Consciousness 151

Clinicians’ opinions appear much more uniform with regard to brain death [34].
As mentioned earlier, the Ad Hoc Committee of the Harvard Medical School went
on to redefine death as a consequence of the technological advancements in the
intensive care, where patients could sustain their severe injuries but maintain the
function of vital organs [2]. It was, hence, possible to dissociate between cardiac,
respiratory, and brain functions, which in turn required an alternative definition of
death, moving from a cardiorespiratory toward a neurocentric formulation (i.e., irre-
versible coma). According to the latter, death can be viewed either as death of the
whole brain or of the brainstem [35] or as neocortical [36]. The first two are defined
as the irreversible cessation of the organism as a whole, differing in their anatomical
interpretation [37], whereas the last solely requires the irreversible loss of the capac-
ity of consciousness and social interaction but has never convinced medical or legal
scholars. The main utility of the introduction of brain death is that it permitted vital
organ procurement for transplantation with the application of ethical restrictions,
such as the dead donor rule (i.e., a patient has to be declared dead before the removal
of life-sustaining organs). Based on the neocortical definition of death, however,
both VS/UWS and MCS patients can be declared dead. It has been argued that the
neocortical definition is conceptually inadequate and practically unfeasible, espe-
cially in lack of a complete understanding of higher-order conscious functioning;
hence, DOC patients are not dead [27] and organ donation options in these patients
should be excluded since they violate the dead donor rule [38]—despite opposing
opinions to abandon this ethical axiom [39].

Legal Issues in Disorders of Consciousness

Managing patients with disorders of consciousness may require the mediation of

legal authorities in order to regulate ambiguous and controversial issues, such as end-
of-life decisions. When end-of-life wishes have not been earlier formulated as an
advance directive (i.e., written statement completed by a competent person in antici-
pation of her/his future incompetence, expressing personal treatment preferences and
formal surrogacy appointment), then a surrogate decision maker is eligible to take
responsibility of the patient’s clinical management. The way the legal representative
should act on behalf of the patient is a progressive one. The surrogate should first
attempt to follow the wishes of the patient as closely as possible the way they were
expressed before the accident, either orally or as advance directives. When the wishes
are unknown and an advance directive is not available, the surrogate decision maker
should try to reproduce the patients’ preferences based on their history and personal
values. When this is not possible the decisions should rely on more objective markers
that determine the patients’ best interest (e.g. likelihood of recovery, pain manage-
ment, impact on family) [25, 40]. The proxy decision maker should mediate trying to
maximize patients’ self-determination and protect their interests on the principles of
beneficence and nonmaleficence (i.e., ongoing treatment may be judged to be futile
on the basis of low likelihood of significant recovery) [41].
152 A. Demertzi et al.

The use of advance directives could also be considered as a means to regulate

cost savings in the end-of-life; once the wishes of a terminal patient are known, care
can be taken as to constrain extraordinary means and spare the available resources
on other urgent cases. However, no such rationale corresponds to reality, and
advance directives, together with hospice care and the elimination of futile care,
have not contributed to the effective regulation of the economics of dying [42].
Treatment resources are not unlimited and despite care for a good death sometimes
physicians need to do with the means they have available. The allocation of resources
and the economics at the end-of-life have not yet been fully determined for DOC
patients. In ICU medicine some unwritten rules can facilitate decisions as to who is
to be treated, like the “first come” principle or “who will most likely benefit from
the intensive care” [43]. However, for chronic DOC cases, information on resource
allocation is often lacking. This may be due to the nature of chronic VS/UWS and
MCS patients. These are severely brain-damaged patients for whom the dilemma on
treating becomes crucial either because treatments are not guaranteed as successful
(i.e., the condition is too bad to be treated) or unkind (i.e., the quality of life of those
surviving is not acceptable) which may lead to an unwise way to allocate the avail-
able resources [43].
The legal provisions concerning the end-of-life issues in DOC differ from coun-
try to country. In the United States, where a patient-centered medical framework has
been adopted, the patient is allowed to participate in the regulation of her/his own
medical care. In the case of DOC, legal representatives in close collaboration with
the clinical staff and in line with the patients’ previously expressed wishes may
decide together about the long-term care of irreversibly comatose patients. There
are times, however, when conflicts of interests arise while making such decisions
either between family and physicians, such as in the Quinlan case [44], or among
family members, like the more recent Schiavo case [45]. As most often such cases
require the mediation of the court, they may have a wider publicity where the public
opinion can come into play and may lead to societal movements on pro-life versus
right-to-die action groups [46]. In Europe, there are more subtle differences in the
way treatment limitation is perceived, especially between Northern (more right-to-
die oriented) and Southern European countries (more pro-life oriented) [33]. In gen-
eral, decisions for treatment limitation (usually concerning ANH) need to be taken
after reference to the court. Exceptions are the Netherlands, Belgium, Switzerland,
and Scandinavian countries where no court mediation is needed for limiting treat-
ment in DOC patients [47].

Conclusions

The ethical issues, accrued from the study and management of DOC patients, are
variant and multifaceted. The medical, legal, and public controversies are partly
shaped by how different people think about these issues and in many cases are
country-dependent. It is, hence, necessary that a uniform ethical framework has to
14 The Ethics of Managing Disorders of Consciousness 153

be shaped to guide clinicians and caregivers in terms of clinical outcome, prognosis,

and medical management.

References

1. Laureys S, Boly M. What is it like to be vegetative or minimally conscious? Curr Opin Neurol.
2007;20:609–13.
2. Report of the Ad Hoc Committee of the Harvard Medical School to Examine the Definition of
Brain Death. A definition of irreversible coma. JAMA. 1968;205:337–40.
3. Zeman A. Consciousness. Brain. 2001;124:1263–89.
4. Demertzi A, Liew C, Ledoux D, et al. Dualism persists in the science of mind. Ann N Y Acad
Sci. 2009;1157:1–9.
5. Posner J, Saper C, Schiff N, Plum F, editors. Plum and Posner’s diagnosis of stupor and coma.
4th ed. New York: Oxford University Press; 2007.
6. Jennett B, Plum F. Persistent vegetative state after brain damage. A syndrome in search of a
name. Lancet. 1972;1:734–7.
7. Giacino JT, Ashwal S, Childs N, et al. The minimally conscious state: definition and diagnostic
criteria. Neurology. 2002;58:349–53.
8. Schnakers C, Vanhaudenhuyse A, Giacino JT, et al. Diagnostic accuracy of the vegetative and
minimally conscious state: clinical consensus versus standardized neurobehavioral assess-
ment. BMC Neurol. 2009;9:35.
9. Owen AM, Coleman MR, Boly M, Davis MH, Laureys S, Pickard JD. Detecting awareness in
the vegetative state. Science. 2006;313:1402.
10. Monti MM, Vanhaudenhuyse A, Coleman MR, et al. Willful modulation of brain activity in
disorders of consciousness. N Engl J Med. 2010;362:579–89.
11. Horne M (2009) Are people in a persistent vegetative state conscious? Monash Bioeth Rev.
28:12–1–12.
12. Kahane G, Savulescu J. Brain damage and the moral significance of consciousness. J Med
Philos. 2009;34:6–26.
13. International Association for the Study of Pain. Classification of chronic pain: descriptions of
chronic pain syndromes and definitions of pain terms, Task force on taxonomy. Seatle: IASP
Press; 1994.
14. Demertzi A, Schnakers C, Ledoux D, et al. Different beliefs about pain perception in the veg-
etative and minimally conscious states: a European survey of medical and paramedical profes-
sionals. Prog Brain Res. 2009;177:329–38.
15. Fins JJ. Affirming the right to care, preserving the right to die: disorders of consciousness and
neuroethics after Schiavo. Palliat Support Care. 2006;4:169–78.
16. Ahronheim JC, Gasner MR. The sloganism of starvation. Lancet. 1990;335:278–9.
17. The Medical Task Force on Anencephaly. The infant with anencephaly. N Engl J Med.
1990;322:669–74.
18. McQuillen MP. Can people who are unconscious or in the “vegetative state” perceive pain?
Issues Law Med. 1991;6:373–83.
19. Schnakers C, Chatelle C, Vanhaudenhuyse A, et al. The Nociception Coma Scale: a new tool
to assess nociception in disorders of consciousness. Pain. 2010;148:215–9.
20. Laureys S, Faymonville ME, Peigneux P, et al. Cortical processing of noxious somatosensory
stimuli in the persistent vegetative state. Neuroimage. 2002;17:732–41.
21. Boly M, Faymonville ME, Schnakers C, et al. Perception of pain in the minimally conscious
state with PET activation: an observational study. Lancet Neurol. 2008;7:1013–20.
22. Demertzi A, Vanhaudenhuyse A, Bruno MA, et al. Is there anybody in there? Detecting aware-
ness in disorders of consciousness. Expert Rev Neurother. 2008;8:1719–30.
154 A. Demertzi et al.

23. Poldrack RA. The role of fMRI in cognitive neuroscience: where do we stand? Curr Opin
Neurobiol. 2008;18:223–7.
24. Schnakers C, Zasler ND. Pain assessment and management in disorders of consciousness. Curr
Opin Neurol. 2007;20:620–6.
25. Bernat JL. Ethical issues in the perioperative management of neurologic patients. Neurol Clin.
2004;22:457–71.
26. Youngner SJ. Do-not-resuscitate orders: no longer secret, but still a problem. Hastings Cent
Rep. 1987;17:24–33.
27. Laureys S. Science and society: death, unconsciousness and the brain. Nat Rev Neurosci.
2005;6:899–909.
28. Boveroux P, Kirsch M, Boly M, et al. Évaluation du pronostic neurologique dans les encéphal-
opathies postanoxiques. Réanimation. 2008;17:613–7.
29. Bernat JL, Beresford HR. The controversy over artificial hydration and nutrition. Neurology.
2006;66:1618–9.
30. Rosner F. Why nutrition and hydration should not be withheld from patients. Chest.
1993;104:1892–6.
31. Steinbrook R, Lo B. Artificial feeding–solid ground, not a slippery slope. N Engl J Med.
1988;318:286–90.
32. Royal College of Physicians. The vegetative state: guidance on diagnosis and management.
Clin Med. 2003;3:249–54.
33. Demertzi A, Ledoux D, Bruno MA, et al. Attitudes towards end-of-life issues in disorders of
consciousness: a European survey. J Neurol. 2011;258(6):1058–65.
34. Bernat JL, Steven L. The concept and practice of brain death. Prog Brain Res.
2005;150:369–79.
35. Bernat JL. Ethical issues in neurology. 2nd ed. Boston: Butterworth Heinemann; 2002.
36. Brierley JB, Graham DI, Adams JH, Simpsom JA. Neocortical death after cardiac arrest. A
clinical, neurophysiological, and neuropathological report of two cases. Lancet.
1971;2:560–5.
37. Bernat JL. A defense of the whole-brain concept of death. Hastings Cent Rep.
1998;28:14–23.
38. Engelhardt K. Organ donation and permanent vegetative state. Lancet. 1998;351:211.
39. Truog RD, Robinson WM. Role of brain death and the dead-donor rule in the ethics of organ
transplantation. Crit Care Med. 2003;31:2391–6.
40. Bernat JL. Clinical ethics and the Law. In: Bernat JL, editor. Ethical issues in neurology.
Boston: Butterworth Heinemann; 2002. p. 79–107.
41. Bernat JL. The persistent vegetative state and related states. In: Bernat JL, editor. Ethical issues
in neurology. Boston: Butterworth Heinemann; 2002. p. 283–305.
42. Emanuel EJ, Emanuel LL. The economics of dying. The illusion of cost savings at the end of
life. N Engl J Med. 1994;330:540–4.
43. Jennett B. Editorial: resource allocation for the severely brain damaged. Arch Neurol.
1976;33:595–7.
44. Beresford HR. The Quinlan decision: problems and legislative alternatives. Ann Neurol.
1977;2:74–81.
45. Quill TE. Terri Schiavo –a tragedy compounded. N Engl J Med. 2005;352:1630–3.
46. Wijdicks EFM. Law and bioethics. In: Wijdicks EFM, editor. The comatose patient. New
York: Oxford University Press; 2008. p. 201–16.
47. Jennett B. Ethical issues. In: Jennett B, editor. The vegetative state: medical facts, ethical and
legal dilemmas. Cambridge: Cambridge University Press; 2002. p. 97–125.
Chapter 15
Future Perspectives of Clinical Coma Science

Steven Laureys and Caroline Schnakers

Redefining Disorders of Consciousness

As illustrated in the previous chapters, the clinical management of disorders of con-

sciousness remains very difficult, but technological advances in neuroimaging and
in EEG-based brain–computer interfaces are now offering new ways to improve the
diagnostic, prognostic, and therapeutic management of these challenging condi-
tions. We believe the recent international clinical research efforts in this challenging
field are heralding a new era of consciousness science and coma management [1].
Following severe brain damage and coma, some patients may awaken (i.e., open
the eyes) but remain unresponsive (i.e., only showing reflex movements). In Europe,
this clinical syndrome was initially termed apallic syndrome and coma vigil; but it
was later redefined as vegetative state (VS). Since its description more than 35 years
ago, an increasing number of functional neuroimaging and event related potential
studies [2] have shown that it sometimes may be difficult to make strong claims
about vegetative patients’ awareness [3]. This situation is further complicated when
patients have underlying deficits in verbal or nonverbal communication functions,
such as aphasia, agnosia, or apraxia [4].

S. Laureys, M.D., Ph.D. (*)

Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du 6 Août, Sart Tilman B30, 4000 Liège, Belgium
Department of Neurology, University Hospital of Liège,
Liège, Belgium
e-mail: [email protected]
C. Schnakers, Ph.D.
Coma Science Group, Cyclotron Research Center, University of Liège,
Allée du 6 Août, Sart Tilman B30, 4000 Liège, Belgium
e-mail: [email protected]

C. Schnakers, S. Laureys (eds.), Coma and Disorders of Consciousness, 155

DOI 10.1007/978-1-4471-2440-5_15, © Springer-Verlag London 2012
156 S. Laureys and C. Schnakers

It appears that part of the healthcare, media, and lay public continue to feel some
unease regarding the unintended denigrating vegetable-like connotation seemingly
intrinsic to the term VS. The European Task Force on Disorders of Consciousness,
therefore, recently proposed an alternative name: unresponsive wakefulness syn-
drome (UWS) [5]. Hence, physicians are recently offered to refer to these patients
as UWS, a more neutral and descriptive term, pertaining to patients showing a num-
ber of clinical signs (hence syndrome) of unresponsiveness (i.e., without response
to commands or oriented voluntary movements) in the presence of wakefulness
(i.e., eye opening). In contrast to coma, which is an acute condition lasting no more
than some days or weeks; VS/UWS can be a chronic condition lasting years or
remain permanent.
Those patients who do recover, classically evolve to a minimally conscious state,
defined by the presence of nonreflex voluntary movements such as orientation to
pain, eye tracking, or reproducible albeit inconsistent command following. By
definition, MCS patients cannot communicate their thoughts or wishes [6]. Finally,
in pseudocoma or locked-in syndrome (LIS), patients may awaken from coma fully
conscious but paralyzed, only able to communicate by eye movements. The medical
and ethical management of these vulnerable patients is very complex and a recent
paper reviewed the even more challenging (but fortunately very rare) situation of
LIS in children [7]. Locked-in patients illustrate how difficult it can be to make
inferences about subjective conscious awareness in severely motor impaired.

A Brief History of Coma

Historically, the seat of consciousness was widely believed to be in the heart, and
the absence of heartbeat was regarded as the clinical sign of death. Neurological
scientific evidence has superseded such thinking and shown that consciousness, an
emergent property of neural activity, resides in the brain [1]. Since the invention of
the positive-pressure mechanical respirator, it has become possible to truly dissoci-
ate cardiac, respiratory, and brain functions in individuals who are in a coma.
Patients who would previously have died from apnea are now able to survive in
profound comatose states that had never been encountered before. This technologi-
cal progress forced modern medicine to redefine the diagnosis of death and move
from its ancient cardiorespiratory-centered definition to a neurocentric one, where
death is defined as the irreversible loss of all brainstem reflexes (including the
breathing reflex). Since the introduction of this clinical definition, not a single
patient who was declared brain dead has ever regained consciousness.
The origin of coma research, as a science, probably dates to 1966 when Fred
Plum and Jerome Posner published the first edition of their classic text The Diagnosis
of Stupor and Coma [8]. For the first time, researchers correlated clinical findings
that were derived from the examination of patients in comatose states with patho-
logical findings, and proposed a pathophysiology of consciousness. In 1974, Bryan
Jennett and colleagues published the Glasgow Coma Scale [9] and, in the next year,
15 Future Perspectives of Clinical Coma Science 157

the Glasgow Outcome Scale [10]. These standardized scoring systems enabled the
performance of multicenter clinical trials and epidemiological studies that resulted
in the development of rational algorithms for the treatment (or withdrawal thereof)
of comatose patients.
Pioneers such as Jennett and Plum revolutionized the field of acute brain injury.
However, the excitement of the 1970s was followed by a return to therapeutic nihil-
ism (i.e., the assumption that patients with chronic disorders of consciousness are
uniformly hopeless cases) and a marked decrease of scientific interest in disorders
of consciousness. Coma research nearly got comatose. Only recently, partly incited
by advances in the field of functional neuroimaging, did the study of coma and
related states undergo a revival [11].

New Ways of Showing Signs of Consciousness

The clinical studies conducted, since the 1970s, have illustrated how challenging it
is to disentangle reflex behavior from voluntary behavior—a problem that can
potentially lead to diagnostic error. We are confined to inferring the presence of
awareness from a patient’s motor responsiveness. The differential diagnosis of dis-
orders of consciousness, hence, may be challenging. The rate of misdiagnosis is
approximately 40%, and new methods are required to complement bedside testing,
particularly if the patient’s capacity to show behavioral signs of awareness is dimin-
ished [12].
Functional neuroimaging now offers the possibility to directly measure the
brain’s activity, not only at rest or during passive stimulation, but also in response to
commands [11]. Assessing resting state brain metabolism using fluorodeoxyglucose
PET, a recent study could disentangle vegetative from locked-in patients by an auto-
matic procedure based on machine learning [13]. A sparse probabilistic Bayesian
learning framework called “relevance vector machine” correctly classified all
included scans as conscious (i.e., locked-in) or unconscious (i.e., vegetative/unre-
sponsive). Such automated user-independent, classification tools of cerebral meta-
bolic images obtained in coma survivors could become a useful tool for the
automated PET-based diagnosis of altered states of consciousness.
In the same line, resting state functional MRI study can now quantify the integ-
rity of the brain’s connectivity pattern in the so-called default mode network in
pathological alterations of consciousness [14]. The default network is defined as a
set of areas, encompassing posterior-cingulate/precuneus, anterior-cingulate/mesio-
frontal cortex, and temporoparietal junctions, which show more activity at rest than
during attention-demanding tasks. Connectivity was investigated using probabilistic
independent component analysis, and an automated template-matching component
selection approach. Connectivity in all default network areas was found to be nega-
tively correlated with the degree of clinical consciousness impairment, ranging from
healthy controls and locked-in syndrome to minimally conscious, vegetative, then
coma patients. Furthermore, precuneus connectivity was found to be significantly
158 S. Laureys and C. Schnakers

stronger in minimally conscious patients as compared with unconscious patients;

and locked-in syndrome patient’s default network connectivity was not significantly
different from controls. Prospective studies in a larger patient population are needed
in order to evaluate the prognostic value of the presented methodology.

Communicating via a Scanner

fMRI activation studies classically measure the brain’s response to external sensory
stimuli. For example, Qin et al. [15] have studied auditory processing in 11 DOC
patients using an auto-referential attention-grabbing stimulus, the patient’s own
name. They reported a linear correlation between activation of the limbic system
(anterior cingulate cortex) and the level of consciousness as quantified by coma
recovery scores. However, such passive stimulation induced by brain activation
does permit to make strong claims about the patient’s consciousness.
Now, Monti and Vanhaudenhuyse et al. [16] have been using a so-called active
paradigm where patients are instructed to follow simple commands. Their study
enrolled 54 DOC cases of different etiology in imaging centers in Cambridge and
Liège, asking patients to perform two mental imagery tasks: imagine playing tennis
(activating motor areas) and imagine walking around in your house (activating para-
hippocampal brain regions). Of the 31 patients previously diagnosed as MCS, only
one was able to show reliable fMRI activation in the expected brain areas, illustrat-
ing that the employed technique only has a sensitivity of 3% in minimally conscious
DOC patients (i.e., has a 97% false negative rate). It is more difficult to calculate the
sensitivity and specificity of the technique in VS/UWS patients given that we do not
have a gold standard of consciousness. Indeed, in the published case series, 4 out of
23 clinically unconscious patients showed fMRI signs of command following and
hence of consciousness (i.e., 17%); but given the known low sensitivity, it could be
that other patients might have been (minimally) conscious yet were unable to do the
task or show statistically significant brain activation. However, in those individual
patients where reliable and long-lasting (>30 s) command-related fMRI activation
was observed, there seems no logical alternative explanation other than that they
repeatedly understood and replied to the commands and hence truly showed con-
scious awareness [17].
In addition to show proof of consciousness, fMRI is being used to communicate
with some (very exceptional) DOC patients. Indeed, one clinically noncommunica-
tive patient studied in the Liège University Hospital was shown to correctly answer
five out of six simple questions regarding his family members’ names [16]. Here, it
was set that to communicate “yes,” one should imagine playing tennis (and hence
motor area activation was identified using an automated user-independent relative-
similarity classifier) and to communicate “no,” the spatial navigation task should be
performed (activating parahippocampal areas). fMRI muscle-independent commu-
nication currently even goes beyond such yes–no answers to closed questions and is
offering the possibility to answer multiple choice questions [18]. Evidently, these
15 Future Perspectives of Clinical Coma Science 159

data should be seen as proofs of concept rather than as a practical means to truly
assure long-term communication. EEG based communication devices, called brain–
computer interfaces, are, therefore, being developed as a more practical, transport-
able, and cheaper alternative.
Similarly, to the described fMRI active paradigms, EEG-based technology will
aim to answer two questions—Is the patient conscious? and, if the answer is
affirmative, What is that consciousness like?—needing functional communication.
The practical utility of demonstrating voluntary brain activity using nonclinical
means has already been translated to clinical use—illustrated in the case of a 21-year
old comatose woman who failed to show any motor sign of conscious awareness up
to 49 days after an extensive brainstem and thalamic stroke [19]. Only EEG evoked-
potential based command following (i.e., counting a target name in a list of names),
shown at the intensive care unit, permitted to make the diagnosis of complete
locked-in syndrome. Future multicentric trials are awaited to validate the possible
prognostic value of these tests.

Predicting and Improving the Outcome

In many patients with disorders of consciousness, it remains difficult to reliably pre-

dict the chances of functional recovery. However, MRI techniques such as spectros-
copy and diffusion tensor imaging are improving our possibilities to quantify the
neuronal damage. Magnetic resonance spectroscopy now permits to reveal axonal
sprouting in exceptional cases of late recovery [20]. Such demonstration of axonal
sprouting after sometimes very long time intervals of several years disproves the old
dogma that neural plasticity is limited to the acute or subacute phase of head trauma.
Finally, the widespread therapeutic nihilism with regards to chronic disorders of
consciousness is being challenged by new evidence derived from the use of deep-
brain stimulation. Posttraumatic minimally conscious state patients are undergoing
bilateral thalamic stimulation, improving cognitive function and resulted in a return
to functional object use and intelligible verbalization [21]. The usefulness of deep
brain stimulation in this setting again awaits confirmation from studies that include
larger cohorts of patients, but the evidence to date indicates that the technique can
offer a real therapeutic option in well-chosen patients (selected on the basis of func-
tional neuroimaging results).
In the recovery from coma, the acquisition of interactive verbal or nonverbal
communication represents an important milestone—the ability to communicate
their wishes concerning treatment options allows these vulnerable patients to exert
their right to autonomy. Rapid advances in the development of brain–computer
interfaces mean that clinically noncommunicative coma survivors can finally have a
voice of their own [22]. We predict that real-time functional MRI and EEG-mediated
technology will offer these patients nonmuscle-dependent channels of communica-
tion, enabling them to express their thoughts, control their environment, and increase
their quality-of-life.
160 S. Laureys and C. Schnakers

Changing the Spectrum of Coma

Clearly, it is a thrilling time for the field of acute-brain injury. The gray zones
between the different disorders of consciousness in the clinical spectrum following
coma are beginning to be better defined by the addition of powerful imaging meth-
odology to bedside clinical assessment. We foresee that in a yet-to-be-discovered
fraction of patients who are currently considered to be vegetative, electromyogra-
phy, event-related potential, and functional MRI techniques will reveal signs of con-
sciousness that are unattainable by bedside clinical assessment. These novel
technological means will change the existing behaviorally defined boundaries
between the various consciousness disorders (Fig. 15.1). In a next step, these high-
tech devices will permit some of these clinically noncommunicative patients to cor-
respond their thoughts and wishes via nonmotor pathways. However, it should be

Emergence

EEG (brain computer interface)

Interactive communication or real-time fMRI communication
not depending on motor pathways

Minimally conscious state

Voluntary movements or Subclinical command following

command following shown by EMG, ERP or fMRI

Vegetative state

Eye opening
only reflex behavior

Coma

Brainstem reflexes

Brain death

Fig. 15.1 Patients in a coma may become brain dead (when brainstem reflexes are irreversibly
lost) or progress to a vegetative state (clinically defined as wakefulness solely accompanied by
reflex behavior); next patients may transition to a minimally conscious state (i.e., the patient shows
signs of command following or other nonreflexive movements); the restoration of interactive com-
munication indicates the emergence from this stage in the clinical spectrum of recovery from
coma. Novel technological means are changing these behaviorally defined boundaries. In the
future, electromyography (EMG), event-related potential (ERP) and functional MRI (fMRI) tech-
niques might reveal signs of consciousness that are unattainable by bedside clinical assessment in
a yet-to-be-discovered fraction of patients who are currently considered to be vegetative. In a next
step, these high-tech devices might permit some of these patients to communicate their thoughts
and wishes via nonmotor pathways
15 Future Perspectives of Clinical Coma Science 161

stressed that these exciting developments are not yet reality. The first hurdle to be
overcome before the methodologies discussed above can enter clinical practice
relates to ethical concerns. We side with a proposed ethical framework that empha-
sizes the need to achieve a balance between the protection of patients with disorders
of consciousness and the ability to perform research that can lead to medical prog-
ress [23]. Most of the advances in coma science discussed above are based on sin-
gle-case studies. Only large-scale multicenter clinical trials will enable these
research tools to find their way into evidence-based care that really influences the
lives of individuals with a disorder of consciousness.

References

1. Laureys S, Boly M. The changing spectrum of coma. Nat Clin Pract Neurol. 2008;
4(10):544–6.
2. Bruno MA, Gosseries O, Ledoux D, et al. Assessment of consciousness with electrophysiolog-
ical and neurological imaging techniques. Curr Opin Crit Care. 2011;17(2):146–51.
3. Monti MM, Laureys S, Owen AM. The vegetative state. Brit Med J. 2010;341:3765.
4. Majerus S, Bruno MA, Schnakers C, et al. The problem of aphasia in the assessment of con-
sciousness in brain-damaged patients. Prog Brain Res. 2009;177:49–61.
5. European Task Force on Disorders of Consciousness. Unresponsive wakefulness syndrome: a
new name for the vegetative state or apallic syndrome. BMC Med. 2010;1(8):68.
6. Giacino JT, Schnakers C, Rodriguez-Moreno D, et al. Behavioral assessment in patients with
disorders of consciousness: gold standard or fool’s gold? Prog Brain Res. 2009;177:33–48.
7. Bruno MA, Schnakers C, Damas F, et al. Locked-in syndrome in children: report of five cases
and review of the literature. Pediatr Neurol. 2010;41(4):237–46.
8. Plum F, Posner JB. The diagnosis of stupor and coma. 1st ed. Philadelphia: Davis, F.A; 1966.
9. Teasdale G, Jennett B. Assessment of coma and impaired consciousness. A practical scale.
Lancet. 1974;2(7872):81–4.
10. Jennett B, Bond M. Assessment of outcome after severe brain damage. Lancet.
1975;1(7905):480–4.
11. Owen A, Coleman M, Boly M, et al. Detecting awareness in the vegetative state. Science.
2006;313(5792):1402.
12. Schnakers C, Vanhaudenhuyse A, Giacino J, et al. Diagnostic accuracy of the vegetative and
minimally conscious state: clinical consensus versus standardized neurobehavioral assess-
ment. BMC Neurol. 2009;9:35.
13. Phillips CL, Bruno MA, Maquet P, et al. “Relevance vector machine” consciousness classifier
applied to cerebral metabolism of vegetative and locked-in patients. Neuroimage.
2011;56(2):797–808.
14. Vanhaudenhuyse A, Demertzi A, Schabus M, et al. Two distinct neuronal networks mediate the
awareness of environment and of self. J Cogn Neurosci. 2011;23(3):570–8.
15. Qin P, Di H, Liu Y, et al. Anterior cingulate activity and the self in disorders of consciousness.
Hum Brain Mapp. 2010;31(12):1993–2002.
16. Monti MM, Vanhaudenhuyse A, Coleman MR, et al. Willful modulation of brain activity in
disorders of consciousness. N Engl J Med. 2010;62(7):579–89.
17. Stins JF, Laureys S. Thought translation, tennis and Turing tests in the vegetative state. Phenom
Cogn Sci. 2009;8:361–70.
18. Sorger B, Dahmen B, Reithler J, et al. Another kind of ‘BOLD Response’: answering multiple-
choice questions via online decoded single-trial brain signals. Prog Brain Res.
2009;177:275–92.
162 S. Laureys and C. Schnakers

19. Schnakers C, Perrin F, Schabus M, et al. Detecting consciousness in a total locked-in syn-
drome: an active event-related paradigm. Neurocase. 2009;15(4):271–7.
20. Voss HU, Uluc AM, Dyke JP, et al. Possible axonal regrowth in late recovery from the mini-
mally conscious state. J Clin Invest. 2006;116(7):2005–11.
21. Schiff ND, Fins JJ. Deep brain stimulation and cognition: moving from animal to patient. Curr
Opin Neurol. 2007;20(6):638–42.
22. Kubler A, Furdea A, Halder S, et al. A brain-computer interface controlled auditory event-
related potential (p300) spelling system for locked-in patients. Ann N Y Acad Sci.
2009;1157:90–100.
23. Fins JJ, Schiff ND, Foley KM. Late recovery from the minimally conscious state: ethical and
policy implications. Neurology. 2007;68(4):304–7.
Index

A BAEP. See Brainstem auditory evoked

ABAB design, 101 potentials (BAEP)
Alpha-coma (AC), 89 Basal ganglia
Alpha rhythms DBS, 143
BCI systems, 70 MRS, 48
EEG interpretation, 57 swallowing, 113
ALS. See Amyotrophic lateral sclerosis (ALS) BCI systems. See Brain-computer interface
Altered state of consciousness (BCI) systems
active fMRI paradigms Bispectral index (BIS), 63
description, 29–30 Bistable dynamics, 81–82
voluntary brain modulation, 30–31 Brain
patient prognosis death
functional recovery, 12 clinical management, 150–151
MCS, 12–21 coma, 160
medical complications, 12 description, 2
mortality factors, 12 diagnostic criteria, 3
recovery, 12 decoding (see Multivariate pattern
VS/UWS, 12–21 recognition analysis)functions
Alzheimer’s disease (AD), 39–40 activation and external stimuli, 28–29
Amantadine activity at rest, 26–28
CNC scale, 128 default mode network, 27–28
CRS-R, 128 differentiated approach, 26
dopaminergic agent, 122 multimodal assessment techniques, 26
GCS, 122 vegetative state, 26, 27
uses, 122, 128 injury
Amyotrophic lateral sclerosis (ALS), 71, 73 DBS, 143
Anterior cingulate cortex MRS, 48
conscious perception, 29 sensory stimulation program, 97–98
swallowing, 112–113 Brain-computer interface (BCI) systems
Apomorphine, 129 coma, 76
Artificial nutrition and hydration (ANH), 150 description, 67–68
diagnostic and communication tools
B EEG-based, 71, 73, 74
Baclofen fMRI-based, 74–75
assumptions, 132 invasive BCIs systems, 75
description, 131–132 multivariate pattern recognition analysis, 41
effects, 132 techniques

C. Schnakers, S. Laureys (eds.), Coma and Disorders of Consciousness, 163

DOI 10.1007/978-1-4471-2440-5, © Springer-Verlag London 2012
164 Index

Brain-computer interface (BCI) systems (cont.) D

EEG, 68–70 Deep brain stimulation (DBS)
neuroimaging, 70–73 brain pacemaker, 140
Brain plasticity, 98 clinical applications, 140
Brainstem auditory evoked potentials criterias, 144
(BAEP), 59 description, 139
Bromocriptine, 129 and DOC treatment
animal studies, 140
intralaminar and rostral-ventral
C thalamic nuclei, 140–141
Cardiopulmonary resuscitation (CPR), 150 intralaminar nuclei of the
Central pattern generator (CPG) thalamus, 143
oral feeding, 117 mechanisms, 142–143
swallowing, 108, 109 median-parafascicular complex, 141
Central thalamic nuclei (CTN), 133–134 midbrain reticular formation, 141
Cerebellum parafascicular intralaminar and
LIS, 92 centromedian thalamic nuclei, 141
swallowing, 114 VS patients, 141–142
Cerebral connectivity measurements. See history, 140
Transcranial magnetic stimulation neurostimulator, 140
(TMS) traumatic brain injury, 143
Cerebral cortex, swallowing, 111 Default mode network, 27–28
Chemical shift imaging (CSI), 48, 50 Diffuse axonal hemorrhagic
CNC scale. See Coma/near coma (CNC) scale lesions, 46, 47
Coma Diffusion tensor imaging (DTI)
AC, 89 description, 49–50
BCI system, 76 FA, 50–52
brain death, 160 restricted isotropic diffusion, 50
description, 4 Disability rating scale, 19, 20
diagnostic criteria, 3 Disorders of consciousness (DOC)
EEG VS, 3, 4
interpretation, 58 behavioral scales
patterns, 89–90 arousal preservation, 5, 6
GOS, 156, 157 detection, 5
history, 156–157 FOUR, 6–7
MRI, 46 GCS, 6
NCS, 8 JFK CRS-R, 7
NREM sleep elements, 90 WHIM, 7
outcome prediction and improvement, 159 brain death, 2, 3
REM sleep elements, 90 coma, 3, 4
SC, 89 description, 1–2
sensory stimulation program, 102 and electrophysiology
spectrum change, 160–161 EEG interpretation, 57–58
Coma/near coma (CNC) scale, 128 electroencephalography, 55–56
Coma recovery scale-revised (CRS-R) evoked potentials, 58–61
altered state of consciousness, 128 qEEG, 62–63
amantadine, 128 LIS, 3, 5
CPG. See Central pattern MCS, 3–5
generator (CPG) NCS, 8
CPR. See Cardiopulmonary pain assessment, 8
resuscitation (CPR) pharmacological treatments (see
CSI. See Chemical shift imaging (CSI) Pharmacological treatments)
CTN. See Central thalamic redefined, 155–156
nuclei (CTN) sleep (see Sleep)
Index 165

Do-not-resuscitate (DNR) orders, 150 legal issues

DTI. See Diffusion tensor imaging (DTI) advance directives usage, 151–152
end-of-life issues, 151, 152
managing patients, 151
E Evoked potentials (ERPs)
Electroencephalogram BAEP, 59
approaches, 56 BCI systems, 69
interpretation components, 59
alpha rhythm, 57 description, 58–59
classification, 57, 58 endogenous components, 59
coma, 58 exogenous components, 59
global cerebral activity, 57 MMN, 60
loss of consciousness, 57–58 N100, 60
reactivity, 57 N400, 61
qEEG, 63 P3, 60–61
sleep, 88–89 SEP, 59
uses, 56
Electroencephalography (EEG)
BCI systems F
alpha rhythms, 70 Fluid attenuated inversion recovery
ALS, 71, 73 (FLAIR), 46–47
description, 68 Fluorodeoxyglucose PET (FDG-PET), 40
ERPs, 69 fNIRS. See Functional near-infrared
LSP, 73 spectroscopy (fNIRS)
P3, 68, 69, 73, 74 Fractional anisotropy (FA), DTI, 50–52
SCPs, 70 Full outline of unresponsiveness scale
SMRs, 70, 73 (FOUR), 6–7
SSVEPs, 69–70 Functional magnetic resonance imaging
categories, 55–56 (fMRI)
description, 55 active paradigm, 158, 159
EEG, 56 active paradigms, 29–31
TMS BCI systems, 70–71, 74–75
anesthetic agents, 83 consciousness, 158–159
brain damaged patients, 82–83 default mode network, 157
deep sedation, 81 diagnosis, altered states
description, 81 of consciousness, 29–31
entropy index, 83 RVM, 157
slow wave sleep, 81–82 signs of consciousness, 157–158
validation studies, 81–82 sleep, 89
ERPs. See Evoked potentials (ERPs) vegetative state vs. MCS, 26
Ethical management Functional near-infrared spectroscopy
clinical management (fNIRS), 71, 75
VS, 148–150 Functional recovery
ANH, 150 MCS patients, 20–21
brain death, 150–151 VS/UWS patients, 18–20
consciousness, 148
CPR, 150
DNR orders, 150 G
LIS, 149 Glasgow coma scale (GCS), 6
MCS, 148–150 amantadine, 122
NCS, 149 oral feeding, 115
pain, 148–149 sensory stimulation program, 99–100
UWS, 148–150 zolpidem, 129
description, 147–148 Glasgow outcome scale (GOS), 19
166 Index

I Mismatch negativity (MMN), 60

Impaired consciousness. See Brain, functions MRI. See Magnetic resonance imaging (MRI)
Insula MRS. See Magnetic resonance spectroscopy
CSI, 48, 50 (MRS)
swallowing, 113 Multivariate pattern recognition analysis
Invasive BCIs systems, 75 AD, 39–40
advantages, 41
applications
J data mining tools, 39
JFK coma recovery scale-revised (CRS-R), 7 diagnostic tool, 39–40
neuroimaging data decoding, 39
voxels relevance, 40–41
L BCI, 41
Lamotrigin, 133 brain reading, 41–42
Language support program (LSP), 73 FDG-PET, 40
Levodopa, 128–129 methodology
Locked-in syndrome (LIS) classification errors types, 37–38
clinical management, 149 graphical representation, 37
description, 5 handwritten digits, 36
diagnostic criteria, 3 learning set, 37
redefining DOC, 155–156 supervised learning, 36
SVM, 38, 39
testing set, 37
M MVPA, 35–36
Magnetic resonance imaging (MRI) RVM, 40
axonal sprouting, 159 SPM, 35
coma, 46 Multi-voxel pattern analyses (MVPA), 35–36
description, 45
diffuse axonal hemorrhagic lesions, 46, 47
FLAIR, 46–47 N
outcome prediction and improvement, 159 N-acetyl-aspartate (NAA), 48
reticular activating system, 46 Neuroimaging
scan, 46 BCI systems
techniques, 46 fMRI, 70–71
Magnetic resonance spectroscopy (MRS) fNIRS, 71
CSI, 48, 50 multivariate decoding, 39
description, 47 Neurostimulator, 140
NAA, 48 Nociception coma scale (NCS)
NAA/Cr ratio, 48, 49 clinical management, 149
proton conduction, 47 pain assessment, 8
SVS, 48, 49 Noncommunicative brain
traumatic brain injury, 48 damaged patients, 79–80
Methylphenidate, 129–130 Non-rapid eye movement (NREM)
Minimally conscious state (MCS) sleep, 87, 88, 90
altered state of consciousness
etiology, 17
functional recovery, 20–21 O
mortality rate, 16, 18 Oral feeding
recovery of consciousness, 20–21 description, 105–106
recovery rates, 16 DOC patients
clinical management, 148–150 GCS, 115
descripton, 4–5 MCS, 115–117
diagnostic criteria, 3 RLA score, 115–117
oral feeding, 115–117 UWS, 115–116
Index 167

swallowing Postcentral gyrus, 113

complex sensorimotor function, 117 Precentral gyrus, 112
and consciousness, 110–114 Prefrontal cortex, 112
dysphagia, 117–118
neurological control, 108–110
neurophysiology, 106–108 Q
Quantitative electroencephalography (qEEG)
anesthesia monitoring, 62
P BIS, 63
P3 correlation methods, 63
BCI systems, 68, 69, 73, 74 description, 62
evoked potentials, 60–61 EEG, 63
Pain objective measurements, 63
assessment power spectral density, 62
coma, 4 source reconstruction methods, 63
NCS, 8
management, 148–149
Pharmacological treatments R
amantadine Rancho Los Amigos (RLA) scale
CNC scale, 128 oral feeding, 115–117
CRS-R, 128 zolpidem, 129
dopaminergic agent, 122 Rapid eye movement (REM) sleep, 87, 88, 90
GCS, 122 Recovery of consciousness
uses, 122, 128 MCS patients, 20–21
apomorphine, 129 VS/UWS patients, 18–20
baclofen Relevant vector machine (RVM)
assumptions, 132 fMRI, 157
description, 131–132 multivariate pattern recognition
effects, 132 analysis, 40
benefits, 135
bromocriptine, 129
description, 121–122 S
drugs, 122–127 SC. See Spindle-coma (SC)
hypotheses, 135–136 SCPs. See Slow cortical potentials (SCPs)
lamotrigin, 133 Sensorimotor rhythms (SMRs), 70, 73
levodopa, 128–129 Sensory perceptions, 28
measurement tools and behavioral scales, Sensory stimulation program
135 ABAB design, 101
methylphenidate, 129–130 chronic patients, 101
possible positive effects coma, 102
CTN, 133–134 description, 97
drug effects, 133 GCS, 99–100
mesocircuit model, 133–134 limitations and perspectives, 100–101
sensory perception, 122 multicentric studies, 101
serotonin-reuptake inhibitors, 132–133 neurorehabilitation field, 99
zolpidem positive effects, 99
awakening agent, 130 sensitive behavioral scales, 101
cerebral metabolism, 131 theoretical principles
effects, 130, 131 brain plasticity, 98
GABA neurotransmitters, 131 enriched environment effect, 98–99
GCS, 130 traumatic brain injury, 97–98
imidazopyridine, 130 valid behavioral scales, 101
RLA scale, 130 SEP. See Somesthetic-evoked potentials (SEP)
Polysomnography (PSG), 86, 88 Serotonin-reuptake inhibitors, 132–133
168 Index

Single voxel spectroscopy (SVS), 48, 49 animal studies, 110–111

Sleep complex sensorimotor function, 117
VS, 91 dysphagia, 117–118
abnormalities, 86 human studies
apallic syndrome, 91 anterior cingulate cortex, 112–113
arousal, 85 basal ganglia, 113
awareness, 85 cerebellum, 114
coma cerebral cortex, 111
AC, 89 insula, 113
EEG patterns, 89–90 mechanisms, 111
NREM sleep elements, 90 parietal regions, 113
REM sleep elements, 90 postcentral gyrus, 113
SC, 89 precentral gyrus, 112
LIS, 92 prefrontal cortex, 112
methodological issues temporal regions, 113
EEG, 88–89 neurological control
fMRI, 89 CPG, 108, 109
PSG, 88 cranial nerves, 110
PSG, 86 interneurons network, 109–110
recovered state, 91–92 motor program, 109–110
slow-wave neurophysiology
description, 81 esophageal phase, 107, 108
loss of integration and differentiation, 82 oral phase, 106–107
studies in DOC pharyngeal phase, 107–108
VS, 87
characterization, 87
coma, 87 T
day-night brain state differences, 87 Transcranial magnetic stimulation (TMS)
description, 86, 93 conscious perception
existence, 87 cerebral activity properties, 80
NREM, 87, 88 integrated information theory, 80–81
REM, 87, 88 and EEG technique
residual homeostasis assessment, 87 anesthetic agents, 83
SCN, 86 brain damaged patients, 82–83
sleep-wake cycle, 86 deep sedation, 81
UWS, 87 description, 81
zeitgeber, 86 entropy index, 83
UWS, 91 slow wave sleep, 81–82
Sleep-wake cycle, 86 validation studies, 81–82
Slow cortical potentials (SCPs), 70
Slow-wave sleep
description, 81 U
loss of integration and differentiation, 82 Unresponsive wakefulness
SMRs. See Sensorimotor rhythms (SMRs) syndrome (UWS)
Somesthetic-evoked potentials (SEP), 59 clinical evolution, 13
Spindle-coma (SC), 89 clinical management, 148–150
Statistical parametric mapping (SPM), 35 mortality and recovery rates
Steady-state visually evoked potentials MCS, 16
(SSVEPs), 69–70 non-traumatic, 15
Support vector machines (SVM), 38, 39 traumatic, 14
Suprachiasmatic nucleus (SCN), 86 oral feeding, 115–116
SVS. See Single voxel spectroscopy (SVS) redefining DOC, 155–156
Swallowing process sleep, 91
Index 169

V W
Vegetative state (VS) Wessex head injury matrix (WHIM), 7
clinical evolution, 13
clinical management, 148–150
description, 4 Z
diagnostic criteria, 3 Zolpidem
mortality and recovery rates awakening agent, 130
MCS, 16 cerebral metabolism, 131
non-traumatic, 15 effects, 130, 131
traumatic, 14 GABA neurotransmitters, 131
redefining DOC, 155–156 GCS, 130
sleep, 91 imidazopyridine, 130
Voxels relevance, 40–41 RLA scale, 130

Download more eBooks here: http://avaxhm.com/blogs/ChrisRedfield