Monitoring the Nervous

System for Anesthesiologists

and Other Health Care
Professionals

Second Edition

Antoun Koht
Tod B. Sloan
J. Richard Toleikis
Editors

123
Monitoring the Nervous System
for Anesthesiologists and Other Health
Care Professionals
Antoun Koht • Tod B. Sloan
J. Richard Toleikis
Editors

Monitoring
the Nervous System
for Anesthesiologists
and Other Health Care
Professionals
Second Edition
Editors
Antoun Koht, MD Tod B. Sloan, MD, MBA, PhD
Departments of Anesthesiology Department of Anesthesiology
Neurological Surgery and Neurology University of Colorado School of
Northwestern University Feinberg Medicine
School of Medicine Aurora, CO, USA
Chicago, IL, USA

J. Richard Toleikis, PhD

Department of Anesthesiology
Rush University Medical Center
Chicago, IL, USA

Videos can also be accessed at

http://link.springer.com/book/10.1007/978-3-319-46542-5

ISBN 978-3-319-46540-1    ISBN 978-3-319-46542-5 (eBook)

DOI 10.1007/978-3-319-46542-5

Library of Congress Control Number: 2017943687

© Springer International Publishing AG 2012, 2017

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.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in
this book are believed to be true and accurate at the date of publication. Neither the publisher nor
the authors or the editors give a warranty, express or implied, with respect to the material
contained herein or for any errors or omissions that may have been made.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG Switzerland
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
A few pioneer neurophysiologists, surgeons, anesthesiologists,
and researchers from many disciplines and countries gave birth
to the field of intraoperative neuromonitoring over 40 years ago
when together they realized that there was a need and a means
for providing better patient care. Their interest and efforts
resulted in numerous subsequent international meetings,
spawned the establishment and growth of various professional
societies, and ignited a growing interest in IOM. All of this
would not have occurred without the contributions and support
of countless individuals who dedicated a significant part of their
professional lives and resources to the field of IOM. It is to these
various pioneers and individuals that we dedicate this book.
Foreword to First Edition: Orthopedic
Spine Viewpoint

We began our Case Western Reserve University (CWRU) efforts to develop a

system for monitoring spinal cord function during scoliosis corrective sur-
gery in the late 1960s. It was prompted by the risks to the spinal cord as a
result of using Harrington Rods for curvature correction. From the very
beginning, we approached it as a team effort. University Hospitals and CWRU
had the expertise to tackle this difficult challenge, but they had to be pulled
together. A young neurosurgeon by the name of Jerald Brodkey had some
experience with a technique of summating distal peripheral nerve stimula-
tions as they were expressed over the cortex, a process reported by Dawson in
the 1950s. At the same time there was a very bright, young master’s biomedi-
cal engineer, Richard Brown working in the CWRU biomedical engineering
laboratories of Drs. Victor Frankel, MD, PhD and Al Burstein,
PhD. Fortuitously, Richard’s undergraduate degree was in electrical engi-
neering and he had some free time available to work on the spinal cord moni-
toring project (which became his PhD thesis!).
The approach taken in the laboratory was to study the effect of graduated
weights applied directly to the thoracic spinal cord of dogs for varying peri-
ods of time on the ability of the cord to transmit trains of stimuli from the
distal extremities to the cortex. In the course of these studies it also became
apparent that pressure, time, and blood pressure were all critical variables.
Then available commercial neuromonitoring systems were used, but from the
beginning Rich Brown recognized that they would not work in the highly
electrically charged environment of an operating room (OR). Thus began his
creation of a stand alone, portable spinal cord monitoring system capable of
accurately recording the very small cortical signals generated in the hostile
atmosphere of the OR. Thus “Big Blue,” as Rich would call it, came to be
originally equipped with four channels, but soon expanded to eight with all
data stored on tape for later analysis. “Real-time” record assessment was
done by holding up a base line printout up to the light with the current record
printout superimposed to visually determine latency and amplitude changes.
Appropriate filtering, stimulus rates, stimulus configuration, and voltages
along with Rich’s primary passion, patient safety, were all factors to be sorted
out. “Warning signs” of changes in latency and amplitude were part of the
equation with the 10 and 50% guidelines becoming evident even then. From
the beginning, Rich’s goal was to produce a system that would prove both
reliable and provide valid data – causes he championed his entire career –
later holding all systems to the same fire he held his own.

vii
viii Foreword to First Edition: Orthopedic Spine Viewpoint

Once the system had proven to be effective in the laboratory by sorting out
the amounts of weight over what periods of time that correlated clinically
with the presence or absence of clinical neurological deficits, it was time to
take it to the OR. It was strongly suspected that the more complex anesthesia
used in humans would have significant effects on the cortex and hence the
records. Accordingly, the next challenge was to have an anesthesiologist who
would help the team sort out this piece of the puzzle. Betty Grundy, MD, was
the person who enthusiastically joined the team and in her own right added a
great deal of knowledge to the process of making spinal cord monitoring a
viable clinical tool in the OR. She also became a voice within the anesthesia
profession that meticulous anesthesia protocols had to be followed for spinal
cord monitoring to be effective. Along the way, Rich became quite knowl-
edgeable regarding the various anesthetic agents used in spinal surgery to the
extent that he was a frequent presenter to anesthesia grand rounds on the
subject of their effects on cortical function. The final addition to the team was
Marianne Wilham, RN, the primary orthopedic OR nurse for the spinal surgi-
cal team and a critical person in maintaining a constant process in the OR. She
and Rich also became quite adept at dealing with teenage patients and parents
as they went through pre-operative spinal cord monitoring testing and the
next day trip to the OR for surgery.
Together this team meticulously developed protocols and systems that
seemed to provide the most consistent and reliable approach to intraoperative
spinal cord monitoring using Somatosensory Cortical Evoked Potentials
(SSEPs). Early in this process, several significant and revealing cases were
performed that were encouraging and confirmed the value of Rich Brown’s
“Big Blue” and the future of SSEPs. It should be noted that the “Wake-Up
Test” of Stagnara came into vogue about the same time as the CWRU work,
and it was adopted by the Case Team as a way to verify the findings of the
intraoperative changes seen in monitoring. One early case was a patient with
scoliosis and diastomatomyelia. It was elected to do the Harrington spinal
corrective surgery before removing the diastomatomyelia. Each time the
Harrington distraction was applied, the signals deteriorated and after removal
returned. The case was aborted with no neurological deficits. The diastomato-
myelia was removed and the subsequent spinal corrective surgery went for-
ward without incidence. Other early cases included a patient with cervical
spinal cord abscess in which artificially raising the blood pressure temporar-
ily restored SSEP responses and the clinical function. There was also a case
of cervical spinal cord hemangioma dissection that was performed success-
fully under the protective umbrella of SSEPs. Thus, these early anecdotal
experiences became convincingly indicative of the potential for intraopera-
tive spinal cord monitoring to make a great contribution to the safety of
patients undergoing major corrective spinal surgery. This monitoring tool
also proved to be one of the critical factors contributing to the development of
more and more powerful and corrective spinal implant systems that could be
applied in a safe manner.
It turns out that during the same time period, Dr. Tetsuya Tamaki and a
team of Japanese researchers including an anesthesiologist, Dr. K. Shimoji,
were independently working on a method for intraoperative spinal cord
Foreword to First Edition: Orthopedic Spine Viewpoint ix

­ onitoring using spinal – spinal evoked potentials. Before long there was
m
communication between the Case Team and Dr. Tamaki’s team to the extent
that a series of international spinal cord monitoring conferences were held,
the first being in Cleveland, Ohio, in 1977. At this first meeting, Dr. Vernon
Nickel, a highly respected orthopedic surgeon remarked to the gathering,
“One day intra-operative spinal cord monitoring will be as accepted and used
as the EKG.” One of the key individuals in this movement to develop intraop-
erative spinal cord monitoring was a neurosurgeon, Dr. J. Schramm, from
Germany. The list of participants continued to grow both in the United States
and throughout the world in great measure because of the encouraging and
engaging efforts of Rich Brown whose nature was to share his ideas and
expertise freely with all who took an interest. Again this welcoming approach
was grounded in Rich’s passion for rigorous process, analysis, expertise, and
training. Similarly he was cautious and scientifically reluctant to prematurely
declare SSEPs as the “Gold Standard” for monitoring spinal cord function
replacing the tried and true “Wake-Up Test”. As a final note, Rich never “went
commercial” with his system and expertise, but rather directed his efforts into
organizing the experts in the field and establishing standards of nomencla-
ture, processes, and technical training. He was an energetic founding member
and later a president of the American Society of Neurophysiologic Monitoring
to which he remained committed and focused until his untimely death.
All who have gone before would applaud this valuable book, and in par-
ticular Rich Brown, PhD, who very early on recognized the critical role that
anesthesia and anesthesiologists would play in the development and practice
of intraoperative spinal cord monitoring.
And the rest is history.

Cleveland, OH Clyde L. Nash Jr., MD

September, 2011
Foreword to First Edition: Peeling Back the
Onion Skin Layers

“As natural selection works solely by and for the good of each being, all corporeal and
mental endowments will tend to progress towards perfection.”
(Charles Darwin: The Origin of The Species, XV, 1859)

My! How times have changed! As I write this, I am looking at a copy of an

anesthetic record from June 14, 1968. Being a perpetual “pack rat,” I made it
a habit over the years to file cases of interest and needless to say, accumulated
quite a library over the past 50 years. This case, (Fig. 1), is that of a 3-month-
old baby with a diagnosis of cranial synostosis with orbital compression and
the operative procedure was in three stages, the final one occurring 2 months
later and involved a ventricular peritoneal shunt using the-then relatively new
silastic Holter valve. In this sick and lethargic baby, local anesthetics (carbo-
caine) supplemented with sedation were used over the period of 3 h and
50 min. Specific monitors included a blood pressure cuff and temperature
probe. For neuromonitoring, we considered ourselves “advanced” as we
employed a unit that we nicknamed the “bullet” or “torpedo,” since it had a
cylindrical shape with a diameter of about 6 in. and a length of 1.0 foot! One
end had a transparent viewplate with the tube containing a cathode ray tube
and the electronics for a one-channel electrocardiogram, lead II, and another
single channel for an electroencephalogram lead, using a parietal presenta-
tion. Since explosive agents were in use at that time, the “bullet” had an
explosion proof casing and elevated on a tripod above the 5 foot explosive
level. So now we were able to visualize the EEG, EKG, and measure the heart
rate with clicks triggered by the Q-T complex. If we now move 16 years to
1994, we can note the emergence of a neuromonitoring culture as demon-
strated by the book edited by Peter Sebel and William Fitch, Monitoring the
Central Nervous System [1]. There, 21 authors discussed a range of topics
which are extraordinary when compared to the availability of neuromonitor-
ing facilities in the 1960s. In this time period the horizon of neuromonitoring
is expanded to not only include physiochemical topics as cerebral blood flow
and metabolism, ICP, and EEG, but critical aspects relating to memory,
recovery from anesthesia, cognitive factors, and brain death. Fast forward to
today and to the wonderful effort made by the authors of the present-day book
to present a sophisticated review of the great advances in neuromonitoring
and its application to patient care as well as increasing our understanding of
the complexities not only of the central nervous system but the incredible
relationships among electrodynamic and electrochemical signaling that lead

xi
xii Foreword to First Edition: Peeling Back the Onion Skin Layers

Fig. 1 A 1968 record of a pediatric case with the anesthesia provided by the author. Total monitoring included systolic
blood pressure, temperature, heart rate, lead II of the EKG, and one EEG lead
Foreword to First Edition: Peeling Back the Onion Skin Layers xiii

to cognitive changes which may affect modalities such as pain. Similarly, the
effects of our monitoring efforts may be in themselves modified by the clini-
cal medium of anesthesia and cause a shift in the paradigm, which in a sense
involves monitoring the monitors and helps to eliminate false assumptions [2,
3]. The expertise and experience of the authors contribute greatly to a sense
of true security that these methodologies have been tested by those knowl-
edgeable in their field. Further cementing the link between development and
application are the hard-nosed Case-Based Presentations of practitioners
often highlighting those on both sides of the procedure table. This type of
hegemony is critical for carrying out many of these procedures. This book has
important source material even for those not directly connected with the
many procedures listed in the Table of Contents, for many of the authors are
not only capable as practitioners, but have had a primary role in developing
the many neuromonitoring techniques listed.
Before terminating this Preface, I must take a moment to pay homage to
one, who, in many ways is regarded as the “Mother” of neuromonitoring in
the anesthesia and neurological community, namely, Betty Grundy, MD. I
have known Betty for more than 40 years and can attest to how hard she has
worked to bring electrophysiological monitoring into the operating room and
clinical arena as well as educating a whole host of superb clinicians and those
doing research in this area.

References
1. Sebel P, Fitch W, editors. Monitoring the central nervous system. London:
Blackwell Science; 1994. p. 479.
2. Kuhn T. The structure of scientific revolutions. Chicago, IL: University of
Chicago Press; 1970. p. 226.
3. Popper K. Falsification versus conventionalism. In: David Miller, editor.
Popper selections. Princeton, NJ: Princeton University Press; 1985:
p.143–151

Birmingham, USA Maurice S. Albin, MD., M.Sc. (Anes.)

September, 2011
Foreword to First Edition: Neurosurgeon’s
Viewpoint

I am honored to have been invited to provide a foreword to this important

volume. As a practicing cerebrovascular surgeon, I have a unique perspective
on the field of neuromonitoring as I function somewhat as a “consumer” of
these very valuable resources. Vascular surgeons are charged with exposing
the brain, retracting brain tissue, reconstructing complex vascular anatomy,
temporarily interrupting cerebral blood flow, and performing complex revas-
cularizations. Not infrequently our target organ is already diseased and dys-
autoregulated at the time we expose it. While we have marvelous technologies
to allow us to perform computerized image guidance, highly magnified
3-dimensional views, and microsurgical instrumentation that allows extraor-
dinary things to be done, we perform this invasive maneuvers blinded as to
how the brain is tolerating these actions. Neuromonitoring, when performed
by skilled technologists and physicians with high expertise in the interpreta-
tion of data provide the surgeon with actionable information that can prove
lifesaving.
From my perspective, one of the most interesting aspects of contemporary
neuromonitoring lies in the domain of systems-based practice and communi-
cation. The surgeon often feels like a pilot of an aircraft in which he or she
has certain control capabilities but because the door is closed behind the pilot,
he or she has essentially no knowledge of what is happening in the rest of the
aircraft. It is critical that in our surgical environments the “door” remains
open and that the key human elements have professional confidence in each
other and communicate openly. At the start of the procedure, everyone
responsible including the physicians, technologists, and nurses must under-
stand the nature of the planned procedure, important details about the patient,
the general phases of the operation expected, and when the critical moments
will be occurring. As each stage of the procedure unfolds, the entire team
must be aware of those transitions. When an unexpected anomaly develops, a
rapid assessment of its significance must be performed followed by direct
communication with the surgeon. A timely but deliberate discussion of the
options to be considered and which one to be pursued assures the optimal
environment for the patient’s successful outcome.
It cannot be over emphasized that successful surgical neuromonitoring
requires a coordinated team effort. The critical elements obviously include
in-depth knowledge of the principles of neuromonitoring and technical profi-
ciency. Yet, without a full understanding of the patient’s physiologic state
prior to surgery and the unique aspects of patient positioning, abnormal data

xv
xvi Foreword to First Edition: Neurosurgeon’s Viewpoint

may be misinterpreted. The principles of the planned surgical procedure must

be understood by all team members and constant communication must occur
among the key participants to assure that proper perspective of the environ-
ment is obtained prior to the announcement of an abnormal finding.
Clearly this book will be a significant benefit to surgeons, technologists,
neurophysiologists, anesthesiologists, and neurologists. The information
contained in these chapters will empower the surgical team members with the
knowledge needed to interpret unexpected changes and to react quickly and
appropriately. This book will be an important reference for all members of
these teams and hopefully enhance our ability to provide safe procedures
with optimal outcomes.
Chicago, IL
September, 2011

 H. Hunt Batjer, M.D., F.A.C.S.

Department of Neurological Surgery, School of Medicine
Northwestern University Feinberg
Chicago, IL, USA
UT Southwestern Medical Center
Dallas, TX, USA
Preface to Second Edition

Intraoperative monitoring (IOM) of the nervous system continues to play a

key role for safeguarding neurological function during surgery and interven-
tional procedures when the nervous system is at risk for injury. For many
procedures, it has been integrated as a key component of decision making and
a variety of studies have shown a clear association of its use with improved
outcomes. The utilization of monitoring continues to evolve as monitoring
techniques are improved and developed and their contributions to improved
patient care are better understood. As such, we are pleased to present this
second edition which reflects these changes.
We continue to be grateful to the many past and present pioneers in the
field who have laid the groundwork for modern day monitoring and who
continue to fuel the evolution of its techniques and applications.
Anesthesiologists have played a key role in this evolution with improvements
in neuroanesthesia and their interface with monitoring. In this capacity, we
wish to celebrate the life and acknowledge the key role of Maurice Albin,
who passed in 2016. We are honored that Dr. Maurice Albin wrote the anes-
thesiologist viewpoint foreword to the first edition of this book. Similarly,
many surgeons and interventionists have expanded the role of monitoring
into existing and innovative new procedures. Finally, many neurophysiolo-
gists have also expanded our understanding of the role of monitoring and
have developed enhanced techniques to meet the needs of various procedures.
We owe a great deal of gratitude to all of these individuals as these develop-
ments have been included in this new expanded text.
As with the first edition of the book, the major theme of this edition is to
emphasize the roles of all members of the procedure team cooperatively
working together to provide the patient with the most effective techniques for
ensuring an optimal outcome. Since this involves education across specialty
boundaries, this book continues to take a holistic approach by discussing
modalities and their application during various procedures. The second edi-
tion now has five news, key learning points, questions and answers and has
expanded the learning opportunities to online resources including videos and
hyperlinks to PubMed so as to enhance the text content. We are grateful to the
publisher and the electronic resources that are available to include these addi-
tions. We hope you will find that they are helpful in your patient care and
eagerly look forward to further advancements in monitoring techniques and
their improved understanding and application.

xvii
xviii Preface to Second Edition

The international reach of the first edition of this book included both
English and Chinese. With this edition, it has been expanded to include the
Japanese and Korean languages as well. We are honored and grateful.

Chicago, IL, USA Antoun Koht

Aurora, CO, USA  Tod B. Sloan
Chicago, IL, USA  J. Richard Toleikis
Preface to First Edition

Intraoperative monitoring of the nervous system (IOM) has become common

place in orthopedics, neurosurgery, otologic surgery, vascular surgery, and
other procedures. In addition to the improvement in patient outcome which
has been observed in several circumstances, the monitoring has been incorpo-
rated into the management of surgical procedures where the nervous system
is at risk. The use is being fueled by the understanding that functional knowl-
edge of the nervous system is an important partner to structural knowledge
and both contribute to the quality of care and patient safety.
IOM is more than just a tool like fluoroscopy, intraoperative MRI, or CT
scanning which gives a structural view of the patient’s anatomy. IOM pro-
vides a means for assessing the nervous system function and determining
how the surgical, anesthetic, and physiological environment are impacting
this function. Pamela Prior expressed it well in 1985 when she said “routine
clinical monitoring of ECG, arterial pressure and blood-gas tensions only
indicates the adequacy of factors supporting brain function. The EEG and
evoked potentials are more valuable because they can monitor continuously
the end result at a neuronal functional level.”[1] This “window to the nervous
system” allows all of us to bring our own contributions to help our patients
have the best possible outcomes. The prompt diagnosis of circumstances
unfavorable to the nervous system will enable timely adjustment of the phar-
macologic and physiologic environment to augment surgical decision.
IOM has evolved in the last 30 years from the lonely somatosensory
evoked potentials (SSEP) modality that was used during spine surgery to now
include; MEPs, both free running and triggered EMG, D waves, the H reflex,
and other monitoring modalities. This expansion was not restricted to spine
surgery but extended to other surgeries including those of the head and neck.
The addition of IOM multimodalities allowed a more comprehensive assess-
ment of the nervous system while adding restraint on the anesthetic tech-
nique. The optimal anesthetics for one modality is often not the same as that
for others, thus a very delicate anesthetic balance is needed and complete
cooperation between the IOM team and the anesthesiologist is invaluable.
This team effort is the key to the best patient outcome. Clearly the moni-
toring is helpful to the surgeon, but it is equally valuable to the anesthesiolo-
gist. In that respect, IOM allows the anesthesiologist to see the impact of the
anesthetic and physiologic management on the functional integrity of the ner-
vous system. For example, there is a growing appreciation that a blood pres-
sure which might be appropriate for one patient may not be adequate for

xix
xx Preface to First Edition

another. Further complicating factors are the aging patients with increased
comorbidities and the more complex surgical procedures with increasing
neurological trespass. IOM can help the anesthesiologist insure that the envi-
ronment of the nervous system is optimal for the individual and to adjust the
patient’s physiology as needed when the surgical procedure places an addi-
tional stress on the nervous system.
This interplay of anesthesia, physiology, and surgery is what makes IOM
different from the use of these techniques for diagnostic assessment of pathol-
ogy in the nervous system. The situation is dynamic with a constantly shifting
equilibrium of the effects of the procedure, the drugs given, and the physio-
logical milieu. This is why IOM, like monitoring of blood pressure, heart
rate, oxygenation, etc., must be done constantly to identify changes that allow
rapid correction while adverse neurological circumstances are still reversible.
Some of that reversibility will be contained in the surgical maneuvers, but
changes in the management of the anesthesia, physiology, and positioning of
the patient can mitigate some of the adverse effects of the procedure.
Well recognized by anesthesiologists, each patient is different, not only in
their pathology and comorbidities, but also in how they will react to anesthe-
sia and the surgical procedure. Each patient therefore presents a different
problem. An injury could develop and progress without the surgeon’s, proce-
duralist’s, or anesthesiologist’s knowledge. This is where IOM can become
valuable to identify functional changes in the nervous system which will not
be observed in structural studies or reflected in other means of traditional
monitoring.
To make the team effort most effective, each member of the team needs to
understand each other’s roles. Like an interlocking crossword puzzle, the
interface of each other’s contribution is made stronger when each one knows
about the other and the more effective the team becomes. This book is
designed to help all members of the operative team to better understand what
each member of the team is doing. It is not designed to provide technical
details since there are many excellent papers and books on that subject.
Rather we have sought to allow everyone an opportunity to gain insights into
each of the operative components.
Many of the early applications of IOM were developed in the 1970s by
surgeons, neurophysiologists, anesthesiologists, and other researchers both in
the USA and Japan, as they recognized that the development of aggressive
treatment programs carried a high risk of secondary spinal cord damage and
that there was a need to develop methodologies for defining and evaluating
spinal cord function. Among these were Clyde Nash, MD, and Richard
Brown, PhD, who pioneered the use of SSEPs during Harrington distraction
of the spine in patients with scoliosis [2]. This advance from the intraopera-
tive wake-up test of Vauzelle and Stagnara would become increasingly impor-
tant as procedures presented multiple possible injurious steps [3]. In patients
with many significant comorbidities, a one-time clinical assessment which
was used in healthy young patients with scoliosis is not applicable. The pio-
neers of IOM not only developed the techniques for monitoring, but they
designed and built equipment to meet the specific challenges present in the
operating room that were not encountered in the diagnostic laboratory. They
Preface to First Edition xxi

also recognized the importance of the team effort and that such things as
blood pressure management during distraction of the spine was essential for
overcoming the effects of the procedure [2]. With an awareness that others
were beginning to address the need for monitoring spinal cord function,
Clyde Nash and Jerald Brodkey invited participants from throughout the
world and hosted the first two symposia on spinal cord monitoring which
were held in Cleveland in September 1977 and St. Louis in January 1979.
These were followed by a series of International Symposia on spinal cord
monitoring, the first of which was held in Tokyo, Japan, in 1981 and was
hosted by Dr. Tetsuya Tamaki. Three years later, Dr. Johannes Schramm
hosted the Second International Symposium held in Erlangan, Germany
(1984). The Third and Fourth Symposia were later held in Annapolis,
Maryland (in 1986) and Niigata, Japan (in 1989), and were hosted by Drs.
Thomas Ducker and Richard Brown and by Dr. Koki Shimoji, respectively.
Special thanks to these early pioneers who recognized the importance of this
new technology and worked to strengthen it and expand its usage. Subsequent
to the International Symposia came the formation of the American Society of
Neurophysiologic Monitoring (ASNM) in 1989, and also the advent of the
International Symposia on Intraoperative Neurophysiological Monitoring in
Neurosurgery held in New York and hosted by Drs. Vedran Deletis and Fred
Epstein (1998–2006). From these latter symposia came the formation of the
International Society of Intraoperative Neurophysiology (ISIN) in 2006.
IOM has evolved from the early days. Some of the techniques currently
used are refinements of the early techniques while others are completely new.
The monitoring professionals have recognized that the changes in neurophys-
iology that result from anesthesia and surgery are different from those seen in
the laboratory which makes diagnostic approaches less applicable. Further,
IOM must be done with constant, rapid updates to provide timely information
about the state of the nervous system. This evolution in techniques has been
accompanied with the development of a new field of intraoperative neuro-
physiology with professionals who have dedicated their career to IOM. The
backgrounds of these individuals are as diverse as the techniques currently
being employed. Some come from the logical pioneering fields of orthopedic
surgery, neurosurgery, neurology, and anesthesiology. But a whole new field
of intraoperative neurophysiology has developed with individuals bringing to
bear their knowledge of intraoperative neurophysiology with the many allied
medical fields to provide focused IOM care. These individuals have been
responsible for many developments in the field and are key to the current
utilization of monitoring for providing excellent patient care.
Many of the early developments of IOM can also be attributed to anesthe-
siologists. Recently, Tamaki (orthopedic surgeon) wrote an article about the
history of EP monitoring and credited Shimoji (anesthesiologist) with intro-
ducing epidural evoked potential monitoring in 1971 [4]. Betty Grundy, MD,
as an anesthesiologist involved in the early applications of IOM recognized
this in 1982 when she wrote about the application of auditory evoked poten-
tials in surgery on the brainstem in the Journal of Neurosurgery “we wanted
early indication of deteriorating function so that we could intervene to pre-
vent permanent injury. We therefore selected an approach similar to that used
xxii Preface to First Edition

for intraoperative monitoring of other physiological parameters such as heart

rate or arterial blood pressure, attempting to correct undesirable trends as
soon as these could be identified with certainty” [5].
Dr. Grundy went on to bring IOM into anesthesiology; her landmark arti-
cle in Anesthesiology in 1983 was a call for anesthesiologists to take an active
role in the team. She noted that “the hope is that deteriorating neurologic
function will be detected early so that the surgeon and/or anesthesiologist can
intervene to optimize function and minimize the possibility of permanent
damage to the nervous system” [6]. Dr. Grundy was to further stress that role
when she wrote in 1984 “the anesthesiologist has important responsibilities
in facilitating the electrophysiological monitoring. A multiplicity of factors
under their control of the anesthesiologist can alter evoked potentials” [7].
Her early experience noted the interaction of anesthesia, physiology, and the
nervous system which supported her recommendations for anesthetic and
physiological management; without IOM many unfavorable interactions
would have gone unrecognized. These observations are still echoed today.
As IOM techniques and applications have evolved, some advancements
have come from anesthesiologists. In particular, the anesthetic techniques and
physiological management that supports IOM, and which have been refined
from observations made by IOM, have also improved patient care. Many
anesthesiologists remain actively involved in IOM and are contributors to this
book.
As the field of IOM has developed, the cadre of IOM professionals that
has emerged to provide the best neurophysiological monitoring has been a
distraction from the integral role of anesthesiologists in the IOM team. As
such, this book is devoted to restoring that role by focusing on the knowledge
and experience gained by anesthesiologists and professionals who are part of
the IOM team. Our goal is to facilitate the most effective team effort by
expanding the interface of knowledge between the surgical, anesthesiologi-
cal, and neurophysiological members.
The first section describes the different techniques used in monitoring.
The goal is to provide insight into the anatomy, physiology, and techniques so
that the information provided by their use can be placed in the context of the
surgical, anesthetic, and physiological management.
The second section seeks to provide basic aspects of anesthetic manage-
ment. Not only will this be helpful to anesthesia providers seeking to refine
their choice of medications, but it also will be helpful to practitioners in other
specialties to understand the challenges inherent in the anesthetic manage-
ment. Some anesthesiologists are concerned about anesthesia without muscle
relaxants while other members of the team may be concerned about any use
of muscle relaxants. The contributions of the authors will be helpful to reas-
sure both that it is possible to meet this need and successfully obtain optimal
signals which will enable the team to effectively monitor the patients and for
the surgeon to make the best decision.
Finally, the book provides case examples of specific types of procedures
where IOM has become a routine part of the management. In each case the
chapter provides an overview of the anatomy, neural physiology, and pathol-
ogy which is central to the procedure. Understanding this allows each ­member
Preface to First Edition xxiii

of the team to understand how the procedure, anesthesia, physiology, and

IOM come to bear on the risks of the procedure and outcome. In each case,
the authors have also presented some examples of typical IOM changes in
these cases. This allows discussion of the differential diagnosis of the effects
which could cause these changes. In that respect, the emphasis has been on
non-surgical effects to allow better insight into the ways that the management
of anesthesia, positioning, and physiology can contribute to improved
outcome.
We have assembled a prestigious group of contributors who are all actively
involved in the team efforts of IOM during various surgical procedures. Each
has contributed their knowledge and experience to improve all of our effec-
tiveness in these procedures. Hopefully, by sharing our knowledge and expe-
rience we can make the fabric of our team efforts stronger and provide the
best possible care of our patients.

References

1. Prior PF. EEG monitoring and evoked potentials in brain ischaemia. Br J

of Anaesth. Jan 1985;57(1):63–81.
2. Nash CL, Jr., Lorig RA, Schatzinger LA, Brown RH. Spinal cord monitor-
ing during operative treatment of the spine. Clin Orthop Relat Res. Jul-­
Aug 1977(126):100–105.
3. Vauzelle C, Stagnara P, Jouvinroux P. Functional monitoring of spinal
cord activity during spinal surgery. Clin Orthop Relat Res. Jun
1973(93):173–178.
4. Tamaki T, Kubota S. History of the development of intraoperative spinal
cord monitoring. Eur Spine J. 2007; 16 Suppl 2:S140–146.
5. Grundy BL. Monitoring of sensory evoked potentials during neurosurgi-
cal operations: methods and applications. Neurosurgery. Oct
1982;11(4):556–575.
6. Grundy BL. Intraoperative monitoring of sensory-evoked potentials.
Anesthesiology. Jan 1983;58(1):72–87.
7. Grundy BL. Evoked potentials in the operating room. Mt Sinai J Med.
1984;51(5):585–591.

Chicago, IL, USA Antoun Koht

Aurora, CO, USA  Tod B. Sloan
Chicago, IL, USA  J. Richard Toleikis
September, 2011
Acknowledgment

We want to acknowledge our families; (TS) Celia, Wendy, and Heather; (JRT)
Sandra, Jennifer Anne, Matthew, Jason, and Jennifer Rachel; (AK) Sonia,
Yara, John, and Alexander who have provided love, support, and understand-
ing during many late hours of monitoring and without which the creation of
this book would not have been possible.

xxv
Contents

Section I Monitoring Techniques

1 Somatosensory-Evoked Potentials...............................................3
Aimee Becker, Corey Amlong, and Deborah A. Rusy
2 Transcranial Motor-Evoked Potentials.......................................19
Leslie C. Jameson
3 Auditory-Evoked Potentials.........................................................35
Christoph N. Seubert and Mary Herman
4 Visual-Evoked Potentials..............................................................51
Sandra C. Toleikis and J. Richard Toleikis
5 Deep Brain Stimulation................................................................71
Jay L. Shils, Diana Apetauerova, Amal A. Mokeem,
and Jeffrey E. Arle
6 Monitoring of Spinal Cord Functions.........................................87
Sumihisa Aida, Tatsuro Kohno, and Koki Shimoji
7 Electromyography.........................................................................103
J. Richard Toleikis
8 The Use of Reflex Responses for IOM........................................125
Ronald Leppanen
9 Brain and Spinal Cord Mapping.................................................155
Charles D. Yingling and Tina N. Nguyen
10 EEG Monitoring...........................................................................169
Ira J. Rampil
11 Clinical Application of Raw and Processed EEG.......................193
Phillip E. Vlisides and George A. Mashour

xxvii
xxviii Contents

12 A Guide to Central Nervous System Near-Infrared

Spectroscopic Monitoring............................................................205
Harvey L. Edmonds Jr., Michael R. Isley, and Jeffrey R. Balzer
13 Transcranial Doppler Ultrasound...............................................219
Harvey L. Edmonds Jr.
14 Monitoring of Jugular Venous Oxygen Saturation....................229
Deepak Sharma and Abhijit Lele
15 Intracranial Pressure Monitoring...............................................243
Ross Martini, Andrea Orfanakis, and Ansgar Brambrink
16 IOM Instrumentation Layout and Electrical Interference.......253
Brett Netherton and Andrew Goldstein
17 Signal Optimization in Intraoperative Neuromonitoring.........269
Robert E. Minahan and Allen S. Mandir

Section II Anesthesia Considerations

18 Anesthesia for Awake Neurosurgery...........................................301

Antoun Koht, Georg Neuloh, and Matthew C. Tate
19 Anesthesia Management and Intraoperative
Electrophysiological Monitoring.................................................317
Tod B. Sloan

Section III Clinical Applications

20 Monitoring Applications and Evaluating Changes....................345

Antoun Koht, Tod B. Sloan, and J. Richard Toleikis
21 Intraoperative Neurophysiological Monitoring
for Intracranial Aneurysm Surgery............................................353
Laura B. Hemmer, Carine Zeeni, Bernard R. Bendok,
and Antoun Koht
22 Intracranial Arteriovenous Malformation Surgery...................367
Laura B. Hemmer and Carine Zeeni
23 Intraoperative Neurophysiologic Monitoring During Surgery
for Supratentorial Mass Lesions..................................................377
Georg Neuloh, Antoun Koht, and Matthew C. Tate
24 Surgery for Infratentorial Mass..................................................385
Michael J. Malcharek and Gerhard Schneider
25 Trigeminal Microvascular Decompression.................................395
Antoun Koht
26 Surgery for Hemifacial Spasm.....................................................409
Raymond F. Sekula Jr., Jeffrey R. Balzer, Jesse D. Lawrence,
and Penny P. Liu
Contents xxix

27 Skull Base Surgery........................................................................425

David E. Traul and Thomas N. Pajewski
28 Surgery for Chiari Type I Malformation....................................435
Penny P. Liu, Chaim I. Nelson, Gregory D. Arnone,
Ashley Kane Palmer, and Raymond F. Sekula Jr.
29 ENT and Anterior Neck Surgery.................................................445
W. Scott Jellish and Michail Avramov
30 Carotid Surgery............................................................................459
Zirka H. Anastasian, Eugene Ornstein, and Eric J. Heyer
31 Anterior Cervical Spine Surgery.................................................473
John F. Bebawy, Antoun Koht, and Srdjan Mirkovic
32 Posterior Cervical Spine Surgery................................................485
Paul D. Mongan and Vikas V. Patel
33 Surgery for Scoliosis Correction..................................................497
Mary Ellen McCann, Robert M. Brustowicz,
and Sulpicio G. Soriano
34 Neurophysiological Monitoring in Thoracic Spine Surgery.....507
Tod B. Sloan, Evalina Burger, Christopher J. Kleck,
and Anthony M. Oliva
35 Intraoperative Neurophysiologic Monitoring
for Lumbosacral Spine Procedures.............................................525
Deborah A. Rusy, Corey Amlong, and Aimee Becker
36 Intramedullary Spinal Cord Surgery.........................................535
Beate Poblete and Karl F. Kothbauer
37 Intraoperative Monitoring in Tethered Cord Surgery..............549
Daniel J. Janik and Claudia F. Clavijo
38 Surgery in the Peripheral Nervous System................................563
Leo T. Happel and David G. Kline
39 Surgery of the Aortic Arch...........................................................575
K. Annette Mizuguchi, Linda S. Aglio, and Laverne D. Gugino
40 Electrophysiological Monitoring During Thoracic Aortic
Aneurysm Surgery........................................................................601
Tod B. Sloan, Leslie C. Jameson, and Claudia F. Clavijo
41 Monitoring During Cardiopulmonary Bypass...........................617
Harvey L. Edmonds Jr.
42 Interventional Neuroradiology....................................................625
Anthony K. Sestokas and Daniel M. Schwartz
43 Intraoperative Neuromonitoring in Pediatric Surgery.............633
Lisa Francis, Veronica Busso, and John J. McAuliffe
xxx Contents

Section IV Intensive Care

44 Monitoring in the Intensive Care Unit........................................ 653

Louanne M. Carabini
45 Epilepsy and Seizures: OR and ICU Applications of EEG.......663
Sabrina G. Galloway and Tod B. Sloan
46 Monitoring Cerebral Blood Flow................................................ 681
W. Andrew Kofke and Bonnie H. Wang
Afterword to First Edition...................................................................  699
Johannes Schramm
Index....................................................................................................... 703
Contributors

Linda S. Aglio, MD Department of Anesthesiology, Perioperative and Pain

Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston,
MA, USA
Sumihisa Aida, MD, PhD Geriatric Health Service Facility, Izumi, Adachi-Ku,
Tokyo, Japan
Corey Amlong, MD, MS Department of Anesthesiology, University of
Wisconsin School of Medicine and Public Health, Madison, WI, USA
Zirka H. Anastasian, MD Department of Anesthesiology, Columbia
University, New York, NY, USA
Diana Apetauerova, MD Department of Neurology, Lahey Hospital and
Health System, Burlington, MA, USA
Jeffrey E. Arle, MD, PhD Department of Neurosurgery, Beth Israel
Deaconess Medical Center and Harvard University, Boston, MA, USA
Gregory D. Arnone, MD Department of Neurosurgery, Allegheny General
Hospital, Pittsburgh, PA, USA
Michail Avramov, MD, PhD Department of Anesthesiology, Loyola
University Health System, Maywood, IL, USA
Jeffrey R. Balzer, PhD Department of Neurological Surgery, University of
Pittsburgh Medical Center, Pittsburgh, PA, USA
John F. Bebawy, MD Department of Anesthesiology and Neurological
Surgery, Northwestern University, Chicago, IL, USA
Aimee Becker, MD Department of Anesthesiology, University of Wisconsin
School of Medicine and Public Health, Madison, WI, USA
Bernard R. Bendok, MD Department of Neurologic Surgery in Arizona,
Mayo Clinic College of Medicine, Mayo Clinic Arizona, Mayo Clinic
Hospital, Phoenix, AZ, USA
Ansgar Brambrink, MD Department of Anesthesiology and Perioperative
Medicine, Oregon Health and Science University, Portland, OR, USA
Department of Anesthesiology, Columbia University College of Physicians &
Surgeons, New York, NY, USA

xxxi
xxxii Contributors

Robert M. Brustowicz, MD Department of Anesthesiology, Perioperative

and Pain Medicine, Boston Children’s Hospital, Boston, MA, USA
Department of Anaesthesia, Harvard Medical School, Boston, MA, USA
Evalina Burger, MD Department of Orthopaedics, University of Colorado,
Aurora, CO, USA
Veronica Busso, MD Department of Anesthesiology, Cincinnati Children’s
Hospital Medical Center, Cincinnati, OH, USA
Louanne M. Carabini, MD Department of Anesthesiology, Section of
Critical Care Medicine, Northwestern University Feinberg School of Medicine,
Chicago, IL, USA
Claudia F. Clavijo, MD Department of Anesthesiology, University of
Colorado School of Medicine, Aurora, CO, USA
Harvey L. Edmonds Jr., PhD Department of Anesthesiology and Perioperative
Medicine, University of Louisville School of Medicine, Louisville, KY, USA
Lisa Francis, DO Department of Anesthesiology, Cincinnati Children’s
Hospital Medical Center, Cincinnati, OH, USA
Sabrina G. Galloway, BS, REEG/EP T, CNIM, CLTM, FASET
Department of Neurology, SUNY Downstate Medical Center, Brooklyn, NY,
USA
Neurodiagnostic Operations, Neuromonitoring Technologies, Inc Glenwood,
MD, USA
Andrew Goldstein, BS, CNIM Manager Biomedical Services, IONM,
SpecialtyCare, Brentwood, TN, USA
Laverne D. Gugino, MD Department of Anesthesiology, Perioperative and
Pain Medicine, Brigham and Women’s Hospital, Harvard Medical School,
Boston, MA, USA
Leo T. Happel, PhD Department of Neurosurgery, LSU Health Science
Center, New Orleans, LA, USA
Laura B. Hemmer, MD Departments of Anesthesiology and Neurological
Surgery, Northwestern University Feinberg School of Medicine, Chicago, IL,
USA
Mary Herman, MD, PhD Department of Anesthesiology, The Geisinger
Health System, Danville, PA, USA
Eric J. Heyer, MD, PhD Department of Neurological Surgery, Columbia
University, New York, NY, USA
Michael R. Isley, PhD, DABNM, FASNM Department of Intraoperative
Neuromonitoring, Orlando Regional Medical Center, Arnold Palmer Hospital
for Children, Orlando, FL, USA
Leslie C. Jameson, MD Department of Anesthesiology, School of Medicine,
University of Colorado, Aurora, CO, USA
Contributors xxxiii

Daniel J. Janik, MD Department of Anesthesiology, University of Colorado

School of Medicine, Aurora, CO, USA
W. Scott Jellish, MD, PhD Department of Anesthesiology, Loyola
University Health System, Maywood, IL, USA
Christopher J. Kleck, MD Department of Orthopaedics, University of
Colorado, Aurora, CO, USA
David G. Kline, MD Department of Neurosurgery, LSU Health Science
Center, New Orleans, LA, USA
W. Andrew Kofke, MD, MBA, FCCM, FNCS Department of
Anesthesiology and Critical Care, University of Pennsylvania, Philadelphia,
PA, USA
Department of Neurosurgery, University of Pennsylvania, Philadelphia, PA, USA
Tatsuro Kohno, MD, PhD Division of Anesthesiology, Niigata University
Graduate School of Medical and Dental Sciences, Niigata, Japan
Antoun Koht, MD Departments of Anesthesiology, Neurological Surgery
and Neurology, Northwestern University Feinberg School of Medicine,
Chicago, IL, USA
Karl F. Kothbauer, MD Division of Neurosurgery, Luzerner Kantonsspital,
Luzern, Switzerland
Jesse D. Lawrence, BS Department of Neurological Surgery, University of
Pittsburgh Medical Center, University of Pittsburgh School of Medicine,
Pittsburgh, PA, USA
Abhijit Lele, MD Departments of Anesthesiology & Pain Medicine and
Neurological Surgery, Harborview Medical Center, University of Washington,
Seattle, WA, USA
Ronald Leppanen, PhD Clinical Neurophysiologist, Knoxville Neurology
Clinic, Knoxville, TN, USA
Penny P. Liu, MD Division of Neuroanesthesia, Department of
Anesthesiology, Tufts Medical Center, Boston, MA, USA
Michael J. Malcharek, MD, PhD Division of Neuroanaesthesia and
Intraoperative Neuromonitoring, Department of Anaesthesiology, Intensive
Care and Pain Therapy, Leipzig, Saxony, Germany
Allen S. Mandir, MD, PhD Department of Neurology, Medstar Georgetown
University Hospital, Washington, DC, USA
Ross Martini, MD Department of Anesthesiology and Perioperative Medicine,
Oregon Health and Science University Hospital, Portland, OR, USA
George A. Mashour, MD, PhD Department of Anesthesiology, 1H247
University Hospital, University of Michigan Medical School, Ann Arbor, MI,
USA
xxxiv Contributors

John J. McAuliffe, MD, MBA, DABNM Department of Anesthesiology,

Cincinnati Children’s Hospital Medical Center, Cincinnati, OH, USA
Mary Ellen McCann, MD Department of Anesthesiology, Perioperative
and Pain Medicine, Boston Children’s Hospital, Boston, MA, USA
Department of Anaesthesia, Harvard Medical School, Boston, MA, USA
Robert E. Minahan, MD Department of Neurology, Medstar Georgetown
University Hospital, Washington, DC, USA
Srdjan Mirkovic, MD Department of Orthopedic Surgery, Northwestern
University Feinberg School of Medicine, Chicago, IL, USA
K. Annette Mizuguchi, MD, PhD, MMSc Department of Anesthesiology,
Perioperative and Pain Medicine, Brigham and Women’s Hospital, Harvard
Medical School, Boston, MA, USA
Amal A. Mokeem, MD Department of Neurosciences, MBC 76, King
Faisal Specialist Hospital & Research Center, Riyadh, Saudi Arabia
Paul D. Mongan, MD Department of Anesthesiology, University of
Colorado Hospital, Aurora, CO, USA
Chaim I. Nelson, MD Department of Anesthesiology PGY3, Tufts Medical
Center, Boston, MA, USA
Brett Netherton, MS, CNIM, FASNM, FASET Signal Gear, LLC,
Prosperity, SC, USA
Georg Neuloh, MD Department of Neurosurgery, University of Aachen,
Aachen, Germany
Tina N. Nguyen, MD Department of Neurosurgery, Kaiser Redwood City,
CA, USA
Anthony M. Oliva, MD, PhD Department of Anesthesiology, University of
Colorado School of Medicine, Aurora, CO, USA
Andrea Orfanakis, MD Oregon Anesthesiology Group, Anesthesiology,
Critical Care Medicine—Anesthesia, Portland, OR, USA
Eugene Ornstein, PhD, MD Department of Anesthesiology, Columbia
University, New York, NY, USA
Thomas N. Pajewski, PhD, MD Division of Neuroanesthesiology,
Department of Anesthesiology, University of Virginia Health System,
Charlottesville, VA, USA
Ashley K. Palmer Charlottesville, VA, USA
Vikas V. Patel, MD Department of Orthopedic Surgery, University of
Colorado, Denver, CO, USA
Beate Poblete, MD Division of Anesthesiology, Luzerner Kantonsspital,
Luzern, Switzerland
Contributors xxxv

Ira J. Rampil, MS, MD Blue Sky Medicine, LLC, Williamson, GA, USA
Deborah A. Rusy, MD, MBA Department of Anesthesiology, University of
Wisconsin School of Medicine and Public Health, Madison, WI, USA
Gerhard Schneider, MD Department of Anesthesiology, Emergency
Medicine and Pain Therapy, University Witten/Herdecke, Helios Clinic
Wuppertal, Wuppertal, North Rhine-Westphalia, Germany
Johannes Schramm, M.D., Ph.D. Department of Neurosurgery, University
School of Medicine, Bonn, Germany
Daniel M. Schwartz, PhD Teaneck, NJ, USA
Raymond F. Sekula Jr., MD, MBA, FACS Department of Neurological
Surgery, University of Pittsburgh Medical Center, University of Pittsburgh
School of Medicine, Pittsburgh, PA, USA
Anthony K. Sestokas, PhD, DABNM, FASNM SpecialtyCare – IONM,
Nashville, TN, USA
Christoph N. Seubert, MD, PhD, DABNM Department of Anesthesiology,
University of Florida College of Medicine, Gainesville, FL, USA
Deepak Sharma, MBBS, MD, DM Departments of Anesthesiology & Pain
Medicine and Neurological Surgery, Harborview Medical Center, University
of Washington, Seattle, WA, USA
Jay L. Shils, PhD, DABNM, FASNM, FACNS Department of
Anesthesiology, Rush University Medical Center, Chicago, IL, USA
Koki Shimoji, MD, PhD Niigata University Graduate School of Medicine,
Niigata, Japan
Standard Medical Information Center, NPO, Pain Control Institute, Inc.,
Tokyo, Japan
Tod B. Sloan, MD, MBA, PhD Department of Anesthesiology, University
of Colorado School of Medicine, Aurora, CO, USA
Sulpicio G. Soriano, MD Department of Anesthesiology, Perioperative and
Pain Medicine, Boston Children’s Hospital, Boston, MA, USA
Department of Anaesthesia, Harvard Medical School, Boston, MA, USA
Matthew C. Tate, MD, PhD Department of Neurological Surgery,
Northwestern Memorial Hospital, Chicago, IL, USA
J. Richard Toleikis, PhD Department of Anesthesiology, Rush University
Medical Center, Chicago, IL, USA
Sandra C. Toleikis, MA Department of Anesthesiology, Rush University
Medical Center, Chicago, IL, USA
David E. Traul, MD, PhD Department of General Anesthesiology,
Cleveland Clinic, Cleveland, OH, USA
xxxvi Contributors

Phillip E. Vlisides, PhD Department of Anesthesiology, 1H247 University

Hospital, University of Michigan Medical School, Ann Arbor, MI, USA
Bonnie H. Wang, MD Department of Neurology, University of
Pennsylvania, Philadelphia, PA, USA
Charles D. Yingling, PhD Golden Gate Neuromonitoring, San Francisco,
CA, USA
Carine Zeeni, MD Department of Anesthesiology, American University of
Beirut Medical Center, Beirut, Lebanon
 Section I
Monitoring Techniques
 Somatosensory-Evoked Potentials
1
Aimee Becker, Corey Amlong,
and Deborah A. Rusy

Key Learning Points electrode placement (recording montage),

and specific equipment parameters, which
• The ultimate goal of intraoperative SSEP mon- include channel availability, filters, averaging,
itoring is to ensure maintenance of neurologic and time base.
integrity throughout a procedure with resultant • Most anesthetic agents have detrimental effects
improved outcome and decreased morbidity. on SSEPs, while a select few actually have
• General consensus is that standard SSEP beneficial effects. In general, cortical effects
recording monitors solely the dorsal column are more pronounced than peripheral effects.
pathway, which mediates mechanoreception • Several physiologic variables can affect the
and proprioception. However, other pathways success or failure of SSEP monitoring, includ-
may contribute to somatosensory function, ing patient temperature, blood pressure,
including the dorsal spinocerebellar tract, the hemoglobin levels, intracranial pressure, oxy-
anterolateral columns, the postsynaptic dorsal genation, and ventilation.
column pathway, and the vagus nerve. • Reproducible baseline waveforms are crucial
• Stimulation and recording are the two major in SSEP monitoring. Evidence-based recom-
technical aspects of SSEP monitoring; under- mendations on when to intervene when SSEP
standing the parameters that affect each is monitoring is altered from baseline are diffi-
critical to successful intraoperative SSEP cult to provide due to the low specificity of
monitoring. Stimulation parameters include SSEP monitoring. However, general consen-
electrode type, electrode placement, stimulus sus is that a 50 % amplitude reduction and
intensity, stimulus duration, stimulus rate, 10 % increase in latency, not attributable to
and unilateral versus bilateral stimulation. anesthetic or physiologic causes, are signifi-
Recording parameters include electrode type, cant changes that warrant intervention.

Intraoperative application of evoked potentials

A. Becker, M.D. (*) • C. Amlong, M.D., M.S.
D.A. Rusy, M.D., M.B.A. has evolved during the past 30 years, and
Department of Anesthesiology, University of somatosensory-­evoked potential (SSEP) monitor-
Wisconsin School of Medicine and Public Health, ing is the method most commonly employed [1].
B6/319, Clinical Sciences Center, 600 Highland The ultimate goal of intraoperative SSEP moni-
Avenue, Madison, WI 53792, USA
e-mail: [email protected]; [email protected]; toring is to ensure maintenance of neurologic
[email protected] integrity throughout a procedure with resultant

© Springer International Publishing AG 2017 3

A. Koht et al. (eds.), Monitoring the Nervous System for Anesthesiologists
and Other Health Care Professionals, DOI 10.1007/978-3-319-46542-5_1
4 A. Becker et al.

improved outcome and decreased morbidity. The Anatomy and Vascular Supply
premise of evoked potentials is simple. When
neural tissue is stimulated, either by true sensory The somatosensory system consists of the dorsal
or artificial electrical stimulation, ascending elec- column–lemniscal pathway (Fig. 1.1), or poste-
trical impulses—or volleys—are sent through rior column pathway, and the spinothalamic
synapses via neural pathways. Depending on pathway. The former pathway mediates mecha-
stimulation site and recording location, there is noreception and proprioception, whereas the lat-
characteristic waveform morphology of the vol- ter mediates thermoreception and nociception.
ley. Near-field potentials result when the neural The general consensus is that standard SSEP
impulse passes immediately beneath the reference recording monitors solely the dorsal column
electrode. Far-field potentials result from impulses pathway. However, other pathways may contrib-
distant to the recording electrodes. SSEPs are, in ute to somatosensory function, including the
general, mixed-field potentials [2]. The value of dorsal spinocerebellar tract, the anterolateral
intraoperative SSEP monitoring is derived columns, the postsynaptic dorsal column path-
from consistency—reproducible, recognizable way, and the vagus nerve [1, 3].
waveforms—such that meaningful conclusions The pathway of the dorsal column–lemniscal
can be extrapolated from data for surgical guid- tract begins with peripheral receptor stimula-
ance. An appreciation of the anatomy and the tion of a first-order neuron in the dorsal root
technical aspects of SSEPs is required for this ganglia. This afferent volley is sent via the ipsi-
consistency and successful intraoperative lateral posterior spinal cord to the medullary
employment. nuclei to synapse on second-order neurons.

Fig. 1.1 The dorsal column pathway. (1) Fibers enter in the medial lemniscus. Maintaining a somatotopic arrange-
the root entry zone and run upward in the dorsal columns ment, they terminate in the ventral posterolateral thala-
to the lower medulla where they terminate in the nucleus mus. (3) Third-order neurons arise in the thalamus and
gracilis and nucleus cuneatus. (2) Second-order neurons project to the parietal cortex (from Lindsay and Bone
decussate as the internal arcuate fibers and pass upward in [83]; with permission)
1 Somatosensory-Evoked Potentials 5

These second-order neurons decussate in the Venous drainage is provided by a large venous
medulla as the internal arcuate fibers and ascend network encircling the spinal cord. This network
in the medial lemniscal pathway to third-order flows to either the median posterior or anterior
neurons in the ventroposterior nuclei of the thal- spinal veins. Venous blood then flows through
amus, maintaining a somatotopic arrangement. numerous radicular veins and ultimately to the
Projections from the thalamus proceed to the azygous and pelvic venous systems [6, 7].
sensorimotor cortex, where further synapsing
occurs. Synapses are believed to be the site of
action for inhalational anesthetics; thus, the Methods
early SSEP response is minimally affected by
inhalational anesthetics. However, as the volley As mentioned, the foremost goal of SSEP moni-
ascends the dorsal column–lemniscal pathway toring should be consistency. Achieving this
and more synapses occur en route to the cortex, consistency requires manipulation of the two
cortical SSEPs are increasingly susceptible to major technical aspects of acquiring SSEPs:
the effects of inhalational anesthetics (see Chap. stimulation and recording. The following recom-
19 for more discussion of anesthesia) [1, 4, 5]. mendations are based largely on published
Perfusion to the dorsal column–lemniscal path- guidelines from “Intraoperative Monitoring
way usually comes from the posterior spinal Using Somatosensory Evoked Potentials: A
arteries in the spinal cord. The posterior spinal Position Statement by the American Society of
artery originates from the vertebral arteries and Neurophysiological Monitoring” [1].
travels bilaterally the length of the spinal cord in
the posterior lateral sulci, supplying the poste-
rior one-third of the spinal cord, including the Stimulation
posterior horns as well as the dorsal column–
lemniscal pathway [6]. The anterior spinal In order to achieve consistent intraoperative
artery, also arising from the vertebral arteries, SSEP monitoring, adequate stimulation must be
supplies the anterior and anterolateral two-thirds applied. Stimulation parameters include elec-
of the spinal cord, including the anterior horns, trode type, electrode placement, stimulus inten-
spinothalamic tracts, and corticospinal tracts. sity, stimulus duration, stimulus rate, and
However, there is a great degree of individual unilateral versus bilateral stimulation. The spe-
variability in origin of vascular supply for both cific hardware and software employed for stimu-
the posterior and anterior spinal arteries, with lation and recording exists in a variety of
each being supported by a varying number of commercially available units [1, 2, 8–10].
radicular arteries, particularly in the thoracic The first step to meaningful intraoperative
spinal cord. Chapter 40 (“Electrophysiological SSEP monitoring is stimulating appropriate
Monitoring During Thoracic Aortic Aneurysm nerves for a given operation. In general [1, 8, 9],
Surgery”) discusses blood supply of the spinal nerves chosen for intraoperative monitoring
cord in greater detail. should be below, with recording sites above, the
As the dorsal column–lemniscal pathway area at risk from surgery such that the monitored
ascends to the medullary nuclei of the brainstem, pathway travels through the neural area at risk.
perfusion comes from both the vertebral artery and For example, during corrective thoracic scoliosis
perforating branches of the basilar artery. While surgery, monitoring solely upper extremity
the somatosensory cortex maintains somatotopic SSEPs would be insufficient as the lower extrem-
arrangement, blood supply is divided into the ante- ity dorsal column tract through the spinal cord
rior and middle cerebral artery. The anterior cere- would be missed. For this example, it would be
bral artery supplies the cortex representing the useful to monitor the upper extremity SSEPs for
lower extremity while the middle cerebral artery position-related injury. The upper extremity
supplies the cortex supplying the face, head, neck, SSEPs would also provide useful information for
trunk, and upper extremity. interpreting the lower extremity SSEPs. In this
6 A. Becker et al.

case, a significant amplitude reduction through- begins by locating the ulnar groove. The cathode
out all waveforms is more likely to be related to is then placed 2 cm proximal to the elbow crease
anesthetic or physiologic parameters than if the at the ulnar groove, while the anode is placed
amplitude change occurred in just the lower 2–3 cm distal over the ulnar nerve. For these
extremity SSEPs. mixed nerves, corresponding muscle twitch (i.e.,
From a hardware standpoint, successful SSEP thumb adduction) with stimulation confirms
monitoring begins with proper electrode selec- appropriate electrode placement [1, 8–10].
tion. Stimulation electrode options include bar Lower extremity peripheral nerves commonly
electrodes, EEG metal disk electrodes, subder- used for intraoperative monitoring include the
mal needle electrodes, and adhesive surface elec- posterior tibial nerve (L4–S3) at the ankle and the
trodes. While each has advantages and peroneal nerve (L4–S2) at the head of the fibula.
disadvantages, adhesive surface electrodes are For posterior tibial nerve stimulation, the cathode
typically used intraoperatively as they are nonin- is placed between the medial malleolus and the
vasive and adhere reliably throughout the Achilles tendon, just proximal to the malleolus;
dynamic intraoperative period (including patient the anode is placed 2–3 cm distal over the poste-
position changes and patient edema). Subdermal rior tibial nerve as it courses around the medial
needle electrodes are also commonly used by malleolus. For peroneal nerve stimulation, the
many providers, especially when stimulation cathode is placed just medial to the head of the
must occur within the sterile field as they can be fibula. The anode is placed 2–3 cm distal. For
placed intraoperatively in a sterile fashion by the these mixed nerves, corresponding muscle twitch
surgeon. Subdermal needle electrodes are also (i.e., plantar toe flexion with posterior tibial nerve
recommended in cases where stimulation needs stimulation and eversion of the foot with pero-
to occur closer to the nerve (e.g., obese or edema- neal nerve stimulation) with stimulation confirms
tous patients). appropriate electrode placement [1, 8–10].
Correct placement of stimulation electrodes The electrical stimulus applied during SSEP
with respect to the nerve is also critical to ade- monitoring is a series of square-wave pulses, with
quate stimulation and subsequent stable SSEPs. durations of 0.1–0.3 ms, at a given intensity [1, 3,
Placement is dependent on both the electrode 8, 9]. When stimulating mixed sensory and motor
being used and the nerve being stimulated (e.g., nerves, the stimulus intensity is adjusted to elicit a
surface electrodes are generally placed 2–3 cm minimal twitch of the distal muscles innervated by
apart, whereas subdermal needles are placed the peripheral nerves. In purely sensory nerves,
1 cm apart) [1, 2, 8–10]. stimulation intensity two to three times the sensory
For upper extremity SSEPs, frequently used threshold is recommended [2]. Typical intraopera-
peripheral nerves include the median nerve (C5-­ tive stimulation intensity ranges from 10 to
T1) at the wrist and the ulnar nerve (C8-T1, ± C7) 50 mA. However, stimulation intensity up to
at the wrist or elbow. For median nerve stimula- 100 mA may be required intraoperatively to elicit
tion, the cathode is placed over the median nerve a reproducible, recognizable waveform, as there
2–4 cm proximal to the wrist crease, and the may be underlying pathology in addition to the
anode is placed 2–3 cm distal over the median deleterious effects of anesthetics on SSEPs [1].
nerve (Note: The cathode is the proximal elec- Possible tissue damage from repeated high
trode connected to the negative pole of the stimu- current at the stimulation sites warrants consider-
lator; the anode is the distal electrode connected ation, but the literature contains no evidence to
to the positive pole; this convention is used to support this concern as long as stimulation is
avoid a phenomenon known as anode blocking). within parameters on commercially available
For ulnar stimulation at the wrist, the cathode is instruments for SSEP monitoring [1]. Use of con-
placed 2–4 cm proximal to the wrist crease and stant current stimulation is recommended to
the anode is placed 2–3 cm distal, both over the compensate for any change in contact resistance.
ulnar nerve. Ulnar nerve stimulation at the elbow This compensation is limited by the maximum
1 Somatosensory-Evoked Potentials 7

output voltage of the stimulator. With constant between the stimulation sites and recording
current stimulation, the output of the stimulator is electrodes, usually on the shoulder [3].
current-limited when contact resistance is exces- As mentioned previously, recording sites for
sive. Most instruments designed for SSEP moni- intraoperative monitoring should be proximal to
toring have a built-in warning for this [1, 8, 9]. the surgical area at risk, with stimulation sites
The frequency of stimulation generally ranges distal. As the neural volley ascends the dorsal
from 2 to 5 Hz [1, 8–10]. To decrease noise with column–lemniscal pathway, different generators
averaging, the rate of stimulation should not be of the potential are recorded by various recording
an integer multiple of the line power supply fre- electrodes.
quency (50 or 60 Hz), the most common noise Recording electrical activity requires mea-
frequency. When excessive noise occurs, small surement of voltage between two electrode sites,
changes in the stimulus rate may improve the an active electrode and a reference electrode.
SSEP quality [1, 11]. These electrode pairs are called recording mon-
Stimulation can be unilateral or bilateral. tages, denoted by: active electrode–reference
Simultaneous bilateral stimulation can enhance electrode. In general, one cortical montage and
SSEPs, while potentially masking unilateral one subcortical montage are used to record the
changes. To effectively and simultaneously mon- ascending neural volley for intraoperative SSEPs.
itor both sides of an extremity pair, interleaved Scalp electrode locations for recording are based
unilateral (alternating left and right) stimulation on the 10–20 International System of EEG
is recommended [1]. electrode placement (Fig. 1.2). An additional
recording site, distal to the stimulation site but
proximal to the surgical site, is often used to
Recording verify peripheral conduction [1].
A recording from a given montage for a spe-
In conjunction with adequate stimulation, appro- cific stimulated peripheral nerve has a characteris-
priate recording techniques must be employed to tic waveform distribution measured in amplitude
achieve consistent intraoperative SSEP monitor- (microvolts) and latency (milliseconds). This is
ing. Recording parameters include electrode recorded on a graph of voltage (microvolts) versus
type, electrode placement (recording montage), time (milliseconds) and represents the SSEP. In
and specific equipment parameters, which general, this characteristic morphology is from
include channel availability, filters, averaging, synapses at sites along the neural pathway. These
and time base. sites are referred to as the generators of the wave-
As with stimulating electrodes, a variety of form. Waveforms are labeled “N” and “P” to rep-
recording electrodes are available, each with resent the polarity of the signal (generally negative
attendant advantages and disadvantages. For is up, positive is down, although the specific con-
intraoperative SSEP recording, subdermal nee- vention used may vary by individual) followed by
dles and metal disk electrodes are used most fre- an integer to represent the poststimulus latency of
quently. Subdermal needles are placed quickly, the wave in normal adults. For example, for corti-
though they must be secured with tape or surgical cal recording from median nerve stimulation,
staples to prevent dislodging. Metal cup elec- characteristic peaks N20 (a negative, or upward,
trodes take longer to secure and require conduc- deflection at about 20 ms) and P22 (a positive, or
tive gel or paste. Corkscrew electrodes, like downward, deflection at about 22 ms) define the
subdermal needles, are quickly placed and have amplitude of the waveform (Figs. 1.3 and 1.4).
the advantage of being fairly secure. For direct The generators of these peaks are thought to be the
cortical recording, as employed during corticog- thalamus and somatosensory cortex [1, 9].
raphy, strip or grid array electrodes are used Table 1.1 is not meant to be an exhaustive list
[1, 10, 12, 13]. A ground electrode is placed of possible recording montages for upper
8 A. Becker et al.

Fig. 1.2 10–20 International System of Electrode stamp for the indication of electrode placements in rou-
Placement. A single plane projection of the head, showing tine recording. “CP” and “FP” locations are midway
all standard positions and locations of the rolandic and between the designated “C” and “P” or “C” and “F” loca-
Sylvian fissures. The outer circle was drawn at the level of tions, respectively, “c” and “i” indicate respective loca-
the nasion and inion. The inner circle represents the tem- tions contralateral or ipsilateral to the side of stimulation,
poral line of electrodes. This diagram provides a useful respectively (from Klem et al. [84]; with permission)

Fig. 1.3 Schematic diagram of normal SSEPs to arm stimulation. Tracings are obtained from the regions identified on
the anatomic model (from Misulis and Fakhoury [2]; with permission)
1 Somatosensory-Evoked Potentials 9

Fig. 1.4 Normal posterior tibial nerve SSEPs. Traces from bottom to top show popliteal fossa potential, lumbar poten-
tial, low thoracic potential, and scalp potential (from Misulis and Fakhoury [2]; with permission)

Table 1.1 Neural generators for median and tibial nerve SEP generatorsa
Median nerve SEP generators Tibial nerve SEP generators
Generator Common Alternate Label Generator Common Alternate
Label channels used labels channels used labels
N9 Brachial plexus EPi-EPc Erb’s Popliteal Tibial nerve Popliteal
action
potential
N11 Spinal nerve root Crv-Fpz N23 Dorsal horn T12-iliac cr. Lumbar
interneurons point
N13a Dorsal horn Crv6-Fpz Cervical, P31 Medulla Crv-Fpz, Cervical
interneurons subcortical Mast-Fpz subcortical
N13b Dorsal column Crv2-Fpz Cervical, N34 Primary Cc-Fpz N37
subcortical sensory
cortex
P13 Spinomedullary Crv-Fpz, Cervical, P38 Primary Ci-Fpz, P39, P40,
junction Mast-Fpz subcortical sensory Cz′-Fpz, cortical
cortex Ci-Cc,
Cz′-Cc
P14 Lemniscal paths, Crv-Fpz, Cervical, N38 Primary Cc-Fpz
cuneate nucleus Mast-Fpz subcortical sensory
cortex
N18 Brainstem/ Ci-noncephalic
thalamic
N19 Primary sensory Cc-Fz, Cc-Ci N20,
cortex cortical
P22 Primary motor Cc-Fz, Cc-Ci
cortex
a
From Minahan and Mandir [82]; with permission
10 A. Becker et al.

extremity and lower extremity peripheral nerve other placed 2–4 cm proximal. For upper extrem-
stimulation. However, it is meant to assist with ity SSEPs this site is the ipsilateral Erb’s point
understanding evoked potentials and provide a (2 cm above the midpoint of the clavicle and at
background for intraoperative monitoring. the posterior border of the head of the sternoclei-
For upper extremity peripheral nerve stimulation, domastoid muscle) referenced to the contralateral
there are several montages commonly used for Erb’s point or a scalp electrode, often Fz [1, 3,
cortical recording. The responses recorded are 10].
most likely generated by the thalamus and After acquisition of the evoked potential,
somatosensory cortex. Since cortical responses some signal manipulation is required to distin-
are characteristically sensitive to general anes- guish the evoked potential from background
thetics, and because patients in the operating noise such as spontaneous EEG activity, ECG
room may have underlying neurologic injury, dif- activity, muscle activity, or 60 Hz noise.
ferent montages may be used to enhance cortical Amplifiers are used to increase the size of the
response amplitude. Montages include CPc–2 cm biologic signal, while filters are used to reduce
posterior to CPc (contralateral cortex to the noise. The signal is averaged over repeated
stimulus; i.e., CP3 for right arm stimulation and stimuli to increase the signal-to-noise ratio [1].
CP4 for left arm stimulation), CPc–Fz (midline Filters should be set to provide quality poten-
frontal electrode), CPc–FPz, and CPc–CPi tials with the least amount of averaging. Low
(cortex ipsilateral to the stimulus) [1, 3, 10]. frequency (high pass) filter and high frequency
For subcortical recording of upper extremity (low pass) filter settings are combined to elimi-
peripheral nerve stimulation, response generators nate components of the acquired potential out-
vary with the montage used and include the side the range of the evoked potential being
spinal cord, the cervicomedullary junction, studied. For most instruments, the standard set-
higher parts of the brainstem, and the thalamus. tings are 20 Hz for the low frequency filter and
Common montages include CPi–Erbc (Erb’s point 3000 Hz for the high frequency filter. Maintaining
contralateral to the stimulus), CvN (posterior standard settings allows a laboratory to make
spinal cervical electrode over the Nth cervical meaningful comparisons for any given patient to
spinous process, typically C6 or C7)–Fz, Fz–A1A2 laboratory normals [1].
(linked ear electrodes), Cz–A1A2, and FPz–A1A2 However, since intraoperative potentials are
[2, 3, 10]. also compared to a patient’s baseline recorded ear-
Cortical recording of stimulation of lower lier in the case, other suggested settings specific to
extremity peripheral nerves represents generators either cortical or subcortical potentials have been
of the neural volley in the somatosensory cortex. suggested. For cortical potentials, these suggested
Recording montages used include CPz–2 cm filter settings are 1–30 Hz for the low frequency
posterior to Cz, CPz–Fz, CPz–CPc, and FPz–Cz filter and 250–1000 Hz for the high frequency fil-
[2, 3, 10]. ter. For subcortical potentials, 30–100 Hz and
The generator source(s) of far-field subcorti- 1000–3000 Hz are suggested, respectively. To
cal potentials from lower extremity peripheral improve cortical SSEPs, setting the high frequency
nerve stimulation are thought to lie in the brain- filter as low as 300–500 Hz may help decrease
stem. Recording montages to acquire these artifact as the relative frequency content of cortical
potentials include CPi–A1A2, CvN–Fz, and potentials is lower than subcortical potentials. The
FPz–A1A2 [2, 3, 10]. 60-Hz rejection filter should be reserved as a last
Peripheral recording of the nerve volley distal resort to improve SSEPs as it can cause “ringing
to the stimulation site but proximal to the surgical artifact” [1, 8–10].
site can confirm the conduction of the peripheral Recorded potentials are averaged over repeated
stimulus. For lower extremity SSEPs, this site is stimuli to increase the signal-to-noise ratio.
the ipsilateral popliteal fossa—one electrode at Guidelines have suggested acquiring 500–2000
the popliteal fossa (4–6 cm above crease) and the trials per averaged response [1, 8, 9]. However,
1 Somatosensory-Evoked Potentials 11

the signal-to-noise ratio and need for prompt Inhalational Anesthetics

intraoperative reporting may dictate the number
of trials averaged. The optimal choice of montage Halogenated inhalational agents produce a dose-­
allows the largest signal-to-noise ratio, which related reduction in amplitude and increase in
minimizes the number of averages needed and the latency of SSEPs. This SSEP decrement is more
acquisition time of a response [12–15]. In addi- pronounced for cortical recordings than subcorti-
tion, in a rare patient, the somatosensory fibers are cal, spinal, or peripheral recordings. [1, 4, 17].
uncrossed such that the ipsilateral and contralat- Nitrous oxide decreases cortical SSEP ampli-
eral cortex need to be evaluated for the maximal tude and increases latency [18, 19]. This effect is
amplitude [16]. synergistic with halogenated inhalational agents
The timebase (milliseconds) for waveform dis- and most intravenous anesthetics [1, 4, 17, 19, 20].
play also needs to be appropriate for the given For example, with equipotent doses, nitrous oxide
potential. Generally, this means 50 ms for upper combined with halogenated agents produces a
extremity potentials and 100 ms for lower extrem- greater decrease in amplitude and increase in
ity potentials [1]. Also, in the presence of underly- latency of the cortical SSEP [15, 19]. As with halo-
ing abnormal neurologic function and subsequent genated agents, the effect on subcortical and
increased latency of SSEPs, the timebase may peripheral SSEPs is minimal [1, 4, 17, 19].
need to be increased to adequately acquire and
display the evoked potential.
Intravenous Anesthetics

Intraoperative Variables Affecting In general, the intravenous drug effects on SSEPs

SSEPS: Pharmacology are less than those from inhalational agents. With
and Physiology the exceptions of etomidate and ketamine, mini-
mal effects on cortical SSEPs are seen with low
In addition to the stimulation and recording doses of intravenous anesthetics. Moderate
parameters discussed earlier, pharmacologic and reduction in amplitude and increase in latency are
physiologic variables can also significantly affect seen with higher doses, again with the exceptions
the reliable recording of evoked potentials. of etomidate and ketamine. Most intravenous
Understanding how these variables influence the agents have negligible effects on subcortical SSEPs.
process is essential to successful intraoperative The following provides details for specific intra-
SSEP monitoring. venous anesthetic effects on SSEPs.
Anesthetic drugs have various effects on Barbiturates produce a short-term dose-­
SSEPs. While the mechanisms of action for spe- dependent reduction in amplitude and increase in
cific anesthetic drugs differ along with each latency of cortical SSEPs, with little effect on
drug’s effect on SSEPs (i.e., some drugs enhance subcortical and peripheral SSEPs [1, 4, 17, 21].
SSEPs, while most decrease SSEPs), all anes- Specifically, the SSEP decrement for induction
thetics share a general mechanism of action by doses of thiopental lasts less than 10 min [20–
either altering the function of synapses or axonal 23]. Infusion of methohexital as part of a total
conduction to change neuronal excitability (see intravenous general anesthetic has been shown to
Chap. 19) [4, 5]. As the number of synapses in a provide excellent conditions for SSEP monitor-
pathway increases, the effect of a given anes- ing [24]. Even at doses causing coma, barbitu-
thetic drug on the SSEP is more pronounced. rates allow the monitoring of cortical SSEPs [1,
Therefore, cortical potentials are more sensitive 4, 21, 25–28].
than subcortical, spinal, or peripheral nerve Propofol influences SSEPs in a similar
recordings to anesthetic effects [1, 4, 17]. This manner to that of barbiturates but with desirable
includes both deleterious and augmentative rapid emergence after prolonged infusion. As a
effects on SSEPs. ­one-­time induction dose, there is no change in
12 A. Becker et al.

amplitude for cortical and subcortical SSEPs spectrum of applications. Adjuvant clonidine [21]
from median nerve stimulation, but there is a and dexmedetomidine [21, 45–47] use is compat-
mild increase in cortical latency [21, 29]. Propofol ible with intraoperative SSEP monitoring.
induction and infusion causes cortical amplitude In general, systemic opioids mildly decrease
reduction with recovery after infusion termina- cortical SSEP amplitude and mildly increase
tion [4, 30]. Propofol has no effect on epidural-­ latency with minimal effect on subcortical and
evoked potentials [4, 31]. Combined with peripheral potentials [1, 4, 19, 21]. Bolus dosing
opioids, propofol produces less cortical ampli- of opioids has a greater impact on SSEP changes
tude depression than nitrous oxide or midazolam than continuous infusion [1]. Therefore, opioid
[1, 21, 32–35]. Compared to equipotent doses of infusions are an important component of anesthe-
halogenated agents [1, 3] or nitrous oxide [1, 36], sia for intraoperative SSEP monitoring.
the amplitude decrement is less with propofol. Remifentanil is often used as it has a short con-
As part of a balanced total intravenous anesthetic, text sensitive half-time and promotes rapid emer-
propofol is compatible with intraoperative moni- gence. Neuraxial opioids, excluding meperidine,
toring of SSEPs [1, 4, 21, 33, 37, 38]. have minimal or no effect on SSEPs [4, 17, 21,
Etomidate and ketamine are unique in that 48–51]. The decreased cortical amplitude and
they increase cortical SSEP amplitude. Etomidate increased cortical latency seen with subarachnoid
produces a marked increase in cortical amplitude meperidine [21, 48] is likely secondary to its
and a mild increase in cortical latency [1, 4, local anesthetic-like qualities. Neuraxial opioid-­
17–22, 37]. Etomidate’s effects on subcortical only techniques can augment analgesia without
amplitude vary from no change to mild reduction affecting intraoperative SSEP monitoring.
[1, 4, 17, 20–22, 37, 38]. Despite this potential Benzodiazepines have mild depressant effects
for subcortical SSEP amplitude reduction and on cortical SSEPs [1, 4, 21]. In the absence of
variable peak specific effects on latency, infusion other agents, midazolam causes mild to no
of etomidate as part of a total intravenous gen- depression of cortical SSEPs, a moderate increase
eral anesthetic has been used to improve cortical in N20 latency, and minimal to no effects on sub-
SSEPs [4, 39, 40], even when intraoperative cortical and peripheral potentials [1, 4, 20, 52].
monitoring was otherwise unobtainable [4, 39]. Used as an intermittent bolus or continuous infu-
Etomidate has the drawback of adrenal sion (50–90 μg/kg/h) to promote intraoperative
suppression. SSEP monitoring [1], midazolam is useful to pro-
Ketamine increases cortical SSEP amplitudes mote amnesia with TIVA and to ameliorate hal-
with no change in cortical latency or subcortical lucinations with ketamine [17].
potentials [1, 4, 21, 41, 42]. The addition of Droperidol, a sedative-hypnotic used in
nitrous oxide [4, 41] or enflurane 1.0 MAC [4, neuroanesthesia, has minimal effects on SSEPs
43] to a ketamine anesthetic decreases SSEP [1, 4, 17]. Concern for QT prolongation is a
amplitude by approximately 50 %. However, ket- consideration.
amine has been used successfully as part of a bal- Neuromuscular blocking agents commonly
anced anesthetic with midazolam and nitrous used during general anesthesia do not directly
oxide for intraoperative SSEP monitoring during affect SSEPs. However, by decreasing electro-
spine surgery [21, 44] and is an acceptable com- myographic artifact and/or interference from
ponent of total intravenous anesthesia (TIVA) for muscle groups near recording electrodes, neuro-
SSEPs [1, 3]. Drawbacks to ketamine include muscular blockers may increase the signal-to-­
hallucinations, long half-life with subsequent noise ratio and improve the quality of SSEP
prolonged emergence, sympathomimetic effects, waveforms [4, 21, 53].
and increased intracranial pressure in the setting Perioperative infusion of systemic lidocaine is
of intracranial pathology. used to decrease postoperative pain. Infusion of
The alpha-2 agonists clonidine and dexme- relatively high-dose lidocaine has been shown to
detomidine are anesthetic agents with a broad decrease cortical SSEP amplitude and increase
1 Somatosensory-Evoked Potentials 13

latency [54], while lower infusion rates have Similar to core temperature, local temperature
been shown to have no effect [55]. changes at anatomic sites can affect SSEPs. For
Summarizing pharmacologic effects, intrave- example, temperature changes at the surgical site
nous anesthetic agents are more compatible with from surgical exposure or cold irrigation in the
intraoperative monitoring of SSEPs than inhala- surgical field can affect SSEPs. Also, stimulating
tional agents. While inhalational agents have an extremity exposed to cold intraoperative tem-
been used in low dose combined with other intra- peratures, with or without cold intravenous fluid
venous agents, TIVA is preferred for consistent infusing, may affect SSEPs [4].
intraoperative SSEP monitoring in patients with
small-amplitude SSEPs. Also, motor-evoked
potentials (MEPs) are frequently paired with Tissue Perfusion
intraoperative SSEP monitoring and are
extremely sensitive to inhalational agents, often Changes in blood pressure can affect tissue per-
requiring TIVA. TIVA can be any combination of fusion and thus can affect SSEPs. If perfusion is
intravenous drugs for end-effects of hypnosis, insufficient to meet basic metabolic demands of
amnesia, analgesia, optimal surgical conditions the tissue, cortical SSEP amplitude begins to
(i.e., an immobile patient), and quick metabolism diminish. With normothermia, this occurs when
for an immediate postoperative neurologic exam- cerebral perfusion decreases to about 18 cm3/
ination. A typical infusion combination is propo- min/100 g of tissue [1, 4, 17, 61–63]. Further
fol and remifentanil with intermittent midazolam, reductions in perfusion below approximately
with or without muscle relaxant. However, as 15 cm3/min/100 g of tissue can cause loss of cor-
mentioned previously, various other hypnotic and tical SSEPs [1, 4, 51, 53, 61–63]. Subcortical
opioid drugs may be used. To help ensure amne- responses are less sensitive to reductions in tissue
sia, a monitor of anesthetic depth may be useful perfusion.
(see Chap. 19 for additional information about Regional ischemia, with or without any
anesthesia considerations). degree of systemic hypotension, can be caused
The physiologic milieu of an intraoperative by local factors that can affect SSEPs. Examples
patient is very dynamic and can affect SSEP include spinal distraction, surgical retractor-
amplitude and/or latency. induced ischemia, position ischemia, tourniquet-
induced ischemia, ischemia from vascular injury,
and vascular clips (either temporary or permanent)
Temperature [4, 64–66].
Oxygen delivery is affected by changes in
Changes in body temperature affect SSEPs. Mild hematocrit, which alters oxygen-carrying capac-
hypothermia increases cortical SSEP latency but ity and blood viscosity. Primate data reveal that
has little effect on cortical amplitude and subcor- in general, mild anemia produces an increase in
tical or peripheral responses [1,]. Mild hypother- SSEP amplitude. Primate data also reveal that
mia (down to 32 °C) may even be associated with reductions in hematocrit beyond mild anemia
increased cortical amplitudes [56–58]. With pro- cause further SSEP amplitude reduction and
found hypothermia, cortical SSEPs disappear. increase in SSEP latency [4, 21, 67, 68].
Subcortical, spinal, and peripheral responses may
remain with increased latency, but they also dis-
appear at lower temperatures [1, 59]. Rewarming Oxygenation/Ventilation
improves the latencies but not in the reverse tra-
jectory as cooling [1, 21]. Mild hyperthermia Variations in both oxygen and carbon dioxide
(39 °C) is associated with a decrease in cortical levels can affect SSEPs. Mild hypoxemia does
and subcortical latencies with no change in not affect SSEPs [4, 69]. A decrease in SSEP
amplitudes [21, 60]. amplitude was reported as a manifestation of
14 A. Becker et al.

intraoperative hypoxemia [70]. Up to a PaCO2 of increase of 10 % or more, not attributable to

50 mmHg, hypercarbia has no effect on human anesthetic or physiologic causes, are considered
SSEPs [21, 71]. Cortical amplitude augmentation significant changes warranting intervention
and a mild decrease in latency occur with hyper- [1, 21, 76, 77]. The validity of these alarm crite-
ventilation in awake volunteers [21, 69]. However, ria has been studied [1, 78, 79].
in isoflurane-anesthetized patients, hypocapnia to
20–25 mmHg caused no change in amplitude and
a mild decrease in latency [21, 72]. Intraoperative Applications
for SSEPs

Intracranial Pressure Intraoperative SSEPs are employed for a wide

range of surgeries. The common goal is to ensure
Increased intracranial pressure decreases ampli- maintenance of neurologic integrity throughout a
tude and increases latency of cortical SSEPs procedure with resultant improved outcome and
[4, 59]. As intracranial pressure increases, there decreased morbidity. Nerve root function can be
is pressure-related cortical SSEP decrements and monitored with SSEPs intraoperatively, although
concurrent loss of subcortical responses with SSEPs may be insensitive to changes in single
uncal herniation [4]. nerve root function (see also Dermatomal-­
Evoked Potentials). Peripheral nerves and bra-
chial plexus monitoring can be used for surgical
Other Physiologic Variables guidance as well as for avoidance of position-­
related neuropraxia during surgeries such as total
A multitude of other physiologic factors may hip arthroplasty and shoulder arthroscopy. Spinal
affect SSEPs, including fluctuations in electro- cord function can be monitored during spine
lytes and glucose, total blood volume, and central fusions, spinal cord tumor removal, arteriove-
venous pressure [4]. nous malformation repair, and abdominal and
thoracic aortic aneurysm repair. The brain stem
and cortical structures can be monitored during
 riteria for Intervention
C tumor resection, carotid endarterectomy, and
During Intraoperative SSEP cerebral aneurysm clipping. Also, SSEPs can be
Monitoring employed to localize the border of the motor cor-
tex intraoperatively [2] (see Chap. 9, “Brain and
Reproducible, recognizable baseline waveforms Spinal Cord Mapping”).
are the foundation of successful intraoperative
SSEP monitoring. It is from these baselines that
intraoperative changes are based. The dynamic Dermatomal-Evoked Potentials
intraoperative milieu, including surgical and
anesthetic influences, can make the process of Evoked potentials elicited by stimulating indi-
SSEP monitoring challenging and complicate vidual dermatomes are called dermatomal SSEPs
the interpretation of the significance of changes (DSSEPs). Surface electrodes are used to stimu-
from baseline. Providing evidence-based alarm late a single dermatome mediated by a unique
criteria for intraoperative changes in amplitude nerve root. Dermatome maps to guide optimal
and latency is difficult. Intraoperative SSEP placement of surface electrodes exist [1, 80, 81].
changes of 45–50 % amplitude reduction and In contrast to SSEPS where supramaximal stimu-
7–10 % latency increases can occur without lation intensities should be used to provide repro-
changes in postoperative neurologic function ducible and reliable evoked responses, high
[21, 73–75]. However, empirically, an amplitude stimulation intensities for DSSEPs can cause cur-
reduction of 50 % or greater and/or a latency rent spread and elicit responses from adjacent
1 Somatosensory-Evoked Potentials 15

dermatomes. Also, stimulus intensity can affect evoked potentials. J Clin Neurophysiol. 1987;4:
397–416.
DSSEP latencies [1, 80]. Therefore, minimally
9. American Electroencephalographic Society. Guidelines
effective stimulation intensities need to be used for intraoperative monitoring of sensory evoked poten-
for DSSEPs. Recording parameters are the same tials. J Clin Neurophysiol. 1994;11:77–87.
for DSSEPs as SSEPs. Cortical responses are 10. International Organization of Societies for
Electrophysiological Technology (OSET). Guidelines
typically larger in amplitude than subcortical
for performing EEG and evoked potential monitoring
responses. Because DSSEPs are sensitive to during surgery. Am J END Technol. 1999;39:
nerve root compression and mechanical stimula- 257–77.
tion [1, 81], intraoperative employment of 11. Stecker MM. Generalized averaging and noise levels
in evoked responses. Comput Biol Med. 2000;
DSSEPs includes the following: pedicle screw
30:247–65.
placement, cauda equina tumor resection, teth- 12. Celesia GG. Somatosensory evoked potentials
ered cord release, and surgical treatment of spina recorded directly from human thalamus and Sm I cor-
bifida. However, due to dermatomal overlap and tical area. Arch Neurol. 1979;36:399–405.
13. Kelly Jr DL, Goldring S, O’Leary JL. Averaged
variability, along with side-to-side relative stimu-
evoked somatosensory responses from exposed cortex
lation intensity, usefulness of DSSEPs can be of man. Arch Neurol. 1965;13:1–9.
compromised [1, 80]. In addition, there are other 14. MacDonald DB, Al Zayed Z, Stigsby B. Tibial
limitations to the intraoperative employment that somatosensory evoked potential intraoperative moni-
toring: recommendations based on signal to noise
make DSSEPs controversial for assessing spinal
ratio analysis of popliteal fossa, optimized P37, stan-
nerve root function. Specifically, a misplaced dard P37, and P31 potentials. Clin Neurophysiol.
pedicle screw is detected only when there is con- 2005;116(8):1858–69.
tact with the nerve root monitored [1, 80]. Use of 15. MacDonald DB, Al-Zayed Z, Stigsby B, Al-Homoud
I. Median somatosensory evoked potential intraop-
DSSEPs is further limited by their exquisite sen-
erative monitoring: recommendations based on
sitivity to anesthetics [81]. signal-to-­noise ratio analysis. J Clin Neurophysiol.
2009;120(2):315–28.
16. MacDonald DB, Streletz LJ, Al-Zayed Z, Abdool S,
Stigsby B. Intraoperative neurophysiologic discovery
References of uncrossed sensory and motor pathways in a patient
with horizontal gaze palsy and scoliosis. Clin
1. Tolekis JR. Intraoperative monitoring using somato- Neurophysiol. 2004;115(3):576–82.
sensory evoked potentials: a position statement by the 17. Sloan T. Evoked potentials. In: Albin MS, editor. A
American Society of Neurophysiological Monitoring. textbook of neuroanesthesia with neurosurgical and
J Clin Monit Comput. 2005;19:241–58. neuroscience perspectives. New York, NY: McGraw-­
2. Misulis KE, Fakhoury T. Spehlmann’s evoked poten- Hill; 1997. p. 221–76.
tial primer. 3rd ed. Woburn, MA: Butterworth-­ 18. Sloan TB, Koht A. Depression of cortical somatosen-
Heinemann; 2001. sory evoked potentials by nitrous oxide. Br J Anaesth.
3. Cruccu G, Aminoff MJ, Curio G, et al. Recommen­ 1985;57:849–52.
dations for the clinical use of somatosensory-evoked 19. Sloan TB. Anesthetic effects on electrophysiologic
potentials. Clin Neurophysiol. 2008;119:1705–19. recordings. J Clin Neurophysiol. 1998;15:217–26.
4. Sloan TB, Heyer EJ. Anesthesia for intraoperative 20. Koht A, Schutz W, Schmidt G, Schramm J, Watanabe
neurophysiologic monitoring of the spinal cord. J Clin E. Effects of etomidate, midazolam, and thiopental on
Neurophysiol. 2002;19(5):430–43. median nerve somatosensory evoked potentials and
5. Sloan T. Anesthetics and the brain. Anesthesiol Clin the additive effects of fentanyl and nitrous oxide.
North Am. 2002;20:1–27. Anesth Analg. 1988;67:435–41.
6. Mullen M, McGarvey M. Spinal cord infarction: vas- 21. Banoub M, Tetzlaff JE, Schubert A. Pharmacologic
cular anatomy and etiologies. In: Wilterdink J, editor. and physiologic influences affecting sensory evoked
Waltham, MA: UpToDate; 2015. http://www.uptodate. potentials: implications for perioperative monitoring.
com/contents/spinal-cord-infarction-vascular- Anesthesiology. 2003;99:716–37.
anatomy-­and-etiologies. Accessed 19 May 2015. 22. McPherson RW, Sell B, Thaystman RJ. Effect of thio-
7. Cheshire WP, Santos CC, Massey EW, Howard Jr JF. pental, fentanyl and etomidate on upper extremity
Spinal cord infarction: etiology and outcome. somatosensory evoked potentials in humans.
Neurology. 1996;47(2):321. Anesthesiology. 1986;65:584–9.
8. American Electroencephalographic Society. 23. Ikuta T. Effects of thiopental on the human somato-
Guidelines for intraoperative monitoring of sensory sensory evoked response. Folia Psychiatr Neurol Jpn.
1966;20:19–31.
16 A. Becker et al.

24. Sloan TB, Vasquez J, Burger E. Methohexital in total motor evoked potentials after high frequency repeti-
intravenous anesthesia during intraoperative neuro- tive electrical stimulation. Electroencephalogr Clin
physiological monitoring. J Clin Monit Comput. Neurophysiol. 1998;108:175–81.
2013;27:697–702. 39. Sloan TB, Ronai AK, Toleikis JR, et al. Improvement
25. Ganes T, Lundar T. The effect of thiopentone on of intraoperative somatosensory evoked potentials by
somatosenory evoked responses and EEGs in etomidate. Anesth Analg. 1988;67:582–5.
comatose patients. J Neurol Neurosurg Psychiatry. 40. Meng XL, Wang LW, Zhao W, Guo XY. Effects of
1983;46:509–14. different etomidate doses on intraoperative
26. Drummond JC, Todd MM, U HS. The effect of high somatosensory-­evoked potential monitoring. Ir J Med
dose sodium thiopental on brainstem auditory and Sci. 2015;184(4):799–803.
median nerve somatosensory evoked responses in 41. Schubert A, Licina MG, Lineberry PJ. The effect of
humans. Anesthesiology. 1985;63:249–54. ketamine on human somatosensory evoked potentials
27. Sutton LN, Frewen T, Marsh R, Jaggi J, Bruce DA. and its modification by nitrous oxide. Anesthesiology.
The effects of deep barbiturate coma on multimodal- 1990;72:33–9.
ity evoked potentials. J Neurosurg. 1982;57:178–85. 42. Kano T, Shimoji K. The effects of ketzmine and neu-
28. Drummond JC, Todd MM, Schubert A, Sang H. roleptanalgesia on the evoked electrospinogram and
Effect of acute administration of high dose pentobar- elecromyogram in man. Anesthesiology.
bital on human brainstem auditory and median nerve 1974;40:241–6.
somatosensory evoked responses. Neurosurgery. 43. Stone JL, Ghaly RF, Levy WJ, Kartha R, Krinsky L,
1987;20:830–5. Roccaforte P. A comparative analysis of enflurane anes-
29. Scheepstra GL, deLange JJ, Booij LH, Ross thesia on primate motor and somatosensory evoked
HH. Median nerve evoked potentials during propofol potentials. Electroencephalgr Clin Neurophysiol.
anesthesia. Br J Anaesth. 1989;62:92–4. 1992;84:180–7.
30. Kalkman CJ, Drummond JC, Ribberink AA. Effects 44. Langeron O, Lille F, Zerhouni O, et al. Comparison of
of propofol, etomidate, midazolam, and fentanyl on the effects of ketamine-midazolam with those of
motor evoked responses to transcranial electrical or fentanyl-­ midazolam on cortical somatosensory
magnetic stimulation in humans. Anesthesiology. evoked potentials during major spine surgery. Br
1992;76:502–9. J Anaesth. 1997;78:701–6.
31. Angel A, LeBeau F. A comparison of the effects of 45. Bloom M, Beric A, Bekker A. Dexmedetomidine
propofol with other anesthetic agents on the centripe- infusion and somatosensory evoked potentials.
tal transmission of sensory information. Gen J Neurosurg Anesthesiol. 2001;13:320–2.
Pharmacol. 1992;23:945–63. 46. Tobias JD, Goble TJ, Bates G, Anderson JT,
32. Schwartz DM, Schwartz JA, Pratt Jr RE, Wierzbowski Hoernschemeyer DG. Effects of dexmedetomidine on
LR, Sestokas AK. Influence of nitrous oxide on poste- intraoperative motor and somatosensory evoked
rior tibial nerve cortical somatosensory evoked poten- potential monitoring during spinal surgery in adoles-
tials. J Spine Disord. 1997;10:80–6. cents. Paediatr Anaesth. 2008;18(11):1082–8.
33. Borrissov B, Langeron O, Lille F, et al. Combination 47. Chen Z, Lin S, Shao W. Effects on somatosensory and
of propofol-sufentanil on somatosensory evoked motor evoked potentials of senile patients using
potentials in surgery of the spine. Ann Francaises d ­different doses of dexmedetomidine during spine sur-
Anesth et de Reanimation. 1995;14:326–30. gery. Ir J Med Sci. 2015;184(4):813–8.
34. Kalkman CJ, Traast H, Zuurmond WW, Bovill 48. Fernandez-Galinski SM, Monells J, Espadaler JM,
JG. Differential effects of propofol and nitrous oxide Pol O, Puig MM. Effects of subarachnoid lidocaine,
on posterior tibial nerve somatosensory cortical meperidine and fentanyl on somatosensory and motor
evoked potentials during alfentanil anaesthesia. Br evoked responses in awake humans. Acta Anaesthesiol
J Anaesth. 1991;66:483–9. Scandinavica. 1996;40:39–46.
35. Laureau E, Marciniak B, Hèbrard A, Herbaux B, 49. Goodarzi M, Shier NG, Grogan DP. Effect of intra-
Guieu JD. Comparative study of propofol and mid- thecal opioids on somatosensory-evoked potentials
azolam effects on somatosensory evoked potentials during spinal fusion in children. Spine. 1996;21:
during surgical treatment of scoliosis. Neurosurgery. 1565–8.
1999;45:69–74. 50. Schubert A, Licina MG, Lineberry PJ, Deers MA. The
36. Boisseau N, Madany M, Staccini P, et al. Comparison effect of intrathecal morphine on somatosensory
of the effects of sevoflurane and propofol on cortical evoked potentials in awake humans. Anesthesiology.
somatosensory evoked potentials. Br J Anaesth. 1991;75:401–5.
2002;88:785–9. 51. Loughman BA, Yau KW, Ransford AO, Hall
37. Kochs E, Treede RD, Schulte am Esch J. Increase of GM. Effects of epidural diamorphine on the somato-
somatosensory evoked potentials during induction of sensory evoked potentials to posterior tibial nerve
anesthesia with etomidate. Anaesthetist. 1986; stimulation. Anesthesia. 1991;46:912–4.
35:359–64. 52. Sloan TB, Fugina ML, Toleikis JR. Effects of mid-
38. Pechstein U, Nadstawek J, Zentner J, et al. Isoflurane azolam on median nerve somatosensory evoked
plus nitrous oxide versus propofol for recording of potentials. Br J Anaesth. 1990;64:590–3.
1 Somatosensory-Evoked Potentials 17

53. Sloan TB. Nondepolarizing neuromuscular blockade 70. Grundy BL, Heros RC, Tung AS, Doyle
does not alter sensory evoked potentials. J Clin Monit. E. Intraoperative hypoxia detected by evoked poten-
1994;10:4–10. tial monitoring. Anesth Analg. 1981;60:437–9.
54. Schubert A, Licina MG, Glaze GM, Paranandi 71. Kalkman CJ, Boezeman EH, Ribberink AA, Oosting
L. Systemic lidocaine and human somatosensory-­ J, Deen L, Bovill JG. Influence of changes in arterial
evoked potentials during sufentanil-isoflurane anaes- carbon dioxide tension on the electroencephalogram
thesia. Can J Anaesth. 1992;39(6):569–75. and posterior tibial nerve somatosensory cortical
55. Sloan TB, Mongan P, Lyda C, Koht A. Lidocaine infu- evoked potentials during alfentanil/nitrous oxide
sion adjunct to total intravenous anesthesia reduces anesthesia. Anesthesiology. 1991;75:68–74.
the total dose of propofol during intraoperative neuro- 72. Schubert A, Drummond JC. The effect of acute hypo-
physiological monitoring. J Clin Monit Comput. capnia on human median nerve somatosensory evoked
2014;28:139–47. responses. Anesth Analg. 1986;65:240–4.
56. Nuwer MR. Evoked potential monitoring in the oper- 73. Mackey-Hargadine JR, Hall III JW. Sensory evoked
ating room. New York: Raven; 1986. responses in head injury. Central Nerv Syst Trauma.
57. Lang M, Welte M, Syben R, Hansen D. Effects of 1985;2:187–206.
hypothermia on median nerve somatosensory evoked 74. LaMont RL, Wasson SI, Green MA. Spinal cord
potentials during spontaneous circulation. J Neurosurg monitoring during spinal surgery using somatosen-
Anesthesiol. 2002;14(2):141–5. sory spinal evoked potentials. J Pediatr Orthop.
58. Zanatta P, Bosco E, Comin A, Mazzarolo AP, Di 1983;3:31–6.
Pasquale P, Forti A, Longatti P, Polesel E, Stecker M, 75. Lubicky JP, Spadaro JA, Yuan HA, Fredrickson BE,
Sorbara C. Effect of mild hypothermic cardiopulmo- Henderson N. Variability of somatosensory cortical
nary bypass on the amplitude of somatosensory evoked evoked potential monitoring during spinal surgery.
potentials. J Neurosurg Anesthesiol. 2014;26(2): Spine. 1989;14:790–8.
161–6. 76. York DH, Chabot RJ, Gaines RW. Response variabil-
59. Stecker MM, Cheung AT, Pochettino A, et al. Deep ity of somatosensory evoked potentials during scolio-
hypothermic circulator arrest: I effects of cooling on sis surgery. Spine. 1987;12:864–76.
electroencephalogram and evoked potentials. Ann 77. Brown RH, Nash CL, Berilla JA, Amaddio MD.
Thorac Surg. 2001;71(1):22–8. Cortical evoked potential monitoring. A system for
60. Oro J, Haghighi SS. Effects of altering core body tem- intraoperative monitoring of spinal cord function.
perature on somatosensory and motor evoked poten- Spine. 1984;9:256–61.
tials in rats. Spine. 1992;17:498–503. 78. More RC, Nuwer MR, Dawson EG. Cortical evoked
61. Branston NM, Symon L, Cortical EP. Blood flow, and potential monitoring during spinal surgery: sensitivity,
potassium changes in experimental ischemia. In: specificity, reliability, and criteria for alarm. J Spinal
Barber C, editor. Evoked potentials. Baltimore, MD: Disord. 1988;1(1):75–80.
University Park Press; 1980. p. 527–30. 79. Wiedemayer H, Fauser B, Sandalcioglu IE, Schafer
62. Nuwer MR. Intraoperative electroencephalography. H, Stolke D. The impact of neurophysiological
J Clin Neurophysiol. 1993;10:437–44. intraoperative monitoring on surgical decisions: a
­
63. Prior PF. EEG monitoring and evoked potentials in critical analysis of 423 cases. J Neurosurg.
brain ischemia. Br J Anaeth. 1985;57:63–81. 2002;96:255–62.
64. Brodkey JS, Richards DE, Blasingame JP, et al. 80. Owen JH, Toleikis JR. Nerve root monitoring. In:
Reversible spinal cord trauma in cats: additive effects Bridwell KH, Dewald RD, editors. The textbook of
of direct pressure and ischemia. J Neurosurg. spinal surgery. 2nd ed. Philadelphia, PA: Lippincott-­
1972;37:591–3. Raven; 1997. p. 61–75.
65. Dolan EJ, Transfeld EE, Tator CH, et al. The effect of 81. Toleikis JR, Carlvin AO, Shapiro DE, Schafer
spinal distraction on regional blood flow in cats. MF. The use of dermatomal evoked responses during
J Neurosurg. 1980;53:756–64. surgical procedures that use intrapedicular fixation of
66. Gregory PC, McGeorge AP, Fitch W, et al. Effects of the lumbosacral spine. Spine. 1993;18:2401–7.
hemorrhagic hypotension on the cerebral circulation. II. 82. Minahan RE, Mandir AS. Basic neurophysiologic
Electrocortical function. Stroke. 1979;10:719–23. intraoperative monitoring techniques. In: Husain AM,
67. Nagao S, Roccaforte P, Moody RA. The effects of editor. A practical approach to neurophysiologic
isovolemic hemodilution and reinfusion of packed intraoperative monitoring. New York: Demos; 2008.
erythrocytes on somatosensory and visual evoked p. 21–44.
potentials. J Surg Res. 1978;25:530–7. 83. Lindsay K, Bone I. Neurology and neurosurgery illus-
68. Dong WK, Bledsoe SW, Chadwick HS, Shaw CM, trated. London, UK: Churchill Livingstone; 2004.
Hornbein TF. Electrical correlates of brain injury p. 198.
resulting from severe hypotension and hemodilution 84. Klem GH, Lüders HO, Jasper HH, Elger C. The ten-­
in monkeys. Anesthesiology. 1986;65:617–25. twenty electrode system of the International
69. Ledsome JR, Cole C, Sharp-Kehl JM. Somatosensory Federation. The International Federation of Clinical
evoked potentials during hypoxia and hypocapnia in Neurophysiology. Electroencephalogr Clin
conscious humans. Can J Anasth. 1996;43:1025–9. Neurophysiol Suppl. 1999;52:3–6.
18 A. Becker et al.

Questions d. In general, the effect of anesthetics on cor-

1. Which of the following statements regarding tical SSEPs is more pronounced than
the dorsal column pathway is incorrect? effects on subcortical SSEPs.
a. The dorsal column pathway is also referred 3. What is the empirically accepted threshold of
to as the dorsal column-lemniscal SSEP latency increase that warrants intervention?
pathway. a. 25 %
b. The dorsal column pathway mediates b. 35 %
mechanoreception and proprioception. c. 10 %
c. The dorsal column pathway does not d. 50 %
decussate in the medulla. 4. What is the empirically accepted threshold of
d. Perfusion of the dorsal column pathway is SSEP amplitude reduction that warrants
typically from the posterior spinal artery. intervention?
2. Which of the following is false regarding a. 25 %
pharmacologic effects on SSEPs? b. 35 %
a. Nitrous oxide, when combined with vola- c. 10 %
tile anesthetic, reduces SSEPs more than d. 50 %
either agent by itself.
b. The effects of bolus narcotics on SSEPs Answers
are less pronounced than those of continu- 1. c
ous infusions of narcotics. 2. b
c. Ketamine and etomidate are unique in that 3. c
they may be beneficial to SSEPs. 4. d
 Transcranial Motor-Evoked
Potentials
2
Leslie C. Jameson

Key Learning Points

Introduction
• The motor-evoked potential response (MEP)
is an indirect complex polyphasic muscle Motor-evoked potentials (MEPs) continue to be
response that requires a coordinated response the most recent addition to routine intraoperative
of the motor neuron pathway and the muscle. neurophysiologic monitoring (IOM). The impor-
• Due to the motor pathway’s blood supply, the tance of MEPs continues to expand primarily due
MEP is more vulnerable to and a better indica- to the ability to isolate perfusion-related neuro-
tor of adequacy of perfusion, particularly spi- logic function in the spinal cord. Initial reports of
nal cord perfusion. improved patient outcomes obtained with the use
• In addition to age, the ability to obtain MEP of somatosensory-evoked potential (SSEP) moni-
responses is impaired by pre-existing medical toring, primarily during scoliosis procedures in
conditions (e.g., diabetes, hypertension, chronic children and young adults, were quickly followed
spinal cord compression, spinal stenosis, nerve by case reports of isolated postoperative motor
root injury, chronic hypoperfusion, brain injury, injury without SSEP or postoperative sensory
and genetic neuromuscular disease). changes. This reflected the reality of the anatomy
• MEPs are vulnerable to hypoperfusion and and physiology of motor/sensory pathways in the
drug effects. Thus, the anesthesia caregiver is brain and spinal cord [1]. MEP and SSEP path-
responsible for selecting an appropriate tech- ways are located in different topographic and vas-
nique and maintaining adequate perfusion cular regions of the cerebral cortex, brainstem,
through maintenance of hemoglobin, blood and spinal cord. MEP pathways are very complex
pressure, and cardiac output. and include the standard voluntary pyramidal and
• MEP change, loss or loss and recovery, has extrapyramidal networks. The more complex
been shown to be a reliable predictor of imme- extrapyramidal network establishes additional
diate and long-term postoperative neurologic motor connections including those to the cerebel-
function. lum [2]. This complex and multiple synaptic
architecture makes motor pathways more sensi-
tive to ischemic insults than SSEP pathways [3].
L.C. Jameson, M.D. (*) Rare isolated motor injury without sensory
Department of Anesthesiology, School of Medicine,
changes after idiopathic scoliosis procedures
University of Colorado, 12401 E 17th Place,
Room 747, Aurora, CO 80045, USA was not the only driving force behind the wide-
e-mail: [email protected] spread adoption of MEP monitoring. Increasing

© Springer International Publishing AG 2017 19

A. Koht et al. (eds.), Monitoring the Nervous System for Anesthesiologists
and Other Health Care Professionals, DOI 10.1007/978-3-319-46542-5_2
20 L.C. Jameson

surgical volume and operative complexity in the the posterior spinal arteries (PSAs). Spinal cord
central nervous system (CNS, spinal cord) and motor tracts are primarily supplied by the ASA,
spine also fueled the need to independently a vascular network that supplies the metaboli-
assess motor function. cally active anterior two-thirds to four-fifths of
MEPs facilitate better intraoperative decision- the spinal cord including the gray matter and
making in all patient groups. As surgical techniques anterior horn cells, all of which are more sensi-
(instrumentation, diagnostic imaging, and intraop- tive to ischemia [3, 22].
erative imaging) advanced and perioperative anes- Both ASA and PSAs arise as branches of the
thetic management options improved, many vertebral arteries in the brainstem and then
patients who were at high anesthetic, surgical, and descend along the spinal cord providing perfora-
medical risk underwent new extensive surgical pro- tors into the spinal cord. The ASA receives blood
cedures. This increased risk of permanent and dev- radicular arteries, which originate in the aorta
astating neurologic complications. MEP monitoring [23]. Typically, there are three cervical and two
became a favored method to help prevent complex thoracic arteries located at T2, 3 and T7-L4, with
surgical intervention from exceeding safe limits the Artery of Adamkiewicz (AA) providing about
where the risk of the potential surgical adverse 75 % of the blood supply to the anterior cord [3,
event exceeds possible functional gain [4]. New 24]. The reduced number of radicular arteries, the
information suggests MEP monitoring, particularly increased distance traversed, and increased meta-
in spine surgery, has a better correlation with good bolic demand make areas of the spinal cord per-
postoperative motor outcome than the use of SSEPs, fused by the ASA more susceptible to
and many experts advocate MEP monitoring for: hypoperfusion. While axons are quite resistant to
ischemia, the anterior cord contains many more
• Surgical correction of all axial skeletal cells and synapses, which explains the rapid
deformities with instrumentation [5–8] changes seen in MEPs when inadequate perfusion
• Intramedullary spinal cord tumors [9–12] occurs. Disruption of blood flow through these
• Intracranial tumors [13–15] vessels due to mechanical or pressure changes
• CNS and spinal cord vascular lesions [16, 17] rapidly leads to deterioration of MEPs and is
• Seizure disorders [18] used to prompt a change in management (e.g.,
improvement in systemic perfusion, cerebrospi-
MEP use continues to expand outside the area nal fluid drainage) [24–26].
of neurosurgical and axial skeletal procedures to The intracranial blood supply to motor areas is
vascular procedures that put perfusion of the also vulnerable. Perforator arteries and lenticulo-
brain or spinal cord at risk like thoracoabdominal striate arteries supply the motor cortex and inter-
aneurysms, aortic arch procedures (both endo- nal capsule; they arise from the middle cerebral
vascular and open procedures) (see Chaps 39 artery. These vessels transverse a significant
and 40), and preemptive assessment of outcome distance and are vulnerable to hypoperfusion
­
in stroke [19–21]. with a decrease in cerebral perfusion pressure
(CPP) from an increase in intracranial pressure
(ICP) or cerebrospinal fluid pressure (CSFP)
Motor Pathway Blood Supply (CPP = MAP−[ICP or CSFP]) or disruption of
the source vessels (e.g., aneurysm or atrial-­
To understand why MEPs provide essential venous malformations (AVM)) or hypotension.
information for surgical procedures where neu- The distance and caliber of these vessels creates
ral tissue perfusion is at risk, it is necessary to a watershed area making motor function more
review the blood supply of the spinal cord and vulnerable to hypoperfusion than the ascending
understand the relationship between ischemia, sensory tracts [27, 28]. The normal spinal cord
electrophysiology, and infarction. A detailed dis- and brain will autoregulate blood flow to main-
cussion is found in Chap. 40. The spinal cord is tain normal perfusion. Autoregulation occurs
supplied by the anterior spinal artery (ASA) and with a CPP approximately between 50 and
2 Transcranial Motor-Evoked Potentials 21

150 mmHg; specific individuals with long-term tial, CMAP) or a wave propagation along the cor-
high (systemic hypertension) or low (infant) BP ticospinal tract (Direct wave or D wave) (Fig. 2.1).
can be outside these limits. If the perfusion pres- In humans, the exact structural connections that
sure falls below this range, autoregulation is lost are activated by evoked potential stimulation have
and spinal cord blood flow is directly dependent not been clearly defined. Structures involved in
on perfusion pressure. Hypoperfusion, as evident voluntary motor activity in animal models have
by a change in evoked potential activity, can also been described. Recordings from deep brain (DB)
be caused by reductions in oxygen delivery (e.g., electrodes used for stimulation and recording in
anemia, hypovolemia). MEP monitoring pro- patients being treated for epilepsy or movement
vides unique information about the functional disorders have led to better definition of motor
status of the anterior spinal cord and internal transmission and its interaction with sensory func-
capsule (see Chap. 21). tion [29]. Use of magnetic stimulation, the only
technique available for eliciting MEPs from awake
humans, has allowed simulation to occur with
 echnical Aspects of MEP
T simultaneous recordings of electroencephalogra-
Monitoring phy (EEG) and electromyography (EMG) as
paired responses. These data suggest that the vari-
MEPs are elicited by transcranial stimulation of ability in MEP recordings is due to normal varia-
the motor cortex using an electrical or a magnetic tions in inhibition and facilitation in the
technique. The stimulation creates motor neuron corticospinal and cortical pathways [30]. Much of
depolarization and a descending response that tra- the latency seen in MEPs is due to the slower con-
verses the corticospinal tracts and eventually gen- ducting areas of the spinal pathway, which may
erates a measurable response either in the form of explain the MEP sensitivity to hypoperfusion and
muscle activity (compound muscle action poten- anesthetic drugs (see Chap. 19) [31, 32]. Continued

Fig. 2.1 Depiction of

the neurologic response
pathway with motor-
evoked potentials.
Stimulation of the motor
cortex (arrow) results in
a response that is
propagated through the
brain and spinal cord to
cause a muscle
contraction. The
response typically is
recorded near the muscle
as a compound muscle
action potential (CMAP)
or EMG. The response
can also be recorded
over the spinal column
as a D wave followed by
a series of I waves (high
frequency repetitive
discharges fro the
corticospinal fibers)
(from Jameson and
Sloan [33]; with
permission)
22 L.C. Jameson

investigation using DB electrodes for therapy in Table 2.1 Effect of varying the interstimulus interval
(ISI) and the stimulus pulse duration on the threshold
movement and seizure disorders will lead to a
stimulus Threshold stimulus, which can be in volts or
clearer picture of the motor pathways activated by mAmps (mA), is the energy required to produce a
diagnostic transcranial MEPs during surgery and response in 50% of the patients
may lead to a better understanding of the difficul- Pulse duration (ms)
ties in eliciting responses. 0.1 ms 0.2 ms 0.5 ms
All IOM MEP responses require continuity of ISI (ms) Mean motor threshold (mA)
the pathway since disruption of any component 2 158 ± 67 105 ± 33 76 ± 26
will change the measured response. Responses 3 140 ± 55 97 ± 33 64 ± 20
are affected by health of the neuron (e.g., periph- 4 126 ± 56 91 ± 35 61 ± 19
eral neuropathy associated with diabetes), 5 179 ± 74 120 ± 45 83 ± 31
strength of the stimulus or number of neurons Stimulus was applied at C3/C4. All combinations of ISI
contributing to the response, propagation dis- and pulse duration are significantly different from each
other at the P value of <0.001. The lowest mean motor
tance (height), sex, and temperature. Standard
threshold occurred at an ISI of 4 ms and pulse duration of
intraoperative transcranial electrical MEP moni- 0.5 ms (adapted from Szelényi et al. [36])
toring in anesthetized patients uses a high-­voltage
electrical stimulus (measured in volts) to stimu-
late pyramidal cells of the motor cortex. This pro- MEP stimulation utilizes a train of usually 3–7
duces a wave of depolarization that is estimated electrical pulses of 100–500 V intensity (maxi-
to activate only 4–5 % of the corticospinal tract. mum 1000 V) applied through corkscrew elec-
The motor pathway descends through the motor trodes most commonly placed a few centimeters
cortex, crosses the midline in the brainstem, and anterior to the somatosensory recording elec-
descends in the ipsilateral anterior funiculi of the trodes at C3′–C4′ (International 10–20 system).
spinal cord (Fig. 2.1) [2, 33]. Standard stimulus pulse durations are 0.2 ms
Attempts to stimulate spinal cord motor tracts with an interpulse or interstimulus interval (ISI)
and then record neurogenic motor-evoked poten- (period between stimuli) between 2 and 4 ms
tials (NMEPs) from peripheral nerves were done (Table 2.1). Corkscrew scalp electrodes increase
with stimulating electrodes placed into the epi- the electrode surface area and reduce the risk of
dural space (see Chap. 6) [8, 34]. An alternate, burns from the high-energy stimulus.
but less successful method was to use needle Manipulation in the number of stimuli, ISI, pulse
electrodes placed near the lamina of the appropri- duration, pulse strength or intensity, and stimu-
ate spinal segment. Beginning in the 1990s, this lating electrode locations allows for adequate
technique for obtaining responses was instituted cortical neuron depolarization. Parameter
to eliminate the difficulties associated with the changes overcome some of the impediments to
effects of anesthesia on the cerebral motor cortex propagation such as the anesthetic effect on the
when trying to elicit MEPs. NMEPs have largely anterior horn cell synapse, preexisting neuropa-
been abandoned as a motor response since cur- thy and myelopathy, distance of the motor cortex
rent evidence indicates that NMEPs are not medi- from the stimuli, loss of motor neurons, comor-
ated by the same motor pathways as MEP but bid conditions, and age. The time required to
instead by antidromic conduction in sensory obtain a MEP is generally less than 10 s. Multiple
pathways. Thus, NMEPs are not a motor response organizations have published best practice
at all [32, 35]. Direct cortical or spinal cord stim- algorithms that in their hands produce the best
ulation using a strip electrode placed directly on signals [36]. ISI manipulation is frequently cited
the spinal cord or cerebral cortex to stimulate as a critical stimulus parameter to adjust to opti-
motor pathways continues to be used to map or mize MEP acquisition (Table 2.1) [7, 36, 37].
identify neural tissue with motor functionality. Once stimulation has occurred, a reliable and
A detailed treatment of spinal cord motor map- easily detected response is required for monitor-
ping techniques with grid electrodes is found in ing purposes. The response typically used is the
Chaps 9 and 36. CMAP recorded from muscle groups in the
2 Transcranial Motor-Evoked Potentials 23

extremities, although percutaneous epidural D prehensive review articles address these issues
and I waves [38], can be used to confirm a and offer solutions to help the IOM team obtain
response (Fig. 2.1). D waves, direct activation of signals [49]. Often the most critical decision in
the corticospinal neurons [38], have a variable obtaining MEP responses, particularly in those
success rate and following them as a sole source with known neurologic, metabolic or muscular
of monitoring is currently uncommon except in diseases, is the selection of the anesthetic man-
specific surgical procedures such as intramedul- agement (see Chap. 19).
lary spinal cord tumors [39, 40]. When spinal cord D and I wave responses are
Standard muscle responses differentiate later- used, they do not differentiate laterality and D
ality and therefore localize neural tissue at risk. waves do not involve a synapse. The D wave cor-
These CMAP or EMG responses are recorded relates with the number of functioning fibers of
using needle or skin electrodes that are placed in the corticospinal tract responding to the stimulus.
hand muscles of the thenar eminence (abductor Thus, D wave amplitude changes have signifi-
or flexor pollicis brevis), in muscles of the lower cance. D waves are more commonly used during
extremities (gastrocnemius, tibialis anterior, and intramedullary spinal cord surgery where record-
abductor hallucis brevis), and trunk muscles ing electrodes are placed by the surgeon in the
(intercostals, rectus abdominis). The “best” (larg- field [33, 50, 51]. Another alternate method of
est and most reproducible) specific muscle producing a motor response is the Hoffmans
response below the site of the surgical procedure reflex (H-reflex). It is the electrical equivalent of
is selected to be followed [36, 40–46]. In our the spinal cord reflex elicited by a tendon percus-
organization, acceptable CMAP responses are sion knee jerk and monitors the sensory and
polyphasic with a consistent latency and an motor efferent axons as well as the spinal gray
amplitude greater than 150–200 μV. We will matter and components of the reflex arc [50].
continue to follow lesser responses but inform Discussion of this response is outside the scope
the surgeon that the information is not reliable. of this chapter (see Chap. 8). CMAPs are by far
Direct motor mapping in the spinal cord or cere- the most common measure of the MEP response.
bral cortex requires needle placement in the mus- The literature evaluating D, I, and H waves is
cle groups that are innervated by the areas being very limited.
stimulated (e.g., homunculus hand representa- CMAPs can demonstrate considerable vari-
tion–abductor or flexor pollicis brevis). This ability even in normal awake subjects [32, 52].
includes those muscles innervated by the cra- The variability is magnified during general anes-
nial nerves (e.g., cranial nerve VII: orbicularis thesia [31, 53]. Most organizations establish
oculi or oris). standardized criteria for a minimum baseline
CMAPs can be difficult to obtain from patients amplitude (difference between positive and nega-
at both extremes of age, elderly and young chil- tive peaks), complexity (number of positive and
dren. In addition to age, adults often have preex- negative wavelets) but not latency (time from
isting conditions such as diabetes, hypertension, stimulus to response). This is necessary to pre-
chronic spinal cord compression, nerve root vent false-positive monitoring alert when the
injury, chronic hypoperfusion, and axonal con- signal changes. Without these waveform compo-
duction changes that reduce CMAP responses nents, a reliable signal was never present. It
[47]. Children, particularly those under 6 years, assures the surgeon that the MEP responses will
have an immature CNS that makes obtaining a be a reliable measure of function throughout the
motor response challenging [48]. CMAP procedure. MEP responses are presented in
responses can be difficult to obtain in procedures Fig. 2.2. What constitutes a CMAP change that
that are performed on patients with substantial must be acted upon has not been universally
neurologic deficits from preexisting brain injury defined. Permanent loss, a straightforward event,
(e.g., cerebral palsy) and genetic diseases that is strongly correlated with permanent neurologic
impair muscle function (e.g., Duchene muscular injury, whereas patients who experience tempo-
dystrophy, Charcot-Marie-Tooth). Recent com- rary loss or alerts (a predetermined decrease in
24 L.C. Jameson

Fig. 2.2 Standard normal MEP responses. The CMAP lower extremity muscle groups are used due to the
response, a large polyphasic wave, is obtained from the increased difficulty obtaining a consistent response par-
upper extremity traditionally using the abductor pollicis ticularly in adults. Other upper and lower extremity mus-
brevis (APB) and from the lower extremity using tibialis cles can be used depending on the needs of the specific
anterior (TA) and abductor hallicus (AH) brevis. Two patient. Obtained from the author’s archive

Fig. 2.3 Normal MEP

baseline responses and
acute injury. Placing the
patient prone resulted in
the loss of responses.
There was a recovery of
response to baseline
configuration after
adjusting the head
position, increasing
blood pressure and
eliminating residual
desflurane. Obtained
from the author’s
archive

amplitude and/or latency) frequently have nor- time. Other suggested criteria include increases
mal function at the end of the procedure and ulti- in stimulus strength greater than 50–100 V,
mately regain full motor function. Some IOM changes in stimuli number or pattern required to
groups use the presence or absence of a CMAP elicit an MEP, or a significant decrease in CMAP
response as their sole criteria for notifying the amplitudes (usually >80 %) from the initial
surgeon about a problem. This criterion allows responses (without muscle relaxant). All are
the use of muscle relaxants as a component of the considered significant changes by some individu-
anesthetic, which is a common surgical request to als (Fig. 2.3). Signal recovery after these changes
eliminate patient movement at an inopportune is reassuring and usually predicts normal postop-
2 Transcranial Motor-Evoked Potentials 25

erative motor function. Loss of CMAP responses tion in anesthetic management to obtain adequate
requires notification of the surgeon and anesthe- waveforms—a point that necessitates negotia-
siologist to correct, when possible, the physio- tions between the anesthesiologist, surgeon, and
logic issues contributing to the MEP change (see IOM team. Many of the older prospective series
Chap. 20) [1, 54–59]. used SSEPs and EMG but only rarely MEPs due
to this issue. In one study of 1055 adult patients
undergoing cervical spine surgery between 2000
Application of MEP Monitoring and 2005, MEP studies were attempted and
obtained in only 26 of 1055 patients due to the
Monitoring during structural spine and spinal cord perceived difficulties [61]. These were the highest
surgery is customarily multimodal and includes risk patients for spinal cord injury. With the cur-
SSEPs, MEPs, and electromyography (EMG, free rent relatively routine availability of total intrave-
running, and stimulated). MEP monitoring is con- nous anesthesia (TIVA) based on propofol, MEPs
sidered essential whenever spinal cord function is can be relatively easy to obtain (see Chap. 19).
at risk. Thus, MEP monitoring is usually per- When used during spine procedures, MEPs had
formed during structural spine surgery from C1 to 100 % sensitivity, 96 % specificity, and a positive
cauda equina whenever the risk of cord injury due predictive value of 96 % for postoperative motor
to stretch, compression, or vascular damage changes [61]. Adults with cervical myelopathy
[56, 60] could occur. “At risk” situations also had about a 12 % incidence of only MEP alerts
include any surgery where a compromise of spinal (no EMG or SSEP changes). These alerts were
cord perfusion or direct injury to motor tracks or usually followed by resolution after alterations in
nerve roots could occur. Consensus opinion is that anesthetic and surgical management occurred.
the evidence supports MEP monitoring in the fol- Nonetheless, these authors believed that the MEP
lowing specific spine procedures: monitoring provided 100 % sensitivity and 90 %
specificity [64].
• Spinal deformities with scoliosis greater than MEP changes are relatively infrequent [7] in
45° rotation pediatric procedures. One group reported that in
• Congenital spine anomalies 172 pediatric spinal deformity corrective proce-
• Resections of intramedullary and extramedul- dures, there were 15 intraoperative MEP alerts, all
lary tumors of which resolved with changes in management.
• Extensive anterior and/or posterior decom- None of the patients (MEP-alert and MEP-­
pressions in spinal stenosis with myelopathy unchanged patients) had new neurologic deficits.
• Functional disturbance of the cauda equina This group concluded that MEP monitoring alone
and/or individual nerve roots was adequate for spinal deformity surgery with a
sensitivity of 1.0 and a specificity of 0.97. Patients
However, the evidence does not meet the with persistent MEP changes had immediate post-
level 1 standard (large randomized, placebo-­ operative motor deficits. SSEP changes, when
controlled, double-blind studies). The evidence present, lagged significantly behind the MEP
is based on a large case series and meta-analysis changes and often did not predict outcome [65].
(level 2, 3 evidence) where MEP changes pre- For adults with spinal cord myelopathy, one of the
dicted immediate postsurgical neurologic find- diagnostic criteria includes changes in MEPs prior
ings [1, 40, 54–56, 58, 59, 61–63]. A recent to surgical intervention; consequently, baseline
evidence-based analysis by the American studies, post anesthesia, and prepositioning are
Academy of Neurology and American Clinical strongly recommended [66, 67].
Neurophysiology Society is strongly supportive Consensus opinion supports and studies sug-
of IOM in spine surgery [46, 56]. gest that the use of intraoperative spinal cord
Obtaining MEPs remains challenging in some motor mapping improves long-term motor func-
patient populations. They often require an altera- tion in intramedullary spinal cord tumor resection
26 L.C. Jameson

(see Chap. 36) [68, 69]. The MEP is the only reli- motor deficits and only 4 patients had deficits
able monitor of motor pathways and is an early (1 %) remaining 3 months after surgery. Total or
predictor of impending damage to the cord due to subtotal tumor resection was done in only 11 % of
the precarious blood supply. For an anterior patients prior to motor mapping but 69.8 % after
approach to an intramedullary spinal cord tumor motor mapping was initiated [74]. A number of
resection, focal injury to the anterior spinal vascu- other groups have published similar reports and
lature or motor tracts is generally detected only noted that the long-term outcome is significantly
minutes after a hypoperfusion event; this is con- improved by more extensive tumor resection in
siderably faster than with SSEP monitoring alone both children and adults for all supratentorial
[11, 68, 69]. tumors [72, 75, 76]. Neurologic injury to the pos-
During intracranial procedures, direct cortical terior fossa can have devastating consequences.
stimulation is likewise used to map motor func- Motor mapping is an effective way to identify both
tion and to delineate the demarcation between tumor margins and safe resection zones, areas
tumor and functional tissue. This can be per- between cranial nerve nuclei, or entry zones into
formed using electrode strips, direct hand-held the floor of the fourth ventricle. Stimulation can be
device, or a Penfield motor stimulation technique. either transcranial or, more frequently, direct
Although the Penfield technique is often pre- brainstem stimulation [77].
ferred in awake patients, the pulse train technique Intracranial aneurysms and arteriovenous mal-
used for MEP is associated with less stimulation- formations can result in areas of hypoperfusion
related seizures and is more effective in produc- during the endovascular embolization, resection,
ing CMAP responses during general anesthesia. or temporary and permanent clipping. MEPs iden-
The motor stimulation may replace or augment tify hypoperfusion in motor areas and adjacent
awake craniotomy procedures in the supratento- areas perfused by vessels involved in the vascular
rial area when eloquent areas (e.g., speech) or lesion. Identification of MEP change followed by
motor pathways (e.g., internal capsule, motor therapeutic intervention appears to substantially
cortex, and premotor cortex) are at risk [70–72]. reduce permanent injury. Two large studies with
In large clinical studies, sensitivity and specificity 108 and 129 patients undergoing supratentorial
are reported to be between 90 % and 100 %, aneurysm clippings found that in cases where
respectively; however, in some reports, Broca’s MEPs were unchanged, none of the patients had
area had a reported specificity of only 64 % and deficits. One study confirmed adequate flow with
Wernicke’s area of just 18 % [73]. MEPs and with microvascular Doppler ultraso-
Excess stimulation strength can cause direct nography. Both studies reported between 13 and
activation of structures at a distance from the stim- 33 % of patients had reversible MEP changes;
ulus location. Focal stimulation often involves these patients had no neurologic changes immedi-
manually stimulating portions of the cortex or ately after the procedure or had only transient neu-
inserting a strip electrode under the dura. The rologic changes from which they fully recovered.
usual pattern of transcranial stimulation is performed Patients with permanent MEP change (about 20 %)
at about 1/10th the MEP stimulus strength. A pat- had permanent neurologic deficits, some quite
tern of gradually increasing stimulus is applied. severe [16, 78, 79]. Small case series generally
Recent large case reports have documented that support these findings. The neurosurgical commu-
MEP monitoring assists in delineating the edge nity has reported improved outcomes during aneu-
between tumor cells and functioning neural tissue. rysm occlusion on basilar, vertebral, and middle
In a study of 404 patients, all with a low-­grade cerebral artery aneurysms when using MEP moni-
glioma, MEP mapping substantially reduced the toring. Publications report MEP changes occur
number and severity of permanent motor deficits rapidly and better reflect long-­ term outcomes
while increasing the number of total resections. when the involved vessels provide perfusion to
One hundred of these 100 patients had temporary motor pathways.
2 Transcranial Motor-Evoked Potentials 27

 ontribution of Anesthesiology
C of the combination of blood pressure and
to Effective MEP Monitoring oxygen-­ carrying capacity. Consequently it
assesses the adequacy of perfusion in specific
Without the cooperation and support of the anes- patients under specific surgical conditions. When
thesia care provider, producing MEP responses IOM signals deteriorate, increasing the systemic
and detecting changes is not possible. Most treat- blood pressure to the patient’s preoperative value
ment options, when MEP change occurs, are in or higher is the most common and most effective
the hands of the anesthesia caregiver. MEP response the anesthesia care team can provide.
change is not only initiated by surgical activity Transfusion is also an effective therapeutic inter-
but by physiologic management and anesthetic vention when appropriate. Maintenance of “nor-
drug choices. Any event that will impact neural mal” physiologic conditions within the brain and
function can impact MEP waveforms (see Chap. spinal cord can be difficult but results in the ideal
19). This reality stresses the importance of the monitoring conditions and the best neurologic
team effort, cooperation between the surgeon, outcomes.
anesthesiologist, and IOM technologist. The impact of dexmedetomidine on MEP
Hypotension is of particular interest since monitoring deserves special comment. Propofol
deliberate hypotension to reduce blood loss was infusion syndrome [82] is diagnosed primarily in
once considered a management technique, par- pediatric patients and can prove fatal (see
ticularly in the idiopathic scoliosis procedures in Chap. 19). Thus, substituting dexmedetomidine
children and during aneurysm clipping. There is for propofol as the “recommended” TIVA hyp-
a growing appreciation that the presumed lower notic when IOM is required has been advocated.
limit of autoregulation is not always adequate Early literature reports suggested that its use
for tissues undergoing surgical stress [80]. Mean caused no negative physiologic effect or impair-
BP that is adequate for a young adult patient may ment in MEP monitoring [83–85]. Two recent
not be adequate for an older adult with many carefully performed studies found a clinically
coexisting diseases. Hence, increasing or main- and statistically significant attenuation in the
taining systemic perfusion pressure effectively amplitudes of MEPs when the targeted plasma
treats many impending hypoperfusion injuries concentrations of demedetomidine exceeded
(Fig. 2.4). 0.6–0.8 ng/mL [83, 86, 87]. Another study in
The acceptable lower limit for hemoglobin which dexmedetomidine was administered in
has come under question. Current recommenda- combination with propofol was discontinued by
tions by the blood banking community are to the safety monitoring board. Reduction or loss of
allow hemoglobin to be as low as 7 g/dL during MEPs occurred in healthy pediatric spines when
acute blood loss, particularly in healthy patients both drugs were used in any combination [83].
[81]. However, anemia can be compensated only Dexmedetomidine also has a long context-­
within the limits of the patient’s physiologic sensitive half-life consequently wakeup times
ability to increase cardiac output to maintain can be prolonged.
local tissue perfusion. Neurologic tissue has a
high metabolic demand and may have compro-
mised perfusion due to pre-existing systemic dis- Risk of MEP Monitoring
ease (hypertension, vascular disease, poor
cardiac output, surgical stress, and inflamma- MEP monitoring is not without risk. The US
tion) as well as regional compression (spinal Food and Drug Administration has specified
cord stenosis, surgical activity, position, acute relative MEP contraindications. The most com-
injury). Thus the acceptable lower limit of blood mon concern was direct cortical thermal injury
pressure and hemoglobin is unlikely to be the (kindling), but over the last 18 years only two
same for all situations and is poorly predictable. cases of cortical thermal injury have been
MEP monitoring allows a functional assessment reported. In a 2002 survey of the literature,
28 L.C. Jameson

Fig. 2.4 Recovery of MEP responses after intraopera- had weakness on the left, which resolved over 2 weeks.
tive loss. During a posterior cervical fusion of C5 to T4 On the right, the patient had a dense hemiparesis that had
the patient abruptly lost MEP responses in both lower not changed 3 months after surgery. (Tibialis anterior–
extremities. After BP elevation and steroid administra- TA, Adductor hallucis–AH) Obtained from the author’s
tion, MEP responses returned only on the left. Patient archive

p­ ublished complications included tongue lacera- onset seizures, epidural hematomas, or infec-
tion (n = 29), cardiac arrhythmia (n = 5), scalp tions from epidural electrodes or movement
burn at the site of stimulating electrodes (n = 2), injuries (e.g., surgical, joint dislocation), neuro-
jaw fracture (n = 1), and awareness (n = 1) [88]. psychiatric disease, headaches, and endocrine
Placing a bite block between both molars can abnormalities have been reported. Relative MEP
ameliorate tongue laceration. Notably no new- contraindications include epilepsy, a cortex
2 Transcranial Motor-Evoked Potentials 29

lesion, skull defects, high intracranial pressure, somatosenoury-evoked potentials from the lower and
upper extremities. Spine. 2003;28(2):194–203.
an intracranial apparatus (electrodes, vascular
7. Hsu B, Cree AK, Lagopoulos J, Cummine JL.
clips, and shunts), cardiac pacemakers, or other Transcranial motor-evoked potentials combined with
implanted pumps. The most common patient response recording through compound muscle action
identified side effect is sore muscles [89, 90]. potential as the sole modality of spinal cord monitor-
ing in spinal deformity surgery. Spine (Phila Pa 1976).
Needle placement will lead to bleeding and
2008;33(10):1100–6. Epub 2008/05/02.
bruising at the insertion site. Infection is always 8. *
Minahan RE, Sepkuty JP, Lesser RP, Sponseller PD,
possible. The prevalence of major and minor Kostuik JP. Anterior spinal cord injury with preserved
problems is astonishingly low. neurogenic ‘motor’ evoked potentials. Clin
Neurophysiol. 2001;112(8):1442–50. Epub 2001/07/19.
9. Deletis V, Sala F. Intraoperative neurophysiological
monitoring of the spinal cord during spinal cord and
Conclusion spine surgery: a review focus on the corticospinal
tracts. Clin Neurophysiol. 2008;119(2):248–64. Epub
2007/12/07.
Clearly the goal of intraoperative monitoring is
10. Yanni DS, Ulkatan S, Deletis V, Barrenechea IJ, Sen
to provide the greatest degree of assistance to C, Perin NI. Utility of neurophysiological monitoring
the operative team for optimal intraoperative using dorsal column mapping in intramedullary spinal
decision-­making. The current literature suggests cord surgery. J Neurosurg Spine. 2010;12(6):623–8.
11. Morota N, Deletis V, Constantini S, Kofler M, Cohen
that MEP monitoring provides excellent specific-
H, Epstein FJ. The role of motor evoked potentials
ity and sensitivity whenever motor tracts are during surgery for intramedullary spinal cord tumors.
involved. As such, the real question for consider- Neurosurgery. 2010;41(6):1327–36.
*
ation is which of the techniques available should 12. Sala F, Bricolo A, Faccioli F, Lanteri P, Gerosa M,
Sala F, et al. Surgery for intramedullary spinal cord
be used to complement MEP monitoring in indi-
tumors: the role of intraoperative (neurophysiological)
vidual patients. monitoring. Eur Spine J. 2007;16 Suppl 2:S130–9.
13. Mikuni N, Okada T, Enatsu R, Miki Y, Hanakawa T,
Urayama S, et al. Clinical impact of integrated func-
tional neuronavigation and subcortical electrical stim-
References1 ulation to preserve motor function during resection of
brain tumors. J Neurosurg. 2007;106(4):593–8.
1. *Raynor BL, Bright JD, Lenke LG, Rahman RK, 14. Neuloh G, Pechstein U, Schramm J, Neuloh G,
Bridwell KH, Riew KD, et al. Significant change or loss Pechstein U, Schramm J. Motor tract monitoring
of intraoperative monitoring data: a 25-year experience during insular glioma surgery. J Neurosurg.
in 12,375 spinal surgeries. Spine. 2013;38:E101–8. 2007;106(4):582–92.
2. Waxman S. Control of movement. In: Waxman SG, 15. *
Neuloh G, Bogucki J, Schramm J. Intraoperative
editor. Clinical neuroanatomy 27/E. 27th ed. preservation of corticospinal function in the brain-
New York: McGraw Hill Professional; 2013. stem. J Neurol Neurosurg Psychiatry. 2009;80(4):
p. 183–94. 417–22.
3. Hickey R, Sloan TB, Rogers JN. Functional organiza- 16. Szelényi A, Langer D, Kothbauer K, De Camargo
tion and physiology of the spinal cord. In: Porter SS, AB, Flamm ES, Deletis V. Monitoring of muscle
editor. Anesthesia for surgery of the spine. New York: motor evoked potentials during cerebral aneurysm
McGraw-Hill; 1995. p. 15–39. surgery: intraoperative changes and postoperative
4. Fehlings MG, Houldon D, Vajkoczy P. Introduction. outcome. J Neurosurg. 2006;105(5):675–81. Epub
Intraoperative neuromonitoring: an essential compo- 2006/11/24.
nent of the neurosurgical and spinal armamentarium. 17. Neuloh G, Schramm J. Monitoring of motor evoked
Neurosurg Focus. 2009;27(4):E1. potentials compared with somatosensory evoked
5. Pelosi L, Lamb J, Grevitt M, Mehdian SM, Webb JK, potentials and microvascular Doppler ultrasonogra-
Blumhardt LD. Combined monitoring of motor and phy in cerebral aneurysm surgery. J Neurosurg.
somatosensory evoked potentials in orthopaedic spi- 2004;100(3):389–99.
nal surgery. Clin Neurophysiol. 2002;113(7):1082– 18. Neuloh G, Bien CG, Clusmann H, von Lehe M,
91. Epub 2002/06/29. Schramm J. Continuous motor monitoring enhances
6. MacDonald D, Zayed Z, Khoudeir I, Stigsby functional preservation and seizure-free outcome in
B. Monitoring scoliosis surgery with combined mul- surgery for intractable focal epilepsy. Acta Neurochir
tiple pulse trascranial electric motor and cortical (Wien). 2010;152(8):1307–14.
19. Corti M, Patten C, Triggs W. Repetitive transcranial
1
Asterisks indicate key references. magnetic stimulation of motor cortex after stroke: a
30 L.C. Jameson

focused review. Am J Phys Med Rehabil. 34. Toleikis JR, Skelly JP, Carlvin AO, Burkus
2012;91(3):254–70. JK. Spinally elicited peripheral nerve responses are
20. Nascimbeni A, Gaffuri A, Imazio P, Nascimbeni A, sensory rather than motor. Clin Neurophysiol.
Gaffuri A, Imazio P. Motor evoked potentials: prog- 2000;111(4):736–42.
nostic value in motor recovery after stroke. Funct 35. Amassian VE, Stewart M, Quirk GJ, Rosenthal JL.
Neurol. 2006;21(4):199–203. Physiological basis of motor effects of a transient
21. Woldag H, Gerhold LL, de Groot M, Wohlfart K, stimulus to cerebral cortex. Neurosurgery.
Wagner A, Hummelsheim H. Early prediction of 1987;20(1):74–93.
functional outcome after stroke. Brain Inj. 36. *Szelényi A, Kothbauer KF, Deletis V. Transcranial
2006;20(10):1047–52. electric stimulation for intraoperative motor evoked
22. Waxman S. Spinal cord. In: Waxman SG, editor. potential monitoring: stimulation parameters and
Clinical neuroanatomy 27/E. 27th ed. New York: electrode montages. Clin Neurophysiol.
McGraw Hill Professional; 2013. p. 43–147. 2007;118(7):1586–95.
23. Crawford ES, Svensson LG, Hess KR, Shenaq SS, 37. *Deletis V. Basic methodological principles of multi-
Coselli JS, Safi HJ, et al. A prospective randomized modal intraoperative monitoring during spine surger-
study of cerebrospinal fluid drainage to prevent para- ies. Eur Spine J. 2007;16 Suppl 2:S147–52.
plegia after high-risk surgery on the thoracoabdomi- 38. Houlden DA, Schwartz ML, Tator CH, Ashby P,
nal aorta. J Vasc Surg. 1991;13:36–45. MacKay WA. Spinal cord-evoked potentials and mus-
24. Schurink GWH, Nijenhuis RJ, Backes WH, Mess W, cle responses evoked by transcranial magnetic stimu-
de Haan MW, Mochtar B, et al. Assessment of spinal lation in 10 awake human subjects. J Neurosci.
cord circulation and function in endovascular treat- 1999;19(5):1855–62.
ment of thoracic aortic aneurysms. Ann Thorac Surg. 39. Costa P, Peretta P, Faccani G. Relevance of intraop-
2007;83(2):S877–81. discussion S90–2. erative D wave in spine and spinal cord surgeries. Eur
25. Okita Y. Fighting spinal cord complication during sur- Spine J. 2013;22(4):840–8.
gery for thoracoabdominal aortic disease. Gen Thorac 40. Gavaret M, Jouve JL, Pereon Y, Accadbled F, Andre-­
Cardiovasc Surg. 2011;59(2):79–90. Obadia N, Azabou E, et al. Intraoperative neurophysi-
26. Wan IY, Angelini GD, Bryan AJ, Ryder I, Underwood ologic monitoring in spine surgery. Developments
MJ. Prevention of spinal cord ischaemia during and state of the art in France in 2011. Orthop
descending thoracic and thoracoabdominal aortic sur- Traumatol Surg Res. 2013;99(6 Suppl):S319–27.
gery. Eur J Cardiothorac Surg. 2001;19(2):203–13. 41. Fernandez-Conejero I, Deletis V. Transcranial electri-
27. Sakuma J, Suzuki K, Sasaki T, Matsumoto M, Oinuma cal stimulation and monitoring. J Neurosurg.
M, Kawakami M, et al. Monitoring and preventing 2014;120(1):291–2.
blood flow insufficiency due to clip rotation after the 42. Joksimovic B, Damjanovic A, Damjanovic A, Rasulic
treatment of internal carotid artery aneurysms. L. Transcranial electric stimulation for intraoperative
J Neurosurg. 2004;100(5):960–2. motor evoked potential monitoring: dependence of
28. Horiuchi K, Suzuki K, Sasaki T, Matsumoto M, required stimulation current on interstimulus interval
Sakuma J, Konno Y, et al. Intraoperative monitoring value. J Neurol Surg A Cent Eur Neurosurg.
of blood flow insufficiency during surgery of middle 2015;76(3):190–8.
cerebral artery aneurysms. J Neurosurg. 43. Ukegawa D, Kawabata S, Sakaki K, Ishii S, Tomizawa
2005;103(2):275–83. S, Inose H, et al. Efficacy of biphasic transcranial
29. Ghitani N, Bayguinov PO, Vokoun CR, McMahon S, electric stimulation in intraoperative motor evoked
Jackson MB, Basso MA. Excitatory synaptic feed- potential monitoring for cervical compression
back from the motor layer to the sensory layers of the myelopathy. Spine (Phila Pa 1976).
superior colliculus. J Neurosci. 2014;34(20): 2014;39(3):E159–65.
6822–33. 44. Yellin JL, Wiggins CR, Franco AJ, Sankar WN. Safe
30. Ferreri F, Pasqualetti P, Maatta S, Ponzo D, Ferrarelli transcranial electric stimulation motor evoked poten-
F, Tononi G, et al. Human brain connectivity during tial monitoring during posterior spinal fusion in two
single and paired pulse transcranial magnetic stimula- patients with cochlear implants. J Clin Monit Comput.
tion. Neuroimage. 2011;54(1):90–102. 2016;30(4):503–6 [Epub ahead of print].
31. Firmin L, Muller S, Rosler KM. A method to measure 45. Kobayashi S, Matsuyama Y, Shinomiya K, Kawabata
the distribution of latencies of motor evoked potentials S, Ando M, Kanchiku T, et al. A new alarm point of
in man. Clin Neurophysiol. 2011;122(1):176–82. transcranial electrical stimulation motor evoked
32. Tsutsui S, Yamada H, Hashizume H, Minamide A, potentials for intraoperative spinal cord monitoring: a
Nakagawa Y, Iwasaki H, et al. Quantification of the prospective multicenter study from the Spinal Cord
proportion of motor neurons recruited by transcranial Monitoring Working Group of the Japanese Society
electrical stimulation during intraoperative motor for Spine Surgery and Related Research. J Neurosurg
evoked potential monitoring. J Clin Monit Comput. Spine. 2014;20(1):102–7.
2013;27(6):633–7. 46. Ney JP, van der Goes DN, Nuwer M, Emerson R,
33. Jameson LC, Sloan TB. Monitoring of the brain and Minahan R, Legatt A, et al. Evidence-based guideline
spinal cord. Anesthesiol Clin. 2006;24(4):777–91. update: intraoperative spinal monitoring with somato-
2 Transcranial Motor-Evoked Potentials 31

sensory and transcranial electrical motor evoked (IONM): lessons learned from 32 case events in 2069
potentials: report of the Therapeutics and Technology spine cases. Am J Electroneurodiagnostic Technol.
Assessment Subcommittee of the American Academy 2011;51(4):247–63.
of Neurology and the American Clinical 59. Malhotra NR, Shaffrey CI. Intraoperative electro-
Neurophysiology Society. Neurology. 2012;79(3): physiological monitoring in spine surgery. Spine.
292–4. 2010;35(25):2167–79.
47. Deiner SG, Kwatra SG, Lin H-M, Weisz DJ. Patient 60. Sutter M, Deletis V, Dvorak J, Eggspuehler A, Grob
characteristics and anesthetic technique are additive D, Macdonald D, et al. Current opinions and recom-
but not synergistic predictors of successful motor mendations on multimodal intraoperative monitoring
evoked potential monitoring. Anesth Analg. during spine surgeries. Eur Spine J. 2007;16 Suppl
2010;111(2):421–5. 2:S232–7.
48. Lieberman JA, Lyon R, Feiner J, Diab M, Gregory 61. Kelleher MO, Tan G, Sarjeant R, Fehlings
GA. The effect of age on motor evoked potentials in MG. Predictive value of intraoperative neurophysio-
children under propofol/isoflurane anesthesia. Anesth logical monitoring during cervical spine surgery: a
Analg. 2006;103(2):316–21. prospective analysis of 1055 consecutive patients.
*
49. Sala F, Manganotti P, Grossauer S, Tramontanto V, J Neurosurg Spine. 2008;8(3):215–21.
Mazza C, Gerosa M. Intraoperative neurophysiology 62. Sutter MA, Eggspuehler A, Grob D, Porchet F,
of the motor system in children: a tailored approach. Jeszenszky D, Dvorak J. Multimodal intraoperative
Childs Nerv Syst. 2010;26(4):473–90. monitoring (MIOM) during 409 lumbosacral surgical
50. Leppanen RE. Intraoperative monitoring of segmen- procedures in 409 patients. Eur Spine J. 2007;16
tal spinal nerve root function with free-run and Suppl 2:S221–8.
electrically-­
triggered electromyography and spinal 63. Eggspuehler A, Sutter MA, Grob D, Jeszenszky D,
cord function with reflexes and F-responses. A posi- Porchet F, Dvorak J. Multimodal intraoperative moni-
tion statement by the American Society of toring (MIOM) during cervical spine surgical proce-
Neurophysiological Monitoring. J Clin Monit dures in 246 patients. Eur Spine J. 2007;16 Suppl
Comput. 2005;19(6):437–61. 2:S209–15.
*
51. Jameson LC, Sloan TB. Neurophysiologic monitoring 64. Kim DH, Zaremski J, Kwon B, Jenis L, Woodard E,
in neurosurgery. Anesthesiol Clin. 2012;30(2):311–31. Bode R, et al. Risk factors for false positive transcra-
52. Wassermann EM. Variation in the response to tran- nial motor evoked potential monitoring alerts during
scranial magnetic brain stimulation in the general surgical treatment of cervical myelopathy. Spine
population. Clin Neurophysiol. 2002;113(7): (Phila Pa 1976). 2007;32(26):3041–6.
1165–71. 65. Haghighi SS, Mundis G, Zhang R, Ramirez
53. Sloan TB. Anesthesia and the brain, does it matter? B. Correlation between transcranial motor and
Anesthesiol Clin North America. 2002;20:1–27. somatosensory-evoked potential findings in cervical
54. Davis SF, Corenman D, Strauch E, Connor myelopathy or radiculopathy during cervical spine
D. Intraoperative monitoring may prevent neurologic surgery. Neurol Res. 2011;33(9):893–8.
injury in non-myelopathic patients undergoing 66. Wilson JR, Fehlings MG, Kalsi-Ryan S, Shamji MF,
ACDF. Neurodiagn J. 2013;53:114–20. Tetreault LA, Rhee JM, Chapman JR. Diagnosis, heri-
55. Avila EK, Elder JB, Singh P, Chen X, Bilsky tability, and outcome assessment in cervical myelopa-
MH. Intraoperative neurophysiologic monitoring and thy: a consensus statement. Spine (Phila Pa 1976).
neurologic outcomes in patients with epidural spine 2013;38(22S):S76–7.
tumors. Clin Neurol Neurosurg. 2013;115(10): 67. Wilson JR, Barry S, Fischer DJ, Skelly AC, Arnold
2147–52. PM, Riew KD, et al. Frequency, timing, and predic-
*
56. Nuwer MR, Emerson RG, Galloway G, Legatt AD, tors of neurological dysfunction in the nonmyelo-
Lopez J, Minahan R, et al. Evidence-based guideline pathic patient with cervical spinal cord compression,
update: intraoperative spinal monitoring with canal stenosis, and/or ossification of the posterior lon-
somatosensory and transcranial electrical motor gitudinal ligament. Spine (Phila Pa 1976). 2013;38(22
evoked potentials: report of the Therapeutics and Suppl 1):S37–54.
Technology Assessment Subcommittee of the 68. Quinones-Hinojosa A, Gulati M, Lyon R, Gupta N,
American Academy of Neurology and the American Yingling C, Quinones-Hinojosa A, et al. Spinal cord
Clinical Neurophysiology Society. Neurology. mapping as an adjunct for resection of intramedullary
2012;78(8):585–9. tumors: surgical technique with case illustrations.
57. Gavaret M, Trebuchon A, Aubert S, Jacopin S, Neurosurgery. 2002;51(5):1199–206. discussion
Blondel B, Glard Y, et al. Intraoperative monitoring in 206–7.
pediatric orthopedic spinal surgery: three hundred 69. Cheng JS, Ivan ME, Stapleton CJ, Quinones-Hinojosa
consecutive monitoring cases of which 10% of A, Gupta N, Auguste KI. Intraoperative changes in
patients were younger than 4 years of age. Spine. transcranial motor evoked potentials and somatosen-
2011;36(22):1855–63. sory evoked potentials predicting outcome in children
58. Eager M, Shimer A, Jahangiri FR, Shen F, Arlet V. with intramedullary spinal cord tumors. J Neurosurg
Intraoperative neurophysiological monitoring Pediatr. 2014;13(6):591–9.
32 L.C. Jameson

70. Balogun JA, Khan OH, Taylor M, Dirks P, Der T, monitoring during scoliosis surgery using a total
Carter Snead Iii O, et al. Pediatric awake craniotomy intravenous anesthetic regimen including dexmedeto-
and intra-operative stimulation mapping. J Clin midine. J Clin Neurophysiol. 2008;25(1):56–61.
Neurosci. 2014;21(11):1891–4. 85. Koruk S, Mizrak A, Kaya Ugur B, Ilhan O, Baspinar
71. Ringel F, Sala F. Intraoperative mapping and monitor- O, Oner U. Propofol/dexmedetomidine and propofol/
ing in supratentorial tumor surgery. J Neurosurg Sci. ketamine combinations for anesthesia in pediatric
2015;59(2):129–39. patients undergoing transcatheter atrial septal defect
72. Bello L, Riva M, Fava E, Ferpozzi V, Castellano A, closure: a prospective randomized study. Clin Ther.
Raneri F, et al. Tailoring neurophysiological strategies 2010;32(4):701–9.
with clinical context enhances resection and safety 86. Mahmoud M, Sadhasivam S, Sestokas AK, Samuels P,
and expands indications in gliomas involving motor McAuliffe J. Loss of transcranial electric motor evoked
pathways. Neuro Oncol. 2014;16(8):1110–28. potentials during pediatric spine surgery with dexme-
73. Trinh VT, Fahim DK, Maldaun MV, Shah K, detomidine. Anesthesiology. 2007;106(2):393–6.
McCutcheon IE, Rao G, et al. Impact of preoperative 87. Bala E, Sessler DI, Nair DR, McLain R, Dalton JE,
functional magnetic resonance imaging during awake Farag E. Motor and somatosensory evoked potentials
craniotomy procedures for intraoperative guidance are well maintained in patients given dexmedetomi-
and complication avoidance. Stereotact Funct dine during spine surgery. Anesthesiology. 2008;
Neurosurg. 2014;92(5):315–22. 109(3):417–25.
74. Bertani G, Fava E, Casaceli G, Carrabba G, Casarotti 88. Legatt AD. Current practice of motor evoked ­potential
A, Papagno C, et al. Intraoperative mapping and mon- monitoring: results of a survey. J Clin Neurophysiol.
itoring of brain functions for the resection of low-­ 2002;19(5):454–60.
*
grade gliomas: technical considerations. Neurosurg 89. Macdonald DB, Skinner S, Shils J, Yingling
Focus. 2009;27(4):E4. C. Intraoperative motor evoked potential monitoring:
75. Sanai N. Emerging operative strategies in neurosur- a position statement by the American Society of
gical oncology. Curr Opin Neurol. 2012;25(6): Neurophysiological Monitoring. Clin Neurophysiol.
756–66. 2013;124(12):2291–316.
76. Sanai N, Berger MS. Glioma extent of resection and 90. Macdonald DB. Intraoperative motor evoked poten-
its impact on patient outcome. Neurosurgery. tial monitoring: overview and update. J Clin Monit
2008;62:753–6. Comput. 2006;20(5):347–77.
77. Morota N, Ihara S, Deletis V. Intraoperative neuro-
physiology for surgery in and around the brainstem:
role of brainstem mapping and corticobulbar tract
Questions
motor-evoked potential monitoring. Childs Nerv Syst.
2010;26(4):513–21. 1. Which of the following does NOT decrease the
78. *Szelényi A, Kothbauer K, de Camargo AB, Langer likelihood that MEP can be acquired in the OR
D, Flamm ES, Deletis V. Motor evoked potential a. Very young age
monitoring during cerebral aneurysm surgery: techni-
b. Diabetes
cal aspects and comparison of transcranial and direct
cortical stimulation. Neurosurgery. 2005;57(4 c. Long-standing hypertension
Suppl):331–8. d. Myelopathy
79. Neuloh G, Schramm J. Motor evoked potential moni- e. All of the above
toring for the surgery of brain tumours and vascular
2. During surgery, MEP change in the tibialis
malformations. [Review] [126 refs]. Adv Tech Stand
Neurosurg. 2004;29:171–228. anterior that is NOT resolved by the conclusion
80. Edmonds Jr HL. Multi-modality neurophysiologic of surgery correlates with
monitoring for cardiac surgery. Heart Surg Forum. a. Loss of proprioception in the feet
2002;5(3):225–8.
b. Loss of vibration sense in the hands
81. Goodnough LT, Levy JH, Murphy MF. Concepts of
blood transfusion in adults. Lancet. 2013; c. Loss of motor function in the leg
381(9880):1845–54. d. Loss of speech discrimination
82. Marik PE. Propofol: therapeutic indications and side-­ e. Loss of visual acuity
effects. Curr Pharm Des. 2004;10(29):3639–49.
3. Which of the following has been associated
83. Mahmoud M, Sadhasivam S, Salisbury S, Nick TG,
Schnell B, Sestokas AK, et al. Susceptibility of tran- with EMG monitoring?
scranial electric motor-evoked potentials to varying a. Epidural D waves
targeted blood levels of dexmedetomidine during b. H reflex
spine surgery. Anesthesiology. 2010;112(6):1364–73.
c. Stimulation of cranial nerve VII during an
84. Anschel DJ, Aherne A, Soto RG, Carrion W, Hoegerl
C, Nori P, et al. Successful intraoperative spinal cord acoustic neuroma
2 Transcranial Motor-Evoked Potentials 33

d. Stimulation of the posterior tibial nerve d. The MEP is less sensitive to ischemia in
e. Neurogenic motor-evoked potentials the spinal cord
4. Which of the following are associated with e. All of the above
deterioration of MEP muscle responses dur- 6. When MEP responses are lost during surgery,
ing surgery? the most frequent rescue technique is
a. Inhalational anesthesia a. Change to volatile anesthesia
b. Hypotension b. Decrease blood pressure
c. Anemia c. There is no effective therapy
d. Administration of muscle relaxant d. Increase BP to preoperative values or higher
e. All of the above e. None of the above
5. Compared to SSEP
a. MEP has the same vascular supply in the Answers
spinal cord 1. c
b. MEP has more synapses in the spinal cord 2. c
than SSEP 3. c
c. MEP is supplied by the posterior spinal 4. e
artery while the SSEP is supplied by the 5. b
anterior spinal artery 6. d
 Auditory-Evoked Potentials
3
Christoph N. Seubert and Mary Herman

Key Learning Points • Brainstem pathways of the auditory system run

predominantly contralateral to the stimulated ear
• Auditory-evoked potentials are most useful to • Brainstem auditory potentials are very resis-
monitor the integrity of the intracranial ­auditory tant to the effects of anesthetic agents
nerve (cochlear portion of cranial nerve 8)
• The electrocochleogram (ECochG) can provide
independent verification of stimulus delivery
• Waves I and V are most robust on the AEP; Anatomy of the Auditory System
wave I originates from the cochlea, which is
typically not directly in harm’s way. Wave V Sound signals are processed by the auditory sys-
originates at the level of the inferior colliculus, tem in a sequential manner. First, the acoustic
medial geniculate body energy of sound is conducted to the cochlea
located within the inner ear, where conversion to a
coded electrochemical signal takes place. This sig-
nal is then transmitted along the auditory pathway
via the eighth cranial nerve, brainstem, and mid-
brain to the primary auditory cortex. Various audi-
Electronic supplementary material: The online version tory-evoked potentials can be recorded from all
of this chapter (doi:10.1007/978-3-319-46542-5_3) these elements (Fig. 3.1, see also supplemental
contains supplementary material, which is available to
authorized users. electronic content (Video 3.1)). The tracings con-
sist of waves with peaks and troughs that corre-
C.N. Seubert, M.D., Ph.D., D.A.B.N.M. (*)
Department of Anesthesiology, University of Florida spond to fluctuations in electrical potential. Each
College of Medicine, PO Box 100254 JHMHSC, wave is described as being either a peak (P = posi-
Gainesville, FL 32610-­0254, USA tive deflection) or trough (N = negative deflection)
e-mail: [email protected] and is further described in terms of both amplitude
M. Herman, M.D., Ph.D. (peak-to-peak height) and latency (time from stim-
Department of Anesthesiology, The Geisinger Health ulus). The waves can be divided on the basis of the
System, GMC Anesthesiology, 100 North Academy
Ave., Danville, PA 17822, USA time between acoustic stimulus and evoked
e-mail: [email protected] response into short-, mid-, and long-latency

© Springer International Publishing AG 2017 35

A. Koht et al. (eds.), Monitoring the Nervous System for Anesthesiologists
and Other Health Care Professionals, DOI 10.1007/978-3-319-46542-5_3
36 C.N. Seubert and M. Herman

Fig. 3.1 Neural pathway of the auditory system and entire neural pathway. Note how the morphology of indi-
recordable potentials. Note that auditory input is transmit- vidual waves projects differently in the individual record-
ted to bilateral primary auditory cortices and ascends ing channels. In addition to Na and Pa waves reflecting
through brainstem and midbrain partially ipsilateral and activation of the primary auditory cortex, the mid-latency
partially contralateral to the side of stimulation (for details, auditory-evoked potential (MLAEP) contains a distinctive
see text). The electrocochleogram (ECochG) contains wave V of the ABR within its first 10 ms (adapted from
near-field signals from cochlea and distal auditory nerve Thornton et al. [47]; with permission). Anatomic and
(adapted from Coats [44]; with permission). The brainstem radiographic representations of the auditory pathway can
auditory-evoked potential (ABR) reflects activity in the be found in the supplemental on-line materials

a­coustic-evoked potentials. The long-latency of recording that will not be discussed is the
acoustic-evoked potentials originate in the associ- recording of compound action potentials from the
ation cortex and require cooperation and attention. auditory nerve, which requires the surgeon to place
They are abolished under anesthesia and will, an electrode intraoperatively on an anatomic struc-
therefore, not be further considered. Another type ture of interest (for a recent review see Simon [1]).
3 Auditory-Evoked Potentials 37

 onduction of Auditory Signals

C the majority of the sound energy travels via this
from Ear to Cochlea ossicular chain to the cochlea. If movement of
the tympanic membrane or ossicular chain is
The ear is subdivided into external, middle, and restricted by fluid or a disease process, a con-
inner parts. The external ear is composed of the ductive hearing deficit results and the far less
auricle that acts to collect and direct sound effective conduction of sound via bone becomes
through the external auditory meatus toward the an important input to the cochlea. Conversely,
tympanic membrane. The tympanic membrane during bone drilling, bone-conducted noise
forms the boundary between the external and may overwhelm the cochlea and lead to a tem-
middle parts of the ear. The tympanic membrane porary hearing deficit.
is covered in a very thin squamous epithelial Two muscles lie within the middle ear and act
layer externally and by a mucous membrane to prevent excessive movement of the ossicles
internally. It moves in response to air vibrations due to loud noises. The tensor tympani muscle
that pass to it through the external auditory arises from the superior surface of the cartilage
meatus. Movements of the tympanic membrane forming the auditory tube, the greater wing of the
are transmitted by three auditory ossicles, the sphenoid bone and the petrous part of the tempo-
malleous, incus, and stapes, through the middle ral bone. It is innervated by the mandibular divi-
ear to the inner ear. sion of the trigeminal nerve and inserts on the
The middle ear lies within the petrous portion handle of the malleolus. The tensor tympani pulls
of the temporal bone. The tympanic cavity lies on the handle of the malleolus, tenses the tym-
directly behind the tympanic membrane and panic membrane, and thus dampens oscillations
shares important anatomical relationships to of the tympanic membrane. The stapedius muscle
neighboring structures. Superiorly, the epitym- arises from the pyramidal eminence on the poste-
panic recess is separated from the middle cranial rior wall of the tympanic cavity. It inserts on the
fossa by a thin roof of bone, the tegmen tympani. neck of the stapes. It is innervated by a branch of
The anterior (carotid) wall separates the carotid the facial nerve. The stapedius pulls the stapes
canal from the tympanic cavity. The eustachian posteriorly and tilts the base of the stapes in the
tube projects through the anterior wall to connect oval window. This acts to tighten the stapes and
the middle ear to the nasopharynx. The floor (jug- reduce excessive movement.
ular wall) is formed by a layer of bone that sepa-
rates the tympanic cavity from the superior bulb
of the internal jugular vein. The medial or laby-  eural Components of the Auditory
N
rinthine wall separates the tympanic cavity from System and Electrical Generators
the inner ear. The middle ear also connects poste- Along the Auditory Pathway
rior and superior with the mastoid air cells
through the mastoid antrum. Cochlea: Electrocochleogram
The auditory ossicles form a chain that The cochlea converts sound waves into action
extends across the tympanic cavity from the potentials in the neurons of the cochlear nerve.
tympanic membrane to the oval window (fenes- Sound waves conducted to the oval window
tra vestibuli). The malleous (hammer) is propagate in the perilymph of the cochlea. The
attached to the tympanic membrane. Its supe- action of these waves on the spiral organ of Corti
rior head lies within the epitympanic recess, generates excitatory synaptic input from the
and its handle is embedded in the tympanic cochlear hair cells, which in turn depolarizes the
membrane. Movement of the tympanic mem- cochlear end of the auditory nerve. This depolar-
brane results in movement of the malleous. The ization leads to the generation of the eighth nerve
head of the malleous articulates with the incus compound action potential.
(anvil). The long process of the incus articu- The electrical activity within the cochlea can
lates with the stapes (stirrup). The base of the be recorded in the form of an electrocochleogram
stapes is attached to the oval window. Typically, (ECochG; Fig. 3.1). The ECochG includes the
38 C.N. Seubert and M. Herman

cochlear microphonic, the summation potential, of stimulation, but mostly cross over to the con-
and the eighth nerve compound action potential. tralateral side (see Figs. 3.1 and 3.2).
The hair cells generate the cochlear microphonic First-order auditory neurons run in the
and summation potential within the cochlea (for cochlear division of CN VIII from the spiral
details see section “Electrocochleogram” below). organ of Corti to the dorsal and ventral cochlear
The eighth nerve compound action potential nuclei in the upper medulla. Myelinated den-
results from depolarization within the distal drites of the auditory nerve pass into and through
(cochlear) portion of the auditory nerve axons. It the spiral ganglia and form a nerve bundle in the
generates wave N1 of the ECochG. It is recorded internal auditory canal. Both the acoustic and
as a phase negativity in the middle ear or extra- vestibular portions of the auditory nerve pass
tympanic recording site. Sounds used to elicit through the temporal bone alongside the intracra-
ECochGs may produce more than one volley of nial portion of the facial nerve. Together, they
action potentials within the auditory nerve, thus exit the internal auditory canal and travel to the
producing N1 and N2 (and sometimes N3) com- brainstem. At the point of exit from the internal
ponents of an eighth nerve compound action auditory canal, both facial and vestibulocochlear
potential. The N1 potential corresponds to wave I nerves make an acute turn from the anteromedial
of the brain stem auditory response discussed trajectory of the internal auditory canal within
below. the petrous pyramid of the temporal bone pos-
terolaterally toward the cerebellopontine angle of
the brainstem. This acute turn “anchors” the ves-
 uditory Pathway from Cochlear
A tibulocochlear nerve and puts it at risk of a
Nerve to Midbrain: Auditory stretch-induced neurapraxic injury during retrac-
Brainstem-Evoked Responses tion of the brainstem, especially if the anatomy of
the posterior fossa is already disrupted by pathol-
Neural transmission of auditory signals starts ogy such as a cerebellopontine angle tumor.
with input from cochlear hair cells into the distal Auditory nerve fibers synapse at either the
auditory nerve, whose anatomic course puts it at posterior ventral cochlear nucleus or the anterior
risk of injury during many procedures in the pos- ventral cochlear nucleus. Fibers that synapse at
terior cranial fossa. Once the signals reach the the posterior ventral cochlear nucleus also have
brainstem, they pass through a complex series of connections with the dorsal cochlear nucleus.
relay and processing stations to the midbrain. From the cochlear nuclei, second-order neurons
The signals travel partially ipsilateral to the side may follow several pathways or combinations of

Fig. 3.2 (continued) The lateral lemniscus ascends medial to nal mesencephalic and central tegmental tracts near the mid-
the intrapontine segment of the trigeminal nerve and lateral to line, decussation of the superior cerebellar peduncle at the
the medial lemniscus and superior cerebellar peduncle to level of the inferior colliculus and the red nucleus located
reach the inferior colliculus. (d) After reaching the inferior between the mid-level of the inferior colliculus, and the lateral
colliculus, auditory information is carried to the medial genic- wall of the third ventricle. (g) Left retrosigmoid view. The left
ulate body by the brachium of the inferior colliculus, which cerebellar hemisphere was removed to expose the dorsolateral
ascends obliquely on the lateral surface of the midbrain. After brainstem and the ventral and dorsal cochlear nuclei, lateral
reaching the medial geniculate body, the auditory pathway lemniscus, and inferior colliculus. Aud. auditory, Cer. cerebel-
crosses beneath the lentiform nucleus in the sublentiform lar, CN cranial nerve, Coch. cochlear, Coll. colliculus, collicu-
pathway to reach the auditory cortex on the most anterior lar; CTT central tegmental tract, Decuss. decussation, Dent.
transverse temporal gyrus, referred to as Heschl’s gyrus. (e) dentate, Dors. dorsal, Flocc. flocculus, Gen. geniculate, Gl.
Posterior view of the midbrain. The nuclei of the superior and gland, Gyr. gyrus, Inf. inferior, Lat. lateral, Lemn. lemniscus,
inferior colliculi are located below the surface. The red Med. medial, Nucl. nucleus, Oliv. olivary, Ped. peduncle,
nucleus is located at a deeper level. (f) Further dissection. Pontomed. pontomedullary, Pontomes. pontomesencephalic,
Structures near the inferior collicular implant, in order from Rad. radiations, Sulc. sulcus, Sup. superior, Tent. tentorium,
dorsal to ventral, are the oculomotor and trochlear nuclei TMT trigeminal mesencephalic tract, Trap. trapezoid, TST
located just ventral to the aqueduct in the midline, the trigemi- trigeminal spinal tract, Vent. ventral
3 Auditory-Evoked Potentials 39

Fig. 3.2 Fiber dissection of the central auditory pathway. (a) facial nucleus, and nucleus ambiguus, which are located ven-
Ventral and dorsal cochlear nuclei. Posterolateral view of the tromedial to the trigeminal spinal tract. The facial nucleus is
junction of the cranial nerve with the brainstem. The ventral hidden deep to the cochlear nuclei. (b) Anterior view. The
cochlear nucleus is situated on the lateral and dorsal cochlear ventral pons was removed to expose the medial and lateral
nucleus on the dorsal surface of the inferior cerebellar pedun- lemnisci and the trapezoid body formed by crossing auditory
cle. They are positioned close to the trigeminal spinal tract, fibers at the level of the lower pons. (c) Left lateral view.
40 C.N. Seubert and M. Herman

Table 3.1 Purported neural generators of ABR peaksa researchers have postulated that each peak cor-
Peak Generator responds to one generator, it appears that most
I Acoustic nerve (extracranial) ABRs are a result of the summation of inputs
II Acoustic nerve (intracranial), cochlear nuclei from multiple generators [2–4]. The pattern of
III Cochlear nuclei connections in the auditory system contributes to
IV Lateral lemniscus, superior olivary complex this complexity, as ascending fibers both cross
V Inferior colliculus; contralateral lateral and bypass relay nuclei [4–6]. Nonetheless, the
lemniscus
information in Figs. 3.1 and 3.2 and Table 3.1 can
VI Medial geniculate nuclei
be used to help localize the site of an injury.
VII Thalamocortical radiations
a
While functional deficits can often be localized
Peaks I, III, and V are considered the most useful for
intraoperative neuromonitoring of ABRs. Most peaks are
when injury or ischemia occurs, the complexity
likely a result of the summation of inputs from multiple of the system may sometimes lead to changes in
generators. While not all of these generators have been ABRs with no change in function [7, 8].
proven, the designations are clinically useful because they Wave I of the ABR arises from action poten-
point out the approximate location of an injury. ABR
brain-stem auditory-evoked potentials
tials in the most distal portion of the myelinated
auditory nerve [9]. Wave I of the ABR is
­equivalent to N1 of the ECochG [10]. Wave I is a
pathways en route to the inferior colliculus. Most near field potential, recorded in the vicinity of the
fibers decussate via the trapezoid body and pass ipsilateral stimulated ear. It represents the periph-
up the lateral lemniscus to the contralateral infe- eral potential of this sensory modality. Loss of
rior colliculus of the midbrain. Some fibers syn- wave I may indicate damage to the inner ear, but
apse in either the medial or lateral superior olivary may also be caused by technical problems in the
nuclei. Others pass through the ipsilateral lateral delivery of acoustic stimuli to the ipsilateral ear.
lemniscus to the ipsilateral inferior colliculus. All When wave I is absent, ABRs cannot be used to
ascending fibers synapse at the inferior colliculus. make inferences about the integrity of the brain-
Third-order neurons from the inferior colliculus stem. Wave II of the ABR occurs at approxi-
ascend to the medial geniculate body at the level mately the same latency as N1 of the compound
of the thalamus on each side. Fourth-­order neu- action potential of the proximal auditory nerve to
rons pass through the internal capsule and then the cochlear nucleus. It occurs on the ipsilateral
form the auditory radiation to the primary audi- side. Wave III predominately originates in the
tory cortex. This complex pattern of tracts and caudal pontine tegmentum and region of the
relay stations is involved in elementary process- superior olivary complex. Near-field activity in
ing of auditory input, e.g., by extracting direc- the ipsilateral cochlear nucleus corresponds with
tional information about the source of sounds or wave III [11]. Ascending projections are bilat-
as the afferent limb of auditory startle responses. eral, so wave III may receive input from both the
Activity in the neural pathways from the ipsilateral and contralateral ear. Scalp-recorded
cochlea up to the midbrain can be recorded in the wave III has been recorded at the same time as
form of auditory brainstem responses (ABRs, near-field activity in the cochlear nucleus [12].
sometimes also called brainstem auditory-evoked Other recordings from the area of the cochlear
responses [BAERs] or brainstem auditory- nucleus in the lateral recess of the fourth ventri-
evoked potentials [BAEPs]; see Fig. 3.1). Peaks cle indicate activity that coincides with wave IIIn
in a recording of ABRs are labeled I through (the negative peak between III and IV) [13, 14].
VII. As with other sensory-evoked potentials, the The auditory nerve may continue to be active
wave amplitude, absolute latencies, and inter- during generation of the scalp-recorded waves III
peak latencies are evaluated to assess the integ- and IV [4]. Waves IV and V are often fused into a
rity of the auditory system. The purported IV–V complex and their generators appear to be
generators for these peaks are shown in Fig. 3.1 in close proximity of each other. Wave IV appears
and are summarized in Table 3.1. Although some to reflect activity in ascending auditory fibers
3 Auditory-Evoked Potentials 41

within the dorsal and rostral pons, caudal to the input to memory and visual association areas.
inferior colliculus with input from both ipsilat- Auditory information passes between the two
eral and contralateral sides. Wave V appears to hemispheres through the corpus callosum, the
predominantly reflect activity at the level of the primary ­connection between the left and right
inferior colliculus, perhaps including activity in hemispheres. The transcollosal auditory path-
the rostral portion of the lateral lemniscus as well way begins at the auditory cortex and passes
as activity in the contralateral lateral lemniscus posteriorly and superiorly around the lateral
as it terminates on the inferior colliculus [4, 11]. ventricles.
Waves VI and VII are inconsistent and variable; Electrical phenomena associated with activa-
therefore, they are not routinely monitored [15]. tion of auditory cortex can be recorded in the
Most intraoperative neuromonitoring utilizes form of mid-latency auditory-evoked potentials
only waves I, III, and V to guide the intraopera- (MLAEPs, see Fig. 3.1). MLAEPs are observed
tive course [6, 16, 17]. 10–60 ms after an auditory stimulus [19]. They
appear to arise from the ventral portion of the
medial geniculate body and primary auditory
 rimary Auditory Cortex: Mid-­
P cortex in the primary pathway and also nonpri-
Latency Auditory-Evoked Potentials mary pathways in the auditory thalamocortical
projection [20, 21].
Tracts carrying auditory information project from MLAEPs consist of four deflections, labeled
the medial geniculate body to the cortex and Na, Pa, Nb, and Pb. Na and Pa latencies are
other brain areas by multiple routes [18]. The between 10–25 and 22–40 ms, respectively [19].
medial geniculate body and cortex are linked by Nb latency is at 40 ms and Pb is at 40–60 ms.
two main routes. The first pathway from the ven- Magnetoencephalographic and intracerebral
tral medial geniculate body carries only auditory recordings suggest that the Na/Pa complex is
input and follows a sublenticular route through generated in the posteromedial part of the first
the internal capsule to Heschl’s gyrus in the supe- transverse gyrus. MLAEPs correlate well with
rior temporal lobe. The second pathway from the wakefulness during general anesthesia when
medial geniculate body projects into the inferior using desflurane or propofol [22] and are associ-
portion of the internal capsule and carries mixed ated with awakening from anesthesia following
auditory, somatic, and possibly visual sensory verbal command [23]. MLAEPs may be abnor-
input. Auditory fibers from the medial geniculate mal in neurologic diseases such as dementia,
body also project to the caudate nucleus, the Parkinson’s disease, multiple sclerosis, and myo-
putamen, and the globus pallidus. tonic dystrophy [24–32].
Intrahemispheric and interhemispheric con-
nections occur within the primary auditory cor-  ascular Supply of Auditory Pathway
V
tex. Multisynaptic pathways likely exist in the Structures
middle and posterior areas of the superior tem- The cochlea receives its blood supply from the
poral gyrus. Fibers also extend from the superior internal auditory artery, which is usually a branch
temporal gyrus to the insula and frontal opercu- of the anterior inferior cerebellar artery. The inter-
lum. The arcuate fasciculus provides auditory nal auditory artery is quite small in diameter and
input from the auditory cortical areas in the tem- passes through the internal auditory canal along
poral lobes to the frontal lobes. Wernicke’s area with the eighth nerve [33]. Damage to this artery
in the temporal lobe and Broca’s area in the will cause cochlear ischemia or infarction.
frontal lobe receive auditory information via the Cochlear ischemia from obstruction or disruption
arcuate fasciculus. Auditory input also passes to of the internal auditory artery may affect the
the hippocampus and occipital regions of the ECochG and wave I of the ABRs resulting in the
brain. Although these areas and pathways are loss of all subsequent components [34]. This may
not ­anatomically defined, they provide auditory occur during tumor resection and lead to postop-
42 C.N. Seubert and M. Herman

erative deafness [35]. These effects may be revers- location of the lesion. Wave III would be lost if
ible if perfusion is restored within 15 min [35]. the lesion is caudal to or at the superior olivary
The brainstem (medulla, pons, and midbrain) complex. However, wave I would remain intact.
receives the bulk of its blood supply from the Interruption of the blood supply proximal to the
vertebrobasilar system [36]. Except for the laby- internal auditory artery may affect wave I. For
rinthine branch and early branches of the verte- example, during posterior fossa vascular surgery,
bral arteries, all other branches supply the damage to the vertebrobasilar system, which sup-
brainstem and medulla. In principle, conducting plies the anterior inferior cerebellar artery and
vessels run along the brainstem surface, whereas the internal auditory artery, could cause ischemic
the nuclei within the brainstem and the fiber cochlear damage and loss of all waveforms.
tracts are supplied by perforating vessels. The
vertebral arteries supply the medulla. The para-
median branches of the basilar artery supply  echniques for Recording Auditory-­
T
medial pontine structures. Short circumferential Evoked Potentials
arteries supply the anterolateral pons. Long cir-
cumferential branches of the basilar artery run Auditory-evoked potentials can be recorded from
laterally over the anterior surface of the pons and all neuronal structures that contribute to the audi-
anastomose with branches of the anterior inferior tory system [40]. The first potentials generated in
cerebellar artery. The inferior colliculus receives response to sound come from the cochlea. They
blood from the anterior inferior cerebellar artery can be recorded in the form of the ECochG.
(caudally) with some reinforcement rostrally Because the cochlea lies well protected in the
from the superior cerebellar artery. Quadrageminal temporal bone, direct damage during surgery is
arteries arise from branches of the basilar artery typically either not a concern or planned as part
and also supply the inferior colliculus. of the surgical access, such as in the translabyrin-
The medial geniculate nucleus lies in the dor- thine approach to the posterior fossa. Therefore,
sal thalamus and receives its blood supply from monitoring of the ECochG is not widely used.
posterolateral arteries (thalamogeniculate), From the cochlea, potentials are carried along the
which arise from the posterior cerebral artery. auditory nerve and the brainstem to the primary
The primary auditory cortex, which lies in the auditory cortex and further to association areas of
superior temporal lobe, is supplied by branches the cortex. MLAEPs reflect activation of the pri-
of the middle cerebral artery and, therefore, by mary auditory cortex occurring 10–50 ms after
the anterior cerebral circulation. Interhemispheric acoustic stimulation. MLAEPs are also sensitive
fibers that connect the left and right auditory cor- to the effects of general anesthetics and are there-
tex pass through the posterior corpus callosum, fore not used for intraoperative monitoring of the
which receives its blood supply from the perical- integrity of the auditory pathway. On the con-
losal artery, a branch of the anterior cerebral trary, because of their sensitivity to anesthetics,
artery [18, 37]. they have been used to monitor cortical anes-
ABRs may change during posterior fossa sur- thetic drug effect in order to help assess “anes-
gery as a result of ischemia or infarction from thetic depth.” Their performance as a monitor of
clipping or compression of arteries that perfuse anesthetic depth is comparable to that of other
the brainstem auditory pathways [8]. Patients monitors relying on the processed electroenceph-
who experience major changes in ABRs that per- alogram (EEG) [41]. Recording MLAEPs
sist to the end of the operation almost always requires both stimulation and recording and is
have new postoperative neurologic deficits [38, therefore technically more elaborate than the set-
39]. Changes in waveforms almost always reflect ­up for processed frontal EEG. This fact has ham-
anatomical deficits. For example, damage below pered commercial exploitation and clinical
the mesencephalon will affect wave V, and wave acceptance of MLAEPs as a monitoring tech-
III may or may not be spared depending on the nique during anesthesia.
3 Auditory-Evoked Potentials 43

Short-latency potentials occur less than 10 ms tions of hearing thresholds. A stimulus intensity
after an acoustic stimulus and originate in the of 70 dB above normal hearing level (70 dB nHL)
acoustic nerve and brainstem. They are typically normally yields maximal auditory responses. In
referred to as auditory brainstem responses the absence of a preoperative audiogram,
(ABRs); sometimes also described as BAERs or 90–95 dB is frequently used, particularly in the
BAEPs. An anesthesiologist is most likely to presence of a preoperative hearing deficit. Note
encounter intraoperative use of auditory-evoked that a decrease in the stimulus intensity as it is
potentials in the form of ABRs. delivered to the cochlea reduces the amplitude of
The recording of auditory-evoked responses the auditory-evoked potentials (see Fig. 3.3).
presents significant technical challenges because Such a reduction can result from a dislodged stim-
the signals originate from anatomical structures ulator, fluid in the middle ear (e.g., from breached
that are far removed from the site of electrode mastoid air cells), accumulation of nitrous oxide
placement on the head’s surface. Because of this in the face of a blocked eustachian tube, or dam-
distance between purported anatomical generator age to the auditory pathway.
and recording electrode, these types of responses Depending on the structures at risk during sur-
are called far-field responses. Their amplitude is gery, stimuli are typically presented unilaterally
small, typically less than 0.5 μV, compared to the even though rostral parts of the auditory pathway
EEG and the electrocardiogram, which are a 100 proceed largely crossed over to the contralateral
and a 1000 times larger, respectively. Because of side, but also to a lesser extent uncrossed (see
their small amplitude, auditory-evoked potentials above). Unilateral presentation of stimuli allows
cannot be seen on continuous recordings of electri- diagnosis of lesions to the cochlea and distal
cal activity. Instead they require signal averaging auditory nerve ipsilateral to the side of stimula-
of the responses to 500–2000 acoustic stimuli. tion. If unilateral stimulation is used, bone con-
duction to the contralateral ear is typically
“blocked” by masking. Masking refers to the
Stimulation continuous administration of white noise to the
nonstimulated ear typically at intensities 30 dB
Acoustic stimuli are presented intraoperatively as less than the click stimulus. An alternative way of
“clicks” of 100-μs duration. The clicks comprise stimulation is the use of interleaved stimuli alter-
a broad spectrum of tone frequencies and thus nating between right and left ear, but sorted into
activate much of the cochlea. This nearly simul- separate averaged recordings for left and right ear
taneous activation of the cochlea triggers a syn- stimulation. Such interleaved stimulation does
chronized volley of action potentials in the not allow for masking, but the small decrement in
acoustic nerve, which in turn can be recorded as sensitivity is made up by the fact that both sides
well-defined peaks in the auditory-evoked are monitored continuously. Bilateral stimulation
response. Clicks can be delivered in three differ- is sometimes used to record MLAEPs.
ent “polarities”—rarefaction, condensation, or Because auditory responses up to the level of
alternating (Fig. 3.3). The description of the click the brainstem occur within 10 ms of stimulation,
polarity refers to the initial movement of the tym- stimuli can be presented as frequently as 30–50
panic membrane away from, toward or alternat- times per second (30–50 Hz). If there is a preex-
ing between away and toward the stimulator. In isting hearing deficit such as that caused by a
practice, either an alternating polarity is used in large acoustic neuroma, slower stimulation at
order to cancel out the stimulus artifact or the 10–15 Hz may be necessary. Stimulus rates
polarity that results in the clearest response. In should never be equal to or a divisor of the line
rare instances bone conduction can be used to frequency of 60 Hz, because signal averaging
deliver the acoustic stimulus. will then tend to augment the electromagnetic
The intensity/volume of the click stimulus can interference from the line frequency rather than
be based on the results of preoperative determina- canceling it out.
44 C.N. Seubert and M. Herman

Fig. 3.3 Effect of click

polarity and stimulus intensity
on brainstem auditory-evoked
potentials. The top panel
shows ABRs recorded in
response to three different
click polarities. Note that click
polarity has notable effects on
the early ABR in the A2-Cz
channel, because it contains
information from the ECochG.
The bottom panel shows the
effect of a stepwise decrease in
stimulus intensity on the ABR.
Note that wave I is lost at
stimulus intensities less than
80 dB, suggesting a problem
with sound delivery, whereas
the desynchronization of
waves III and V at low
stimulus intensities are
indistinguishable from changes
caused by damage to the
auditory system

Physically, the stimuli can be presented either Electrocochleogram

with earphones or with foam ear inserts con-
nected by a tube to stimulators placed at a short Recordings of the ECochG require placement of
distance (i.e., <10 cm) from the ear. Earphones a primary electrode close to the cochlea. During
are used less frequently, because they put a middle ear surgery, such an electrode can be
source of electromagnetic interference close to placed on the promontory or the round window.
the generators of the auditory response. Ear A noninvasive recording is possible from the
inserts are less bulky and do not contribute to external ear canal, which is preferred over a more
noise. The acoustic transmission of sound from distal mastoid electrode [42]. The secondary or
the stimulator through the tube insert to the tym- reference electrode can be at the contralateral ear
panic membrane delays auditory responses by or at Cz. The filter bandpass is set to 5000–
less than 1 ms. 3000 Hz. Stimulation parameters are typically
3 Auditory-Evoked Potentials 45

Fig. 3.4 ECochG and effects of anesthesia on brainstem and addition emphasizes the summating potential. Anesthesia
MLAEPs. The ECochG contains a prominent wave N1 with halothane differentially affects ABRs and MLAEPs.
that coincides with activation of the distal auditory nerve. Whereas the latency of wave V of the ABR increases by
N1 is larger than a typical ABR and therefore easier to less than 1 ms and the amplitude is unaffected (left panel),
record than wave I of the ABR. Electrical activity within the mid-latency response nearly vanishes at high concen-
the cochlea is reflected in the cochlear microphonic and trations of halothane. (Right panel adapted from Coats
summating potential. Subtraction of ECochG traces in [44]; with permission. Right panel bottom adapted from
response to rarefaction and condensation clicks empha- Thornton et al. [47]; with permission.)
sizes the cochlear microphonic (middle panel), whereas
46 C.N. Seubert and M. Herman

the same as those described above, even though containing seven waves (see Fig. 3.1), a mini-
the two cochlear potentials depend on stimulus mum of waves I, III, and V should be present for
duration and are more pronounced, when longer the signal to be useful for intraoperative moni-
stimuli are used [43]. Typical timebases used for toring. A typical montage includes electrodes on
ECochGs are less than 10 ms. the ipsilateral ear and one at the vertex, i.e.,
Three potentials characterize the ECochG A1-Cz and A2-Cz for left and right sides, respec-
(Fig. 3.3). In sequence of activation, they are the tively. Additional channels can be used to aid in
cochlear microphonic, the summating potential, the identification of individual peaks (see
and the N1 potential. Based on the sequence of Fig. 3.1). Specifically, a cervical midline elec-
steps that translate auditory input into nerve trode referenced to Cz sometimes aids in identi-
impulses, the cochlear microphonic and summat- fication of wave V, whereas an A1-A2 or A2-A1
ing potential originate in the hair cells of the organ channel can aid in identification of wave I if the
of Corti and the N1 potential originates in the dis- ECochG is not monitored separately. With the
tal auditory nerve. The cochlear microphonic is an exception of wave I, all potentials recorded as
AC voltage that mirrors the waveform of the ABRs originate from structures deep within the
acoustic stimulus. Therefore, it can be minimized head and are considered far-field potentials.
by stimulating with alternating polarity and can Therefore, wave I shows up best in channels that
be augmented by subtraction of traces recorded contain an electrode near the ipsilateral ear.
with alternating polarity (Fig. 3.3) [44]. In con- Subsequent waves, in contrast, can be activated
trast, the summating potential is a DC current by stimulation from either side and contain lim-
thought to reflect the fact that transduction of the ited information that allows assignment of
stimulus by the hair cells does not occur uni- changes to either the right or left side. In a base-
formly and at the same time throughout the line recording, it is important to clearly identify
cochlea. Therefore, the summating potential is wave I and compare it with that from the contra-
increased in patients with inner ear diseases such lateral ear. Clear identification of wave I guards
as Meniere’s disease that further distort the trans- against situations when the stimulators for the
duction of acoustic stimuli in the inner ear [44]. right and left sides have mistakenly been
With the short clicks typically used for stimula- reversed. Furthermore, the presence of wave I
tion, the summating potential is reflected as a assures delivery of an adequate stimulus and
“shoulder” on the much larger N1 potential. The thus allows identification of unilateral damage to
final potential recorded in the ECochG is the N1 the ipsilateral auditory nerve.
potential, which reflects activation of the distal The typical epoch for ABRs is 10 ms although
auditory nerve and thus the same physiologic phe- the stimulus delay imposed by ear insert stimula-
nomenon as wave I of the ABR. Because it is a tors and the presence of hearing loss sometimes
near-field potential, it is large in amplitude and requires a longer epoch. The bandpass is set from
requires fewer averages than a full ABR. Those 100 or 150 to 3000 Hz and a notch filter for the
laboratories that use ECochG for intraoperative 60-Hz line frequency can be added.
monitoring typically focus on the N1 potential to
quickly ascertain successful stimulation of the
auditory system similar to the more familiar Erb’s MLAEPs
point recording used with SSEPs.
Auditory-evoked potentials beyond the brain-
stem are recorded either for research purposes or
Brainstem Auditory-Evoked to measure anesthetic drug effect [45, 46].
Potentials Examples of late potentials include the event-­
related potential P300, which occurs about
A normal ABR recording should show at least 300 ms after an appropriate stimulus or the mis-
three clearly identifiable waves/peaks. Although match negativity that occurs late after an oddball
an ABR recording is classically described as sequence of stimuli. Both reflect elements of
3 Auditory-Evoked Potentials 47

higher order processing and are absent under accumulation in the middle ear. Conduction in
anesthesia. Latency and amplitude of early peaks the middle ear diminishes when fluid enters the
of the MLAEP correlate with anesthetic drug middle ear either in the form of irrigation fluid,
effect (see Fig. 3.3) [47]. Two commercial moni- e.g., during drilling in the mastoid bone, or as
tors of anesthetic drug effect have been devel- blood from any of the anatomic structures in the
oped based on MLAEP technology, although vicinity. Accumulation of nitrous oxide in the
neither is widely used [48]. Stimulation for face of a blocked eustachian tube can also
MLAEPs can be done as described above, decrease conduction in the middle ear. Note that
although stimulus durations up to 500 μs and the changes caused by diminishing acoustic input
simultaneous stimulation of both ears are look very similar to progressive damage of the
described. The typical band pass is set to auditory pathway if changes to wave I are not
15–250 Hz. The stimulation rate needs to be less assessed (see Fig. 3.2). Finally, drilling of bone
than 10 Hz because the epoch is at least 50 ms. causes noise of an intensity that overwhelms the
The montage can be mastoid—Cz, especially cochlea, prevents recording of ABRs during drill-
under anesthesia, but can also be a midline mon- ing, and alters ABRs recorded shortly after the
tage of a cervical electrode referenced to Cz or cessation of drilling.
Fz. The benefit of using a midline montage is that Physiologic factors that affect the entire ABR
it avoids recording the postauricular muscle are interruption or vasospasm of the cochlear
response [45]. The postauricular muscle response artery or avulsion of fascicles of the distal
is an involuntary muscle reflex in response to ­auditory nerve within the inner auditory canal.
loud acoustic stimuli. Its amplitude may exceed Both diminish or abolish wave I and all subse-
that of the MLAEP and its latency of 15–20 ms quent waves of the ABR and result in diminished
coincides with the early peaks of the hearing or deafness, respectively.
MLAEP. Furthermore, because it is triggered by The intracranial portion of the auditory nerve
the acoustic stimulus, signal averaging will not can be affected by traction on either nerve or
remove it. brainstem resulting in an increase in the latency
between waves I and III. This change occurs only
ipsilateral to the side of the injury. The degree of
 nesthetic and Physiologic
A desynchronization of wave III reflects the sever-
Considerations for Monitoring ity of the insult and frequently the rate of change
of Auditory Brainstem Responses of the potential is inversely related to reversibil-
ity, i.e., a signal that changes profoundly and rap-
Auditory brainstem responses are very resistant idly is less likely to recover [49]. Similar changes
to the effects of general anesthetic agents (see can be caused by cold irrigation or heat from the
Fig. 3.3) [47]. Therefore, no modification of the cautery or drying of the auditory nerve [50, 51].
anesthetic approach to a patient is needed because Application of papaverine to relieve vasospasm
of ABR monitoring. The small increases in [52] or aggressive attempts to fill the subarach-
latency caused by anesthetic agents are not clini- noid space with irrigation fluid prior to dural clo-
cally significant and are easily distinguished sure may lead to ABR changes [53].
from changes in ABRs caused by technical and Damage to the brainstem either in the form of
physiologic factors. direct trauma or through compromise of blood
Sometimes technical problems cause gradual supply or blood flow will be reflected in ABRs to
or abrupt changes in ABRs intraoperatively in the the extent that the auditory pathway is involved.
absence of physical damage to the auditory path- While persistent ABR changes nearly always
way (see also Chap. 29, “ENT and Anterior Neck predict brainstem dysfunction, damage to the
Surgery”, Tables 29.2 and 29.3). Diminished brainstem may still occur even though ABRs
input can occur abruptly, e.g., by kinking the tube remain unchanged. Many centers use ABRs
of an ear insert on the down ear in a lateral or together with other modalities such as
park bench position, or gradually, e.g., by fluid somatosensory-­ evoked potentials (SSEPs) and
48 C.N. Seubert and M. Herman

motor-evoked potentials (MEPs) to monitor the 12. *Møller AR, Jannetta PJ. Monitoring auditory func-
tions during cranial nerve microvascular decompres-
integrity of the brainstem. Again persistent
sion operations by direct recording from the eighth
changes in monitored signals typically predict nerve. J Neurosurg. 1983;59:493–9.
new postoperative deficits, but unchanged sig- 13. Møller AR, Jannetta PJ, Jho HD. Click-evoked
nals do not rule out the potential for injury to the response from the cochlear nucleus: a study in
humans. Electroencephalogr Clin Neurophysiol.
brainstem. That is because the monitored path-
1994;92:215–24.
ways only cover a small part of the cross-­sectional 14. *Møller AR, Jho HD, Yokota M, Jannetta PJ.
area of the brainstem. Thus, multimodality moni- Contribution from crossed and uncrossed brainstem
toring has good specificity, but limited sensitivity structures to the brainstem auditory evoked potentials
(BAEP): a study in humans. Laryngoscope. 1995;105:
for assessing brainstem integrity.
596–605.
15. Chiappa KH, Roppper AH. Evoked potentials in clini-
cal medicine (first of two parts). N Engl J Med.
References1 1982;306:1205–11.
16. Duncan PG, Sanders RA, McCullough DW.
Preservation of auditory-evoked brainstem responses
1. *Simon MV. Neurophysiologic intraoperative moni-
in anaesthetized children. Can Anaesth Soc
toring of the vestibulocochlear nerve. J Clin
J. 1979;26:492–5.
Neurophysiol. 2011;28:566–81.
17. Raudzens PA, Shetter AG. Intraoperative monitoring
2. Jewitt DL, Willliston JS. Auditory-evoked far fields
of brain-stem auditory evoked potentials. J Neurosurg.
averaged from the scalp of humans. Brain.
1982;57:341–8.
1971;94:681–96.
18. Musiek FE, Weihing JA, Oxholm VB. Anatomy and
3. Picton TW, Hillyard SA, Krausz HI, Galambos
physiology of the central auditory nervous system: a
R. Human auditory evoked potentials. I. Evaluation of
clinical perspective. In: Roeser RJ, Valente M,
components. Electroencephalogr Clin Neurophysiol.
Hosford-Dunn H, editors. Audiology diagnosis, vol.
1974;36:179–90.
2. New York: Thieme Medical; 2007. p. 50–6.
4. Møller AR. Neural generators for auditory brainstem
19. Brunner MD, Umo-Etuk J, Sharpe RM, Thornton
evoked potentials. In: Burkard RF, Eggemont JJ,
C. Effect of a bolus dose of midazolam on the audi-
Manuel D, editors. Auditory evoked potentials: basic
tory evoked response in humans. Br J Anaesth.
principles and clinical applications. Baltimore:
1999;82:633–4.
Lippincott Williams & Wilkins; 2007. p. 336–54.
20. Deiber MP, Ibanez V, Fischer C, Perrin F, Mauguiere
5. Strominger NL, Nelson LR, Dougherty WJ. Second
F. Sequential mapping favours the hypothesis of dif-
order auditory pathways in the chimpanzee. J Comp
ferent generators for Na and Pa middle latency audi-
Neurol. 1977;172:349–66.
tory evoked potentials. Electroencephalogr Clin
6. Grundy BL, Jannetta PJ, Procopio PT, Lina A, Boston
Neurophysiol. 1988;71:187–97.
JR, Doyle E. Intraoperative monitoring of brain-stem
21. Thornton RM, Sharpe RM. Evoked responses in
auditory evoked potentials. J Neurosurg. 1982;57:
anaesthesia. Br J Anaesth. 1998;81:771–81.
674–81.
22. Dutton RC, Smith WD, Rampil IJ, Chortkoff BS,
7. Friedman WA, Kaplan BJ, Gravenstein D, Rhoton Jr
Eger II EI. Forty-hertz midlatency auditory evoked
AL. Intraoperative brain-stem auditory evoked poten-
potential activity predicts wakeful response during
tials during posterior fossa microvascular decompres-
desflurane and propofol anesthesia in volunteers.
sion. J Neurosurg. 1985;62:552–7.
Anesthesiology. 1999;91:1209–20.
8. *Legatt AD. Mechanisms of intraoperative brainstem
23. Goto T, Nakata Y, Saito H, Ishiguro Y, Niimi Y, Morita
auditory evoked potential changes. J Clin
S. The midlatency auditory evoked potentials predict
Neurophysiol. 2002;19:396–408.
responsiveness to verbal commands in patients emerg-
9. Legatt AD, Arezzo JC, Vaughn Jr HG. The anatomic
ing from anesthesia with xenon, isoflurane, and sevo-
and physiologic bases of brainstem auditory evoked
flurane, but not with nitrous oxide. Anesthesiology.
potentials. Neurol Clin. 1988;6:681–704.
2001;94:782–9.
10. Gersdorff MCH. Simultaneous recordings of human
24. Kileny P, Dobson D, Gelfand ET. Middle-latency
auditory potentials: transtympanic electrocochleagra-
auditory evoked responses during open-heart surgery
phy (ECoG) and brainstem-evoked responses (BER).
with hypothermia. Electroencephalogr Clin
Arch Otorhinolaryngol. 1982;234(1):15–20.
Neurophysiol. 1983;55:268–76.
11. Legatt AD. Brainstem auditory evoked potentials:
25. Woods DL, Clayworth CC, Knight RT. Middle latency
methodology, interpretation, and clinical application.
auditory evoked potentials following cortical and sub-
In: Aminoff MJ, editor. Electrodiagnosis in clinical
cortical lesions. Electroencephalogr Clin Neurophysiol.
neurology. New York: Churchill Livingstone; 2005.
1985;61:51.
p. 489–523.
26. Woods DL, Clayworth CC, Knight RT, Simpson GV,
Naeser MA. Generators of middle- and long-latency
1
Asterisks indicate key references. auditory evoked potentials: implications from studies
3 Auditory-Evoked Potentials 49

of patients with bitemporal lesions. Electroencephalogr terior fossa aneurysm surgery: a comparison of two
Clin Neurophysiol. 1987;68:132–48. modalities. Can J Anaesth. 1994;41:92–7.
27. Buchwald JS, Erwin RJ, Van Lancker D, Cummings 40. *Martin WH, Stecker MM. ASNM position statement:
JL. Midlatency auditory evoked responses: differen- intraoperative monitoring of auditory evoked poten-
tial abnormality of P1 in Alzheimer’s disease. tials. J Clin Monit Comput. 2008;22:75–85.
Electroencephalogr Clin Neurophysiol. 1989;74: 41. Bruhn J, Myles PS, Sneyd R, Struys MM. Depth of
378–84. anaesthesia monitoring: what’s available, what’s
28. Green JB, Flagg L, Freed DM, Schwankhaus JD. The validated and what’s next? Br J Anaesth.
middle latency auditory evoked potential may be 2006;97:85–94.
abnormal in dementia. Neurology. 1992;42:1034–6. 42. Krieg SM, Kempf L, Droese D, Rosahl SK, Meyer B,
29. Versino M, Bergamaschi R, Romani A, Banfi P, Lehmberg J. Superiority of tympanic ball electrodes
Callieco R, Citterio A, et al. Middle latency auditory over mastoid needle electrodes for intraoperative
evoked potentials improve the detection of abnormal- monitoring of hearing function. J Neurosurg.
ities along auditory pathways in multiple sclerosis 2014;120:1042–7.
patients. Electroencephalogr Clin Neurophysiol. 43. Ferraro JA. Clinical electrocochleography: overview
1992;84:296–9. of theories, techniques and applications. http://www.
30. Green JB, Elder WW, Freed DM. The P1 component audiologyonline.com/articles/pf_article_detail.
of the middle latency auditory evoked potential pre- asp?article_id = 238. Accessed 14 July 2010.
dicts a practice effect during clinical trials in 44. Coats AC. The summating potential and Menière’s
Alzheimer’s disease. Neurology. 1995;45:962–6. disease. Arch Otolaryngol. 1981;107:199–208.
31. Çelik M, Seleker FK, Sucu H, Forta H. Middle 45. Bell SL, Smith DC, Allen R, Lutman ME. Recording
latency auditory evoked potentials in patients with the middle latency response of the auditory evoked
parkinsonism. Parkinsonism Relat Disord. 2000;6: potential as a measure of depth of anaesthesia. A tech-
95–9. nical note. Br J Anaesth. 2004;92:442–5.
32. Arakawa K, Tomia H, Tobimatsuc S, Kirab J. Middle 46. Plourde G. Auditory evoked responses. Best Pract Res
latency auditory-evoked potentials in myotonic dys- Clin Anaesthesiol. 2006;20:129–39.
trophy: relation to the size of the CTG trinucleotide 47. Thornton C, Heneghan CPH, James MFM, Jones
repeat and intelligence quotient. J Neurol Sci. JG. Effects of halothane or enflurane with controlled
2003;207:31–6. ventilation on auditory evoked potentials. Br
33. Kim HN, Kim YH, Park IY, Kim GR, Chung J Anaesth. 1984;56:315–23.
IH. Variability of the surgical anatomy of the neuro- 48. Nishiyama T. Comparison of the two different audi-
vascular complex of the cerebropontine angle. Ann tory evoked potentials index monitors in propofol-­
Otol Rhinol Laryngol. 1990;99:288–96. fentanyl-­nitrous oxide anesthesia. J Clin Anesth.
34. Nadol Jr JB, Levine R, Ojemann RG, Martuza RL, 2009;21:551–4.
Montgomery WW, de Sandoval PK. Preservation of 49. Ying T, Thirumala P, Chang Y, Habeych M,
hearing in surgical removal of acoustic neuromas of Crammond D, Balzer J. Empirical factors associated
the internal auditory canal and cerebellar pontine with brainstem auditory evoked potential monitoring
angle. Laryngoscope. 1987;97:1287–94. during microvascular decompression for hemifacial
35. Levine RA, Ronner SF, Ojemann RG. Auditory spasm and its correlation to hearing loss. Acta
evoked potential and other neurophysiologic monitor- Neurochir. 2014;156:571–5.
ing techniques during tumor surgery in the cerebellao- 50. *Sloan TB. Evoked potential monitoring of the central
pontine angle. In: Loftus CM, Traynelis VC, editors. nervous system intraoperatively. Anesthesiol Clin
Intraoperative monitoring techniques in neurosurgery. North America. 1997;15:593–611.
New York: McGraw-Hill; 1994. p. 175–91. 51. Legatt AD. Brainstem auditory evoked potentials
36. Yasargil MG. Microneurosurgery in CNS tumors, vol. (ABRs) and intraoperative ABR monitoring.
1. Stuttgart: Thieme Medical; 1996. p. 95–108. Handbook Clin Neurophysiol. 2010;9:282–302.
37. Bogousslavsky J, Caplan LR. Stroke syndromes. 2nd ed. 52. Chadwick GM, Asher AL, Van Der Veer CA, Pollard
New York: Cambridge University Press; 2001. p. 146. RJ. Adverse effects of topical papaverine on auditory
38. Little JR, Lesser RP, Luders H, Furlan AJ. Brainstem nerve function. Acta Neurochir (Wien). 2008;150:901–9.
auditory evoked potentials in posterior circulation 53. Jo KW, Lee JA, Park K, Cho YS. A new possible
surgery. Neurosurgery. 1983;12:496–502. mechanism of hearing loss after microvascular
39. Mannimen PH, Patterson S, Lam AM, Gelb AW, decompression for hemifacial spasm. Otol Neurotol.
Nantau WE. Evoked potential monitoring during pos- 2013;34:1247–52.
 Visual-Evoked Potentials
4
Sandra C. Toleikis and J. Richard Toleikis

individual variability, instability and unreliability

Introduction of the visual responses [32–36], their susceptibility
to anesthetics, particularly inhalational agents
One of the important goals of surgical procedures [32–37], and lastly and most damning, the poor
involving the visual pathways (retina, optic nerve correlation of IOM results to postoperative func-
(ON), optic chiasm, optic tracts, lateral geniculate tional outcomes. All of these findings have led to a
nucleus in the thalamus, optic radiation, and occip- general disenchantment with their intraoperative
ital visual cortex) is the preservation of visual func- use [7, 12, 18, 32–35, 37–45].
tion and in cases of visual impairment, where Still, on a case-by-case basis, monitoring of
possible, its improvement [1–4]. With these goals visual-evoked potentials (F-VEPs) has helped
in mind, efforts to evaluate and enhance the useful- guide surgeries of the orbit [46–48], and ante-
ness of intraoperative monitoring (IOM) of the rior visual pathways during tumor or lesion
visual pathways that began in the early 1970s have removal where its use has helped identify
continued. Wright et al. [5] are generally credited encroachment of tumors on the optic chiasm,
with the first report of a method for continuous and has aided in the differentiation of normal
monitoring of the visual pathways in 1973; utiliz- ON tissue from tumor tissue; especially when
ing brief flashes of light to evoke electroretino- the tumor encompasses the ON [2, 12–15, 19, 20].
graphic (F-ERGs) and visual-­ evoked potentials Direct ON stimulation has helped navigate dur-
(F-VEPs) during orbital surgery. This triggered a ing surgical removal of tumors involving the
number of other researchers to test their usefulness anterior visual pathway and skull base tumors
[2, 3, 6–21]. While some have reported favorable [13–15, 49, 50] with good outcomes. Though
results and outcomes [5, 22–31], others have dis- improved microsurgical techniques during pro-
missed their use, citing technical difficulties asso- cedures involving the sellar and parasellar
ciated with the delivery of visual delivery in an regions [19] have significantly reduced the inci-
operating room (OR) setting, large inter- and intra-­ dence of visual complications related to ON or
chiasmal manipulation and/or devasculariza-
tion, the potential remains for real-time, inad-
S.C. Toleikis, M.A. (*) • J.R. Toleikis, Ph.D.
vertent, and potentially harmful maneuvers that
Department of Anesthesiology, Rush University
Medical Center, 1653 W. Congress Parkway, Jelke may cause prolonged or intense indirect traction
739, Chicago, IL 60612, USA or compression of the ON to go unnoticed.
e-mail: [email protected] Concern for preventing devastating outcomes to

© Springer International Publishing AG 2017 51

A. Koht et al. (eds.), Monitoring the Nervous System for Anesthesiologists
and Other Health Care Professionals, DOI 10.1007/978-3-319-46542-5_4
52 S.C. Toleikis and J.R. Toleikis

the visual pathway has encouraged a small but  natomy and Physiology
A
dedicated group of researchers to pursue refinement of the Visual System
of the techniques for IOM of the visual pathways
[2, 3, 13, 15, 19–25]. The optic structures of the eye project images
Other reports of beneficial use of IOM visual onto the light-sensitive receptors of the retina,
pathway monitoring have been contained in the where a surprisingly high degree of neural
literature. F-ERGs [51, 52] and F-VEPs [36, 37, processing is accomplished through the retina’s
53, 54] have been reported to be helpful in assess- complex pattern of interconnections between
ing the depth of anesthesia. F-ERGs have been excitatory and inhibitory neurons. Some nerve
utilized in monitoring retinal function during eye fibers have small excitatory fields surrounded by
surgery [46, 47, 54–56] and endovascular proce- inhibitory areas, and others have inhibitory cen-
dures involving orbital or periorbital vascular ter areas surrounded by excitatory ones. As a
lesions [21] with good outcomes. They also have result, a good stimulus for exciting the visual
been used to monitor retinal perfusion during pathway would be one that undergoes changes in
procedures employing extracorporeal circulation contrast gradients (i.e., pattern-reversal).
and hypothermic circulatory arrest [11, 57]. Because, in general, patient cooperation is not
F-VEPs have reportedly been helpful for ana- possible for the majority of surgical procedures,
tomic navigation of the optic radiations during it is not feasible to utilize stimulation using high-­
surgical treatment of an occipital arteriovenous contrast pattern reversing checkerboard stimuli
malformation [17], and by use of diffusion tensor that are used in diagnostic testing. Hence, the fre-
imaging-based tractography for functional moni- quently employed stimulus for eliciting VEP
toring of the visual pathway [58]. Finally, visual-­ responses for monitoring purposes has been flash
evoked responses obtained from direct stimuli [24, 61–63]. While the spatial distribution
stimulation of the optic tract have been used as a of the light over the visual field of each eye is
method for globus pallidus internus (GPi) target- transmitted to the brain through the optic nerves,
ing during pallidotomy [59, 60] and deep brain very little information regarding the temporal
stimulation (DBS) interventions for treatment of variations in illumination is conveyed. Therefore
Parkinson’s disease where such procedures are it is key to note that when flash stimuli are
performed under general anesthesia or for employed for IOM monitoring purposes, what is
patients who otherwise are unable to cooperate actually being monitored is the visual pathways
during the procedure [16]. Although many of the for light perception and not for visual acuity [64].
above reports involved case report(s) or series, The neural information of flash stimuli travels
their findings suggest that further examinations from the optic chiasm onward, via the optic tracts
of these monitoring methodologies and applica- to the lateral geniculate body in the thalamus,
tions are needed. A better understanding of visual which then projects via connections to the visual
stimuli, the portions of the visual pathways that cortex (Fig. 4.1) [64]. Though coding of the visual
are stimulated, the methods for recording neuro- system has been intensively studied yielding a
physiologic responses, the effects of surgical wealth of information about the retina’s complex
manipulation, anesthetic management, and other neural network responses [65], information about
perioperative factors on responses, will hopefully the gross response from the ON and lateral genic-
lead to improved IOM results. This in turn may ulate body to flash stimuli remains relatively
spark renewed interest in research to further sparse, and in general, early cortical activation
enhance techniques and outcomes for IOM of the following flash visual stimulation is not well
visual pathways. understood in humans [65]. It is important to note
4 Visual-Evoked Potentials 53

ensure retinal and visual pathway stimulation.

They also have been used to monitor surgeries of
the orbit and as a measure of anesthetic depth [46,
47, 51, 52]. The main components of the F-ERG
are a negative-going a-wave with latencies occur-
ring between 24.2 ± 1.1 and 27.2 ± 3.7 ms and a
positive-going b-wave with latencies of
45.0 ± 1.5–55.1 ± 7.4 ms. The a-wave, in response
to a bright flash, largely reflects photoreceptor
function but there may be a contribution from
postreceptoral structures [67]. The b-wave, which
is of higher amplitude than the a-wave in normal
subjects, reflects postphototransduction activity.
The origin of the ERGs’ a- and b-waves is report-
edly a combination of activities that include pho-
toreceptor potentials, potassium-mediated current
flow, and DC potentials within Müller cells [67].
The F-ERG to a flash stimulus is a mass response;
thus F-ERG responses can appear normal when
dysfunction is confined to small retinal areas (e.g.,
Fig. 4.1 Schematic drawing of the visual pathway. OC
macular dysfunction). It has been reported that
optic chiasm, SC superior colliculus, LV lateral ventricle despite the macula’s high photoreceptor density,
(from Moller et al. [64]; with permission) an eye with purely macular disease has a normal
bright single flash ERG response [67]. Typical
that the optic pathways cross at the chiasm such ERG-VEP responses to flash stimuli delivered by
that monocular or binocular flash stimuli used in light-­emitting diode (LED) goggles are included
monitoring will produce bilateral pathway activa- in Fig. 4.2.
tion behind the chiasm unless a means of hemi-
field visual stimulation can be utilized (as can be
done in awake subjects). Flash VEP Response

The F-VEP is generated by postretinal areas of

 liciting and Recording Flash
E the central nervous system in response to visual
Visual-Evoked Potentials light stimulation and is a reflection of activity in
segments of the primary visual pathway that
Depending on the portion(s) of the visual path- project through the lateral geniculate body to the
ways at risk during surgical procedures, a num- cortical visual fields. The response is composed
ber of strategies for stimulation of those pathways of a triphasic waveform with an initial small pos-
have been explored. itive deflection (40–50 ms), followed by a second
negative deflection at 70–89 ms (often referred to
as N70 or N1), followed by a positive wave at
Flash Electrographic Response about 100 ms (P100 or P1) [68]. Though not
unequivocally documented, the generator sites
Responses generated by stimulation of the retina for the three waves of the F-VEP to LED stimula-
(F-ERG) have played a fundamental role in the tion are believed to originate from the lateral
diagnostic evaluation of retinal health [66]. For geniculate, striatum, and areas 17, 18, and 19 of
IOM purposes, they have been primarily used to the visual cortex [27, 68, 69].
54 S.C. Toleikis and J.R. Toleikis

Fig. 4.2 An example of intraoperative flash electroreti- angiography with stenting and coiling for treatment of a
nograms (F-ERGs) recorded from surface electrodes right ophthalmic artery aneurysm. The rate of flash stim-
placed in the orbital notches bilaterally and referenced ulation was 1.1/s. An average of 100 responses was
3 cm laterally and flash visual-evoked responses obtained. Anesthetic management included induction
(F-VEPs) recorded from electrodes placed at left occipi- with thiopental, sufentanil and dexmedetomidine (1 μg/
tal (O1), mid-occipital (Oz), and right occipital (O2) kg), and maintenance with sufentanil infusion, 0.5 MAC
scalp locations is shown. The responses were obtained at of isoflurane and continuous infusion of dexmedetomi-
baseline for a patient undergoing endovascular cerebral dine (0.7 μg/kg/h)

 echniques for Eliciting F-ERG

T structures related to visual acuity, would cer-
and F-VEP tainly be diminished if not lost if one were to try
to deliver these stimuli through closed eyelids.
The lack of suitable equipment for visual stimu- Because flash stimuli do not require patient fixa-
lation has been a severe handicap to the use of tion and cooperation and can be delivered through
ERG-VEP for IOM. Pattern-reversal stimuli or closed eyelids, they have been the most fre-
multifocal electroretinographic stimuli (which quently used stimulus in the OR setting.
are routinely utilized in diagnostic settings to Unfortunately, early research demonstrated that
evaluate retinal function related to acuity) cannot even when flash stimuli were employed for clini-
be utilized on unconscious patients in the OR set- cal diagnostic use (e.g., for assessment of ON
ting given the requirement for their cooperation pathology in multiple sclerosis patients), VEP
and visual fixation on the stimuli. Moreover, the responses to bright flashes were found to be nor-
shift in dark vs. light pattern/contrast of these mal while their responses to pattern-reversal
stimuli, which is key to stimulating the retinal stimulation were abnormal [68]. Moreover, the
4 Visual-Evoked Potentials 55

amplitude and latency of VEPs elicited by the the technique (e.g., lid sutures to keep the corneal
above flash stimulation methods unfortunately lens recording/stimulating device in place) and
demonstrate considerable patient inter- and intra-­ the potential risk of corneal abrasions and ulcer-
variability of amplitude both in diagnostic and ations with their use. The American
intraoperative settings [70]; especially in neo- Electroencephalographic Society (AEEGS)
nates, where maturation of the cerebral cortex Guidelines for IOM recommend that such hard
appears to play a factor [71]. So although flash lens stimulators be kept on the eyes for a maxi-
stimulation has largely been the stimulus of mum of 45 min (limiting their practical use for a
choice for IOM, it is not the best stimulus for majority of surgical procedures) and that users
assessing the preservation of fine visual acuity need to carefully examine the safety data from
and function. Rather it has been used to assess in the stimulator’s developer before employing
a rudimentary fashion whether response to stimu- such lenses in the OR [73].
lation can be conveyed along various points of the Because most shy away from the use of con-
visual pathways [71]. Even so, it is important to tact lens stimulators, LEDs mounted in eye
note that when there is preexisting visual dysfunc- patches [23–25] or goggles [9, 27] have been
tion that may disrupt the ability to convey such used, but the latter are bulky, require a headband,
stimulation, recording responses may be compro- and can interfere with surgery. Their use also car-
mised, although that has not been entirely con- ries some risks. Care needs to be taken to ensure
firmed. Multiple studies [2, 3, 20, 23–25] reported that the goggles are well-supported by the bony
that preexisting severe visual dysfunction (even ridge of the patient’s orbit and that they remain in
for patients with preexisting visual acuities of place during the surgery. Should they inadver-
<0.4 (20/50)) negatively impacted their ability to tently slip down and put direct pressure on the
record F-VEPs while others indicate that F-VEPs globe of the eye, they can cause central retinal
for IOM can be obtained successfully from some artery thrombosis [9].
patients with severe visual deficits [3, 19], when
onset of the dysfunction was acute.  ype of Light Stimulation
T
The AEEGS issued a recommendation that flash
 evices for Flash Stimulation
D VEPs induced by white stroboscopic light
Researchers have worked to develop improve- (F-VEPs) be differentiated from those induced by
ments in devices for the delivery of flash stimuli red LEDs (LED-VEPs) [73] by utilizing the appro-
for IOM use, since the problems of delivering priate abbreviations. However, in general, in the
stimuli may be key to developing more effective IOM literature, the terms F-ERG and F-VEPs
activation of the pathways that correlate with have been used to represent both. However, abbre-
functional vision. Initially a traditional strobe viations aside, the two methods of stimulation may
light was used to elicit responses but was found actually activate different retinal and cortical path-
to be cumbersome in the OR and delivery of the ways [12]. Newer LED stimulating goggles and
flash could be ineffective since it might be par- patches provide significantly increased luminosity
tially obscured by the scalp flap and drapes if a [23–25]. Although safety data regarding their use
bicoronal scalp incision is used, thus requiring a is scant, authors reporting on their IOM use have
modification of the surgical approach. not reported any postoperative sequelae, including
Fiberoptic haptic lens [72] and scleral contact cases involving lengthy neurosurgical procedures
lens [26] connected to photostimulators were [2, 3, 20, 23–25].
developed to provide flash stimuli in the 1970
and 1980s. Although some claim the resulting  onocular vs. Binocular Stimulation
M
responses were more robust than those later Although binocular stimulation has been used
obtained with LED stimulation (especially for some studies, when the goal is to evaluate
through dilated eyes) [72], use of contact lens for individual retinal and anterior pathways, the
IOM has lost favor due to the invasive nature of visual stimulus should be delivered and
56 S.C. Toleikis and J.R. Toleikis

recorded ­monocularly with responses from the subjects [73], each patient with a reproducible
contralateral eye, if clinically unaffected, used response should serve as his or her own control
as a control [73]. in the IOM setting. A simultaneous recording of
F-ERG responses is useful for confirmation of
Pupillary Size and Retinal Luminance retina stimulation.
When employing flash stimuli to elicit F-ERGs Before the surgeon approaches the optic path-
and F-VEPs, efforts to maintain pupil size and ways, the monitoring team should have identified
retinal luminance throughout the surgical proce- reproducible waves to be used as benchmarks for
dure are important because these parameters meaningful assessment during subsequent moni-
affect the response latency and amplitude. Kriss toring during critical stages of the surgery.
et al. [74] showed that F-VEP latency through Stacked plots of sequentially recorded averages
closed eyes is increased when compared to those are indispensable for assessing changes during
recorded through open eyes. Intraoperative use of surgery. Interpretation of intraoperative VEPs
narcotics often results in miotic pupils that reduce should be done in relation to pharmacologic,
retinal luminance, and staged dosing of narcotics physiologic, and surgical factors. Change in
during the surgical procedure may induce response trends should be reported as soon as
changes in latency and amplitude that may be identified, and immediate steps should be taken
misinterpreted as a surgically related event. It has to prevent the risk of lasting damage and to opti-
been recommended that maximal dilation of the mize normal function [73].
pupils be done through use of conjunctival instil-
lation of mydriatics at the start of the case [73], Stimulus Color
but that may be contraindicated when periopera- The color of the flash or LED (white vs. red)
tive assessment of pupillary response is required should be indicated on the record and kept con-
or preferred. Of note, recent and notably success- stant throughout the case.
ful reports of intraoperative recordings of F-ERG
and F-VEPs did not employ pupil dilation as part Stimulus Rate
of their total intravenous anesthetic and monitor- For transient F-ERGs and F-VEPS, a stimulus
ing management: propofol [2], administration of rate of 1–2.5 Hz is suggested and for steady-state
opioids [23–25], or remifentanil [2]. It is impor- responses a rate of 8–30 Hz may be used. Steady-­
tant to note that those with successful IOM state stimulation has not gained wide use for
recordings of F-VEPs employed newer and monitoring purposes. In a study published in
brighter 16 LED arrays embedded in soft round 2004, their use during surgery for monitoring
silicone disks [2, 20, 23–25] or in goggles [3, 18] purposes did not facilitate improved or more sta-
with luminosity adjustable from 500 to 20,000 ble F-VEP recordings [42].
Lumens (Lx), which helped to ensure retinal illu-
mination and stimulation.  ecording Electrodes and Their
R
Placement
 ecording F-ERGs and F-VEPs
R Corneal or scleral lenses for IOM recording of
The recommended standards for IOM of VEPs ERG responses were supplanted in the 1980s
were issued in 1987 by the American EEG with less invasive devices and methods. Corneal
Society [73]. To obtain consistent results, the recording devices using a Burian Allen electrode
guidelines advised that the following parame- and other devices placed on or near the cornea
ters be documented and remain constant yielded larger, robust responses but are rarely
throughout the IOM period. Though no such used due to the risks of corneal abrasion and
standards for IOM recording of ERG have been ulceration with IOM use. F-ERGS can be
published, the following would likely apply to recorded from skin with subcutaneous needles or
them as well. Given the expected high degree skin disc electrodes placed at the center of the
of F-VEP response variability even in awake right lower lid proximal to the lid margin
4 Visual-Evoked Potentials 57

(­ infraorbital notch), and referenced to an electrode above inion) to midline frontal (MQ-MF or
2 cm lateral to the lateral canthus where the larg- Oz-Fz) positions. The AEEGS guidelines [73]
est amplitudes of ERG responses from skin can recommend a second channel be recorded for sig-
be obtained [75, 76]. Esakowitz et al. [77] com- nal confirmation (however, few studies employ
pared the relative amplitudes of F-ERG b-wave them). When used, the secondary set of electrodes
responses when recording with corneal versus would be placed at midline occipital and refer-
other electrode types placed on the skin. The enced to linked earlobes (MQ-Ipsilateral Earlobe
largest response amplitudes were obtained with (AI)/Contralateral Earlobe (A2) or Oz-AI/A2)
the Burian Allen electrode (125 uV, 100 %), with locations. Typically the ground electrode is placed
other corneal electrodes yielding responses of at CZ. Additional channels may be used to study
less amplitude in proportion to those made with the scalp distribution of VEPs [73].
the Burien Allen electrode: JET (93 %), C-glide
(78 %), gold foil (60 %), and DTL (60 %).  nalysis Periods for F-ERGs and F-VEPs
A
Recordings from skin electrodes yielded the The AEEGS guidelines recommend the use of
smallest response amplitudes (14 %). Obviously, analysis periods of 100–200 ms for F-ERG [79]
one has to weigh the risks and benefits of the use and 250–500 ms for F-VEP [73]. A total of
of these devices for recording F-ERGs [77]. 50–200 responses are commonly recorded per
F-ERGs recorded with noncorneal versus corneal average [73], with the caveat that the number per
electrodes, except for amplitude differences, are average should remain constant throughout the
nearly identical in both time (a- and b-wave monitoring period. At least two consecutive ERG
latencies) and frequency domains (dominant and VEP averages should be acquired to confirm
power spectrum peaks), meaning noncorneal the reproducibility of the ERG and VEP wave-
F-ERGs do not differ significantly from corneal forms after setup and prior to surgical incision to
ERGs aside from amplitude [78]. establish a baseline (control) recording [23, 73].
One intriguing way to record F-ERGs was
recently reported by Houlden et al. [3]. In their  mplifier Settings for Recording F-ERG
A
small sample of study patients (N = 12), they and F-VEP
were able to reliably record F-ERGs from EEG For years the standard “clinical diagnostic set-
electrodes placed at Fz′ (2 cm behind Fz) and ref- tings” for system bandpass settings for F-VEPs
erenced to FPz (10–20 International EEG place- have also been employed for IOM purposes, which
ment) to confirm retinal stimulation in the are 1 to 200–300 Hz (−3 dB) with filter roll-off
patients (N = 12) evaluated in their study. Though slopes not exceeding 12 dB/octave for low fre-
a novel and appealing recording methodology for quencies and 24 dB/octave for high frequencies. If
F-ERGs, further studies are needed in larger irreducible artifacts occurred, filter settings could
patient populations to confirm its feasibility and be adjusted to 5–100 Hz. Digital smoothing and
utility for IOM. filtering could also be employed to reduce artifact,
For F-VEPs, standard subdural needle or sur- and filter settings should remain constant through-
face EEG electrodes may be used to record scalp out the monitoring session [73].
responses [73]. Ota et al. [29] reported that Houlden et al. [3] recently suggested that the
F-VEPs acquired with subdural electrodes better difficulties in recording F-VEPs in IOM settings
reflect cortical activity since they have consider- may be due to the high mean alpha EEG ampli-
ably greater spatial resolution and amplitude tudes (>50 μV) in patients, contributing “noise”
when compared to responses acquired from sur- that impedes recording the F-VEP “signal.” For 9
face EEG electrodes placed on the scalp. Single-­ of 12 patients with low mean alpha EEG ampli-
channel response recordings are acquired using tudes (<30 μV), IOM F-VEPs were reproducible,
the 10–20 International nomenclature for the including one whose vision was limited to finger
placement of EEG electrodes with electrodes counting. To improve the recording of responses
typically placed at the midline occipital (5 cm in three patients with high mean alpha amplitudes
58 S.C. Toleikis and J.R. Toleikis

(>50 μV), Houlden et al. [3] elected to see if rais- Kodama et al. [23] and Sasaki et al. [24, 25] have
ing the low pass settings had any effect. In these used a warning criterion whereby a F-VEP ampli-
patients, they simultaneously recorded EEG from tude decrease >50 % from baseline control levels
Oz–Fz and Fz–Fpz′ using 3-, 10-, and 30-Hz low prompted cessation of the surgical procedure
cut filters (six independent recording channels) until recovery of the F-VEP occurred and pro-
and two channels of F-VEP using low cut filters vided that other factors (e.g., anesthesia, use of
settings of 10 and 30 Hz. They found that F-VEP bipolar cautery) could not be used to explain the
amplitude reductions were minimal for low pass amplitude changes. Martinez Piñeiro et al. [81]
filter settings of 3–10 Hz but at 30 Hz, the reported on monitoring patients undergoing
F-VEP’s N1–P1 amplitude decreased by about endovascular treatment of their occipital arterio-
40 % and its morphology was significantly venous malformations (AVM) with F-VEPs. He
altered. They also found that as the low pass filter reported successful intraoperative recordings and
setting increased, the “noise” contribution to the postoperative outcomes employing the same
averaged F-VEP associated with electrocautery stimulation methods and parameters that Kodama
blocking time was also reduced. Based on those et al. [23] and Sasaki et al. [24, 25] used as well
findings, Houlden et al. recommended using as their warning criteria. On the other hand,
15–20 Hz as a low pass filter setting for F-VEP Kamio et al. [20], who examined the use of
responses. Houlden et al. points out that over 30 F-VEPs for patients undergoing transphenoidal
years ago, Nuwer and Dawson [80] recom- surgery for tumor removal, employed a warning
mended increasing the low cut filter settings from criterion of either an increased or decreased
1 to 30 Hz to improve intraoperative somatosen- amplitude of greater than 50 % compared to con-
sory-evoked potential (SSEP) reproducibility, trol levels. In a study utilizing a similar study
but at the time, did not offer any reasons why the population, Chacko et al. [40] used a complete
change improved the SSEP response. Houlden loss of visual responses as the warning criteria to
et al. [3] suggested that the improvement was due halt surgery until responses returned to baseline.
to reduction from the averaged response of the Hussain et al. [44], reporting on the use of moni-
patient’s alpha EEG and artifact due to electro- toring F-VEPs from five patients undergoing
cautery amplifier blocking. Though it seems a functional endoscopic sinus surgery, was the only
simple enough change to implement, only one study involving the anterior visual pathway to
recent study, by Kamio et al. [20], has employed employ an increase in F-VEP P100 latency as an
a higher low bandpass setting (20 Hz) during indicator of optic nerve compression. They noted
recordings of F-VEP. Hopefully others will take that for this criterion to be useful for IOM, the
the opportunity to test Houlden et al.’s hypothe- patient’s intraoperative diastolic blood pressure
ses and determine whether indeed raising the low had to remain higher than 50 mmHg, oxygen
pass filter setting improves the reproducibility of saturations 98 % or higher, and bleeding to be
F-VEP responses in IOM settings. minimized. Given the notable disparity of F-VEP
IOM warning criteria in the previously discussed
Monitoring Criteria studies, it is clear that more research is needed to
Given the documented variability of flash VEPs, better define IOM F-VEP warning criteria that
responses recorded from patients in operative better correlate with patient outcomes.
settings cannot be universally characterized [73]. Regarding warning criteria for studies utilizing
That and perhaps the technical difficulties associ- F-ERGs for monitoring, Padalino et al. [21] used
ated with obtaining reproducible F-VEPs in oper- them to monitor retinal perfusion during a single
ative settings have contributed a lack of clear case involving endovascular treatment of a dural
guidelines for warning criteria to be used during AVM supplied by the bilateral superficial tempo-
surgery to preserve the visual pathway. For recent ral, ophthalmic, and the right middle meningeal
neurosurgical procedures involving removal of arteries. The intraoperative monitoring warning
intraorbital, parasellar, and cortical lesions, criteria that they employed was a 10-ms increase
4 Visual-Evoked Potentials 59

in the F-ERG latency and a 30 % decrease in its and has a blood barrier analogous to the blood–
amplitude compared with baseline and control brain barrier. It would seem reasonable to assume
responses from the other eye. As with the warning that the changes observed in the retinal microcir-
criteria for F-VEPs during IOM, further study is culation also occur in brain. The central nervous
needed to define and confirm warning criteria for system can suffer from the same pathophysiolog-
use of F-ERGs for IOM. ical entities that affect the retina and the ONH
during hypothermic CPB procedures.” During
 ther IOM Applications with F-ERG
O CPB procedures, retinal ischemia and infarction
and F-VEP due to emboli, anterior ischemic optic neuropa-
Keenan et al., Burrows et al., and Reilly et al. [11, thy (AION), posterior ischemic optic neuropathy
82, 83] have explored the F-VEP as an objective (PION), damage of nerve fibers, chorioretinal
measure of the short-term effects of various car- hypoperfusion and hypoxia secondary to hemo-
diopulmonary procedures on neurophysiological dynamic and hematologic changes have resulted
function given the cortex’s sensitivity to small in profound visual deficits and other neuro-oph-
changes in cerebral perfusion due to its proximity thalmological complications [86]. Because
to the watershed area of the posterior and middle F-ERG responses are sensitive to (1) alteration of
cerebral arteries. Reilly et al. [83] showed that blood flow as a consequence of the reduction of
F-VEPs are a more sensitive indicator of central perfusion pressure and (2) to body hypothermia,
nervous system (CNS) stress provoked by com- associated with hemodilution, which helps to
bined hypothermia and hypoxia than depress neural function and neural tissue oxygen
EEG. Burrows et al. [11] found F-VEPs to be an requirements, but may also bring on tissue
objective measure of neurophysiologic function hypoxia, it stands to reason that focus be given to
in the visual pathway during profound hypother- the use of F-ERGs as a tool to monitor patients
mic circulatory arrest (PHCA) in neonates and undergoing cardiosurgical procedures with
infants undergoing surgical correction of congen- CPB. A recent, intriguing recent paper by Brandli
ital heart defects. Although their findings seemed and Stone [87], while not directly related to intra-
promising as an IOM tool for such cases, operative monitoring and performed in rats, sug-
Markand et al. [84] found VEPs to be too incon- gests that F-ERGs are indeed sensitive to
sistent during the surgical course of hypothermia ischemia, even when the induced ischemia is
and recovery; and that their disappearance at remote to the retina. A pilot study done by
temperatures below 25 °C made them less than Nebbioso et al. [57] examined the use of F-ERGs
ideal for monitoring brain function during hypo- for monitoring patients during extracorporeal cir-
thermia. Even Burrows and Bissonnette [85] culation (ECC), both hypothermic and normo-
appear to have abandoned the use of F-VEPs for thermic with some promising results. Under
this application, opting instead to use other mea- hypothermic ECC, they reported that the ampli-
sures of cerebral blood flow (transcranial Doppler tude of the F-ERG response decreased by 50 %
sonography) for monitoring perfusion during while those under normothermic ECC only
CPB surgery in their subsequent study. decreased 10 %. Recovery of response ampli-
However, the use of F-ERGs to monitor car- tudes to baseline levels occurred upon the end of
diac surgery and extracorpeal circulation may ECC and with the rewarming of the patient, with
serve as a new area to consider for IOM purposes. the exception of one individual whose F-ERG
Indeed, monitoring the retina as an extension of response recovery was very prolonged. That
the brain for such cases has been encouraged by patient required postoperative ventilatory sup-
Nenekidis et al. [86]. They believed that better port and a long stay in the intensive care unit
quantification of the hemodynamic state of ret- [57]. Time and further studies will tell whether
ina–optic nerve head (ONH) during on-pump use of F-ERGs will become a useful tool for
CPB, is needed, stating that, “The retina provides a monitoring and maintaining adequate retinal as
‘window’ for the study of cerebral microcirculation; well as cerebral microcirculation and perfusion
it lies in the territory of the internal carotid artery during surgical procedures.
60 S.C. Toleikis and J.R. Toleikis

 etinal Stimulation and Intracranial

R cases where little or no anesthesia is used during
Recording of Responses surgery, the option to use pattern reversal and
other visual stimuli that provide more informa-
Recording directly from cortical structures for tion in visual pathway function (e.g., visual acu-
lesions of the visual pathways, though feasible ity), may open up a new era for using VEPs for
(albeit in studies with relatively small sample monitoring purposes. Indeed, their use during
sizes), has limited utility but has yielded other “awake” procedures (e.g., endovascular
responses with greater amplitudes and better stenting and coiling procedures) when visual
signal-­to-noise ratio [45]. Møller et al. [88] function is at risk, has yet to be fully explored.
recorded compound action potentials directly
from optic nerve (ONEP) elicited by short light
flashes via LEDs during tumor resection in two  irect Electrical Stimulation of Optic
D
patients and was able to record responses that had Nerve
an initial small positive deflection, with a latency
of about 45 ms, followed by a negative wave with Bošnjak and Benedičič [14, 15, 50] evaluated the
a latency of 60–70 ms, although considerable feasibility and utility of recording scalp VEP
individual variation in the shape and size of those responses to direct electrical stimulation (eVEP)
responses was observed [88]. of the ON during tumor removal surgery involv-
In another application of this type of record- ing the anterior visual optic pathways [14], skull
ing, Curatolo et al. [89] found that monitoring the base [15], and during orbital enucleation due to
responses recorded from the visual cortex during malignant melanoma of the choroid or the ciliary
photic stimulation proved to be a reliable tech- body [91]. To acquire cortical potentials elicited
nique for preserving central vision during occipi- by electrical epidural stimulation of the optic
tal lobe surgery [89]. Ota et al. also evaluated the nerve (ON), insulated platinum needle-­stimulating
usefulness of cortically recorded VEPs as an electrodes with a noninsulated ball tip were
IOM tool of the posterior visual pathway in 17 attached epidurally to both sides of the
patients who underwent posterior craniotomy for ON. Bošnjak et al. [50] used the following proce-
lesions or epileptic foci in the parietal, posterior dure for placement of these electrodes, noting that
temporal, and/or occipital lobes; reporting detec- “When the exit of the optic nerve (ON) from the
tion of VEPs in over 90 % of cases with preserved periorbit is fully visualized through a small fenes-
vision that was independent of the type of anes- tration of the orbital apex, needle electrodes are
thesia employed for those procedures [29]. placed in contact on each side of the ON into the
In still another application of this type of cleft between the nerve itself and the basal rem-
recording, photic stimulation while recording nants of the lateral walls of the optic canal. The
averaged visual responses in the optic tract dur- needles are manipulated during positioning with
ing pallidotomy and deep brain stereotaxic sur- bipolar forceps through grip connectors. After
gery has been reported to be useful in guiding placing the epidural stimulating electrodes, their
those surgeries [16, 59]. Such recordings have position is secured with wet cotton patties laid
also been useful in determining the generators for over the orbital apex and leads.” Monopolar optic
scalp-derived VEP responses. Tobimatsu et al. nerve potentials after retinal flash or ­electrical
[90] was able to record VEPs using pattern-­ epidural stimulation of the ON were then recorded
reversal stimuli in eight awake patients with with insulated platinum ball-tipped wire elec-
Parkinson’s disease undergoing stereotactic palli- trodes placed on the surface of the ON using an
dotomy using a depth recording electrode located extracephalic reference electrode. The distance
at or below the stereotactic target in the ventral between the stimulating and recording electrodes
part of the GPi and dorsal to the optic tract. They was approximately 25 mm. The same recording
simultaneously recorded VEPs from the scalp to electrodes were used for monopolar recordings
provide information for differentiation of the from structures outside of the visual pathway to
generators of the scalp VEP components. In such collect control data. The electrical stimulus con-
4 Visual-Evoked Potentials 61

sisted of a rectangular current pulse of varying

intensity (0.2–5.0 mA) and duration (0.1–0.3 ms)
using a stimulation rate of 2 Hz [14, 15]. The
bandpass filter settings utilized in previous studies
were 1–1000 Hz when recording these cortical
potentials after electrical epidural stimulation of
the ON. The analysis time used was between 10
and 300 ms. Each trace was generated from the
average of 100 responses. Of note, considerable
stimulus-related artifact from direct ON stimula-
tion does present a technical hurdle to recording
these potentials [14, 15]. Using this stimulation
and recording technique, Benedičič and Bošnjak
[14, 15] concluded that it was beneficial in pre-
venting ON damage and improving outcomes.
They did not report any warning criteria used in
the studies and their sample sizes were small [4].
The typical eVEP they recorded consisted of N20
and N40 waves (Fig. 4.3) [50]. Considerable vari-
ability in the amplitude of the responses was
observed (e.g., N40 wave amplitudes prior to
Fig. 4.3 (a) An example of a typical electrically elicited
tumor removal varied as much as 25 %). Not sur-
visual-evoked potential (eVEP) response is depicted. The
prisingly, artifact was observed with use of bipo- response consists of a larger N20 wave and a smaller N40
lar coagulation, ultrasonic aspirator, laser, and wave. The stimulus duration was 0.5 ms, and its frequency
craniotome-­hampered IOM [15]. In one patient 2 Hz. An average of 100 responses was obtained. (b) The
eVEP responses recorded from a patient with an optic
with an ON sheath meningioma and vision lim-
nerve sheath meningioma and visual perception of light
ited to light sensation, only the N20 wave was only is shown for which a decreased N20 wave and absent
observed (see Fig. 4.3) [50]. In their subsequent N40 wave are observed (from Bošnjak and Benedičič
report of IOM monitoring for a very small sample [50]; with permission)
(N = 3) of patients undergoing orbital enucleation
due to malignant melanoma of the choroid or the junction. They used a 5-mm spaced bipolar elec-
ciliary body, both F-VEPS and cortical potentials trode to deliver biphasic current stimuli (pulse fre-
from direct stimulation of the optic nerve were quency of 60 Hz, single-pulse duration of 1 ms,
inconsistent or absent in patients with a history and amplitude of 5 mA) to cortex involving the
(>3 months) of severe visual deterioration, but optic tracts, in a nonsedated patient, to guide tumor
obtainable from a single patient with a short his- removal. By mapping the optic radiations using
tory of mild visual impairment [91]. Clearly, more this electrical stimulation and obtaining the
studies are needed to confirm the utility of direct patient’s report of visual effects related to that
electrical stimulation of the optic nerve for visual stimulation, they were able to detect the posterior
pathway IOM use and for the development of and deep functional boundary of the tumor resec-
effective preoperative criteria for patients in tion, and to avoid production of a postoperative
whom these techniques may be useful, as well as symptomatic homonymous hemianopsia. Given
warning criteria that correlates with and improves the incidence of visual field defects following sur-
patient outcomes. geries involving the posterior temporal lobe and
Duffau et al. [13] described the use of intraop- temporo-parieto-occipital junction (TPOJ) with
erative electrical stimulations (IES) during surgery considerable risk for the occurrence of permanent
to help identify and preserve afferent visual fibers homonymous hemianopsia, it may be of great
during removal of a low-grade glioma invading interest for surgeons and IOM practitioners to
the whole temporal lobe and temporo-­occipital consider conducting further research in the
62 S.C. Toleikis and J.R. Toleikis

application of this direct electrical stimulation Wongpichedchai et al. [93] evaluated the effects
technique to help preserve visual function and of halothane in a pediatric population, on dark-
improve patient outcomes for those undergoing adapted (scotopic) and light-adapted (photopic)
such procedures [13]. F-ERGs and found it had little effect on the a-wave
and b-wave latency and amplitude of the scotopic
F-ERG and amplitudes and latencies of the F-ERG
Effects of Temperature photopic responses to red flashes and 30 Hz flick-
ering white light. In another study, Tremblay et al.
In conscious humans, F-VEP latency is 10–20 % [66] retrospectively compared the effects of seda-
longer at 33 °C than at 37 °C. With increasing tives and inhalational agents on scotopic and phot-
hypothermia, progressive increases in F-VEP opic F-ERGs in a small sample of pediatric
latency and decreases in F-VEP amplitude occur patients diagnosed free of retinal disease. In that
with complete loss of the components of the study, F-ERGs were recorded in subjects who
responses at 25–27 °C. When cooling occurred either were (1) conscious (no anesthetic or sedative
more swiftly, F-VEP responses disappeared at medications) [n = 9]; (2) under sedation (chloral
higher temperatures than when a slower cooling hydrate [75–125 mg/kg] and pentobarbital sodium
process was utilized [92]. sedation [5–6 mg/kg]) (n = 9); or (3) under general
anesthesia (intravenous injection of propofol
[2 mg/kg] with or without fentanyl 4 μg/kg, and
Effects of Anesthesia on F-ERGs maintained with isoflurane 2–3 % or halothane
1–2.4 % with 50 % O2 and 50 % N2O [n = 9]). They
Although for IOM purposes, F-ERG responses found that sedation appeared to decrease a- and
are most frequently being employed for confir- b-wave amplitudes of the scotopic bright-flash
mation of stimulation of the retina, gaining an F-ERG responses, without affecting the responses’
understanding of the reported effects of anes- latencies. Though Tremblay et al. [66] reported
thetic and sedative agents on retinal responses, that F-ERG responses recorded under photopic
and conducting additional studies to further elu- conditions showed minimal changes in latencies
cidate the effects of these agents on retina physi- and amplitudes, if one examines the table provided
ology, is of interest. in the paper [66] (Table 4.1), it appears that while

Table 4.1 Summary of statistical analysis performed on ERG parameters obtained after 5 min in photopic conditions
in the conscious (C), sedated (S), and anesthetized (A) patients
ERG parameter Conscious Sedated Anesthetized C-S C-A S-A
Amplitude (μV ± 1 SD)
a-wave 76 ± 20 63 ± 9 54 ± 20 − + −
b-waveparameters 216 ± 49 186 ± 35 163 ± 47 − − −
OP2 19.1 ± 0.9 15.2 ± 4.0 17.1 ± 4.6 − − −
OP3 21.1 ± 2.7 17.3 ± 7.2 18.5 ± 11.7 − − −
OP4 35.5 ± 14.5 22.6 ± 11.6 9.6 ± 4.7 − + −
OP5 2 9.2 ± 10.7 20.1 ± 8.6 8.8 ± 2.0 − + −
Implicit time (ms ± 1 SD)
a-wave 13.7 ± 0.4 14.1 ± 0.5 16.2 ± 1.0 − + +
b-wave 33.1 ± 0.8 34.4 ± 1.5 46.8 ± 4.9 − + +
OP2 16.4 ± 0.3 17.2 ± 0.4 20.0 ± 0.8 + + +
OP3 24.3 ± 1.1 25.4 ± 1.6 29.3 ± 1.6 − + +
OP4 32.0 ± 1.1 32.6 ± 1.5 44.3 ± 5.2 − + +
OP5 40.8 ± 1.5 41.1 ± 1.7 52.6 ± 4.8 − + +
Stars in the three rightmost columns indicate statistical difference between paired-wise groups after post hoc Bonferonni/
Dunn correction (P>0.016) (from Tremblay et al. [66]; with permission)
4 Visual-Evoked Potentials 63

the amplitude of the responses are not affected, the s­urgery under general anesthesia induced with
latencies of the photopic responses recorded under sevoflurane (8 %) in 100 % oxygen (O2) and main-
anesthesia versus consciousness are significantly tained with sevoflurane (range, 0.05–0.31 %,
and statistically increased and that the same holds mean 0.22 ± 0.07 %) and nitrous oxide (mixture of
true for F-ERG responses recorded under sedation 33 % O2 and 66 % N2O). The F-ERG responses
versus under anesthesia [66]. were recorded preoperatively from unpremedi-
Early studies by Raitta et al. [94] evaluated cated American Society of Anesthesiologist
F-ERG responses recorded in ten adults before (ASA) classification I and II patients, and again
and 15–20 min after induction of anesthesia with immediately following discharge from the recov-
a combination of thiopentone sodium, halothane, ery room, and 24 h following sevoflurane/N2O
and nitrous oxide (N2O) and found that the ampli- anesthesia. Ioholm et al. [95] found that the
tudes of F-ERG a- and b-waves were significantly F-ERG b-wave latency in these subjects was
decreased when compared with preoperative lev- increased at both postoperative time points com-
els, but that latencies were unchanged. The find- pared with preoperative responses and that
ings of a study by Yagi et al. [52] evaluated the b-wave amplitudes were also decreased postop-
effects of enflurane on F-ERG in a small sample eratively compared with their preoperative levels.
of patients undergoing surgical procedures and Similar findings were obtained in a subsequent
found that use of enflurane significantly increased study conducted by Ioholm [96]. Of interest,
latencies of F-ERG a-waves and b-waves and Sasaki et al. [24] reported that “after induction of
decreased their amplitudes but did not have any inhalation anesthesia with sevoflurane, ERG data
significant effects on F-ERG b-wave responses, were not reproducible” [24]. An example of the
with increasing concentrations of enflurane (0, comparison of the F-ERG responses obtained by
0.8, and 1.7 %) [52]. Interestingly, a study con- Sasaki et al. [24] with anesthesia employing sevo-
ducted by Ioholm et al. [95] in adult populations flurane versus total intravenous infusion of propo-
examined photopic F-ERGs before and after fol with fentanyl is shown in Fig. 4.4. Given the

Fig. 4.4 Illustrates the reproducibility of ERG and VEP (left), F-ERG responses were not reproducible and the
following induction of inhalation and venous anesthesia in VEP amplitude was also affected. Conversely, following
the same patient, same eye (without visual dysfunction). induction of propofol-based infusion anesthesia, reproduc-
F-ERG and VEPs were obtained twice to confirm the ibility of ERG and VEPs were both good (from Sasaki
reproducibility of the data in the absence of surgical proce- et al. [24] with spelling modifications: revised
dures. Following induction with inhalation anesthesia “Seboflurene” to “Sevoflurane”; with permission)
64 S.C. Toleikis and J.R. Toleikis

variance in the reports regarding the effects on published on the effects of bolus administrations
F-ERG responses to different combinations of of fentanyl or other opioids on F-ERGs in humans
sedative and halogenated agents, one would have during surgery, bolus administrations (additional
to agree with Tremblay’s conclusion that normal injections of fentanyl every 60 min) were used in
retinal physiology is affected by sedation and anes- Sasaki’s et al. study [24], with what can pre-
thesia through different mechanisms that remain to sumed to be little detriment to their use of F-ERG
be fully elucidated by future research [66]. responses (chiefly to ensure retinal stimulation)
Total intravenous anesthesia (TIVA) utilizing for IOM purposes.
propofol and opioid medication is touted as
F-ERG friendly. Indeed F-ERGs obtained from
20 normal children undergoing evaluations of Effects of Anesthesia on VEPs
their visual function under anesthesia with pro-
pofol and fentanyl versus topical anesthesia A summary of the effects of anesthesia on
showed that F-ERG b-wave response latencies F-VEPS prepared by Banoub et al. is shown in
were only slightly increased and b-wave ampli- Table 4.2 [1]. F-VEPs are very sensitive to the
tudes decreased when propofol and fentanyl ver- effects of anesthetics and physiologic factors
sus topical anesthesia were employed and were because they represent polysynaptic cortical
not statistically different [97]. This finding was activity. Because flash stimulation activates both
further supported by an animal study conducted in temporal and nasal parts of the retina and the
pigs [98]. Although no specific reports have been nasal fibers cross to the contralateral side at the

Table 4.2 Summary of the effects of anesthetics on visual-evoked potentials responses

Anesthetic drug Dose/concentration Latency of P-100 Amplitude
Halothane [102] 1 MAC ≈10 % ↑ Inconsistent
Isoflurane [41, 103] 0.5 MAC 10 % ↑ 40 % ↓
1.0 MAC 20 % ↑ 66 % ↓
1.5 MACa 30 % ↑ 80 % ↓
1.0 MAC + 70 % N2O Abolished Abolished
1.5 MAC + 70 % N2O Abolished Abolished
Sevoflurane [104] 0.5 MAC + 66 % N2O 5–10 % ↑ 20 % ↓
1 MAC + 66 % N2O Abolished Abolished
1.5 MAC + 66 % N2O Abolished Abolished
1.4–1.7 MAC Abolished Abolishedb
Nitrous oxide [105–107] 10–50 % No effect 25–80 % ↓c
Propofol [108] 2 mg/kg + 10 mg kg−1 h−1 Negligible ≈20 % ↓
Thiopental [109] 3 mg/kg <10 % ↑ No change
6 mg/kg Abolished Abolished
Etomidate [109] 0.3 mg/kg <10 % ↑ No change
Fentanyl [99] 10–60 μg/kg <10 % ↑ 30 % ↓
Ketamine [108] 1 mg/kg + 2 mg kg−1 h−1 Negligible ≈60 % ↓
Morphine scopolamine 0.2 mg/kg morphine + 0.4 mg scopolamine No change ≈20 % ↓
(premedication) [99]
Neuroleptanalgesia [110] 10 % ↑ No change
Fentanyl, droperidol nitrous oxide
From Banoub et al. [1]; with permission. All data are from humans
MAC minimum alveolar concentration, N2O nitrous oxide, ↑ increase, ↓ decrease
a
In a substantial fraction of patients, waveforms were not recordable at this concentration
b
During electroencephalogram suppression; visual-evoked potentials reappeared during electroencephalogram
bursts [111]
c
Some report a 40 % increase in N-70–P-100 amplitude [108] (Fig. 4.10)
4 Visual-Evoked Potentials 65

level of the optic chiasma, retrochiasmatic lesions was decreased. They posited that those decreases
cannot be monitored [33]. In addition, VEPs are may be due to changes to retinal luminance related
highly dependent on appropriate stimulation of to pupillary-induced constriction associated with
the retina and may be unduly affected by narcotic-­ fentanyl bolus administration [99]. Accordingly,
induced pupillary constriction [99]. bolus administraton of fentanyl, while not pre-
The findings included in Table 4.1 suggest cluded during such surgeries, may indirectly
volatile anesthetics prolong VEP latency and affect F-VEP responses induced by retinal flash
decrease F-VEP amplitudes in a dose-dependent stimulation, and therefore should be taken into
fashion. Nakagawa et al. [37] found that even at a account if responses change post-administration.
1 % (0.5 minimum alveolar concentration [MAC]) Loughnan et al. [100] showed that neither fen-
concentration of sevoflurane, responses were sig- tanyl, 200 μg, nor diazepam, 20 mg administered
nificantly decreased. At 1.5 MAC, responses intravenously, significantly changed F-VEP
could not be interpreted [37]. However, as men- latency or amplitude, suggesting that an anes-
tioned previously, conflicting information about thetic technique based on these two drugs might
the effect of low-dose sevoflurane on F-VEP be suitable when intraoperative evoked potential
responses have been reported, with one researcher monitoring is required to assess ischemia and
finding amplitude decreases with its use [37] preservation of visual-evoked responses [100].
while another did not find any such decrease [29], With respect to anesthetic techniques that
although the latter recorded responses directly employ infusion of propofol, although most of
from cortex and those responses are reportedly the new studies for visual pathway IOM espouse
not as susceptible to the effects of inhalational the use of propofol, Neuloh [45] points out that
agents as those recorded from scalp [45, 88]. TIVA alone cannot ensure success, as one recent
Nitrous oxide (N2O) alone considerably reduces study found that “a satisfactory rate of successful
VEP amplitude. Its use in addition to volatile VEPs could not be achieved despite use of TIVA
anesthetics can make VEP responses unrecord- for anesthetic management” [41]. Moreover, a
able. Increased concentrations of nitrous oxide couple of studies have found that the amplitude
significantly increase VEP latencies. of VEP is strongly affected by the concentration
In general, it appears that opioid and ketamine of propofol and that caution and perhaps further
or propofol-based anesthetic techniques (TIVA), studies are needed in evaluating VEP in patients
along with those employing low-dose volatile undergoing propofol anesthesia. Nakagawa et al.
anesthetics without nitrous oxide, seem to facili- [37] found that at a propofol concentration of
tate intraoperative recording of VEPs but do not 3.0 μg/mL (80–100 μg/kg/min), VEP amplitudes
ensure it. In some cases, the use of these anes- were decreased significantly compared with the
thetic protocols may involve a high incidence of amplitude at 1.5 μg/mL concentration (40–50 μg/
false-positive and false-negative results [33]. kg/min). It led him to conclude that a propofol-­
Opioids (e.g., fentanyl, alfentanil, sufentanil, based TIVA technique appears to induce less
and remifentanil) reportedly have a very mild change in evoked potentials, including VEP, than
effect on other evoked responses [100], which halogenated agents [37]. Hamaguchi et al. [36]
presumably extend to VEP responses. However, it further investigated the influence of propofol
is important to keep in mind that bolus adminis- concentration on F-VEP components in three
tration of opioids has been reported to signifi- patients with cranial aneurysm and four with
cantly reduce the amplitude of scalp-recorded brain tumor. Anesthesia was maintained with
responses [101]. Chi et al. [99] studied the effects intravenous propofol using target controlled infu-
of incremental doses of fentanyl (10 μg/kg) given sion. Changes in F-VEP amplitude and latency
every 10 min for a total dose of 60–90 μg/kg for were measured during three propofol concentra-
patients undergoing coronary artery bypass graft tions (effect site concentrations of 1.5, 2.0,
procedures and observed that while fentanyl and 3.0 μg × mL−1, and correlated with bispectral
administration did not affect latency, amplitude index (BIS) readings at each concentration.
66 S.C. Toleikis and J.R. Toleikis

At 3.0 μg × mL−1 propofol concentration, F-VEP anesthetic management, it is still not clear that the
amplitude was decreased significantly compared use of TIVA with propofol ensures F-VEP record-
with the amplitude at 1.5 μg × mL−1 concentration. ing in all patients [41]. Still, there clearly is a need
No significant change was observed with the for continuous monitoring during cases involving
latency of F-VEP. The value of BIS at 3.0 μg × mL−1 the visual pathways, and indeed studies indicating
propofol concentration also decreased significantly that they were able to detect ischemic response
compared with 2.0 μg × mL−1 concentration. changes in the F-VEP that would have been
Thankfully, neuromuscular blocking drugs do missed by imaging data [58] make a compelling
not directly influence F-VEP responses. In fact, argument to encourage those who would continue
their use may contribute to an improved signal-­ efforts to optimize the methodologies for use of
to-­noise ratio by eliminating electromyographic VEPs for IOM purposes. Certainly, replication of
artifact [12]. their protocols and confirmation of their results
will help solidify their methods. Only time will
tell if others take up that cause [5, 45]. The devel-
Conclusion opment of visual stimulation methodologies that
can better assess visual acuity during surgical pro-
A number of researchers [2, 3, 13–15, 19–21, 23, cedures, coupled with improved anesthetic man-
25, 78, 79] have continued their efforts to over- agement techniques may serve to revitalize the
come the poor reputation that IOM of the visual efforts to confirm the usefulness of IOM of the
pathways has had for several decades. By employ- visual pathways: a plea that was issued years ago
ing new and brighter stimuli, direct cortical stim- and still remains true [73].
ulation methods, and monitoring new types of
cases, these researchers hope to spawn revitaliza-
tion of research that will help evaluate, establish,
and improve the usefulness of visual pathway
References
intraoperative monitoring. Certainly, the success 1. Banoub M, Tetzlaff JE, Schubert A. Pharmacologic
of Sasaki et al. [24, 25] (93.5 %), Kodama et al. and physiologic influences affecting sensory evoked
[23] (97 %), and subsequent research [19, 20] for potentials: implications for perioperative monitor-
acquisition of stable scalp F-VEPs recordings ing. Anesthesiology. 2003;99:716.
2. Luo Y, Regli L, Bozinov O, Sarnthein J. Clinical
during surgeries involving the anterior pathways, utility and limitations of intraoperative monitoring
with good correlation of monitoring results and of visual evoked potentials. PLoS One. 2015;10(3):
visual outcomes, have been encouraging. Their e0120525.
ability to achieve those goals has largely been 3. Houlden DA, Turgeon CA, Polis T, Sinclair J,
Coupland S, Bourque P, et al. Intraoperative flash
attributed to the combined use of (1) a brighter VEPs are reproducible in the presence of low ampli-
reusable (sterilizable); (2) flexible LED stimulat- tude EEG. J Clin Monit Comput. 2014;28:275–85.
ing devices that guarantee supramaximal retinal 4. Duffau H. Intraoperative monitoring of visual func-
stimulation, even when VEPs cannot be recorded; tion. Acta Neurochir (Wien). 2011;153:1929–30.
5. Wright JE, Arden G, Jones BR. Continuous monitor-
and (3) use of TIVA with propofol to minimize the ing of the visually evoked response during intra-­
effects of anesthetics on the responses. However, orbital surgery. Trans Ophthalmol Soc U K.
monitoring of patients with preoperative visual 1973;93:311–4.
deficits remains controversial. Indeed, Kodama 6. Handel N, Law J, Hoehn R, Kirsch W. Monitoring
visual evoked response during craniofacial surgery.
et al. limited their monitoring to patients with pre- Ann Plast Surg. 1979;2:257–8.
operative acuities of less than 0.4 (20/50) [23]. 7. Allen A, Starr A, Nudleman K. Assessment of sen-
Because impaired preoperative vision is a major sory function in the operating room utilizing cere-
predictor of postoperative deterioration [45], bral evoked potentials: a study of fifty-six surgically
anesthetized patients. Clin Neurosurg. 1981;28:
eliminating such patients from monitoring limits 457–81.
the broad usefulness of VEPs for IOM of these 8. Grundy BL. Intraoperative monitoring of sensory-­
surgical procedures. Regarding enhancements in evoked potentials. Anesthesiology. 1983;58:72–87.
4 Visual-Evoked Potentials 67

9. Albright AL, Sclabassi RJ. Cavitron ultrasonic surgi- artery during aneurysm clip placement surgery.
cal aspirator and visual evoked potential monitoring J Neurosurg Pediatr. 2007;107:865–7.
for chiasmal gliomas in children. Report of two 23. Kodama K, Goto T, Sato A, Sakai K, Tanaka Y,
cases. J Neurosurg. 1985;63:138–40. Hongo K. Standard and limitation of intraoperative
10. Costa e Silva I, Wang AD, Symon L. The application of monitoring of the visual evoked potential. Acta
flash visual evoked potentials during operations on the Neurochir (Wien). 2010;152:643–8.
anterior visual pathways. Neurol Res. 1985;7:11–6. 24. Sasaki T, Itakura T, Suzuki K, Kasuya H, Munakata
11. Burrows FA, Hillier SC, McLeod ME, Iron KS, R, Muramatsu H, et al. Intraoperative monitoring of
Taylor MJ. Anterior fontanel pressure and visual visual evoked potential: introduction of a clinically
evoked potentials in neonates and infants undergo- useful method. J Neurosurg. 2010;112:273–84.
ing profound hypothermic circulatory arrest. 25. Sasaki T, Ichikawa T, Sakuma J, Suzuki K,
Anesthesiology. 1990;73:632–6. Matsumoto M, Itakura T, et al. Intraoperative moni-
12. Sloan TB. Evoked potential monitoring. Int toring of visual evoked potentials [in Japanese].
Anesthesiol Clin. 1996;34:109–36. Masui. 2006;55:302–13.
13. Duffau H, Velut S, Mitchell M-C, Gatignol P, 26. Feinsod M, Selhorst JB, Hoyt WF, Wilson
Capelle L. Intra-operative mapping of the subcorti- CB. Monitoring optic nerve function during crani-
cal visual pathways using direct electrical stimula- otomy. J Neurosurg. 1976;44:29–31.
tions. Acta Neurochir (Wien). 2004;146:265–9. 27. Herzon GD, Zealear DL. Intraoperative monitoring
discussion 269–70. of the visual evoked potential during endoscopic
14. Benedičič M, Bošnjak R. Optic nerve potentials and sinus surgery. Otolaryngol Head Neck Surg.
cortical potentials after stimulation of the anterior 1994;111:575–9.
visual pathway during neurosurgery. Doc 28. Zaaroor M, Pratt H, Feinsod M, Schacham SE. Real-­
Ophthalmol Adv Ophthalmol. 2011;122:115–25. time monitoring of visual evoked potentials. Isr
15. Benedičič M, Bošnjak R. Intraoperative monitoring J Med Sci. 1993;29:17–22.
of the visual function using cortical potentials after 29. Ota T, Kawai K, Kamada K, Kin T, Saito
electrical epidural stimulation of the optic nerve. N. Intraoperative monitoring of cortically recorded
Acta Neurochir (Wien). 2011;153:1919–27. visual response for posterior visual pathway.
16. Landi A, Pirillo D, Cilia R, Antonini A, Sganzerla J Neurosurg. 2010;112:285–94.
EP. Cortical visual evoked potentials recorded after 30. Wilson WB, Kirsch WM, Neville H, Stears J,
optic tract near field stimulation during GPi-DBS in Feinsod M, Lehman RA. Monitoring of visual func-
non-cooperative patients. Clin Neurol Neurosurg. tion during parasellar surgery. Surg Neurol.
2011;113:119–22. 1976;5:323–9.
17. San-Juan D, de Dios Del Castillo CJ, Villegas TG, 31. Koshino K, Kuroda R, Mogami H, Takimoto
Elizondo DL, Torrontegui JAF, Anschel DJ. Visual H. Flashing diode evoked responses for detecting
intraoperative monitoring of occipital arteriovenous optic nerve function during surgery. Med J Osaka
malformation surgery. Clin Neurol Neurosurg. Univ. 1978;29(1–2):39–47.
2011;113:680–2. 32. Cedzich C, Schramm J, Fahlbusch R. Are flash-­
18. Chung SB, Park CW, Seo DW, Kong DS, Park evoked visual potentials useful for intraoperative
SK. Intraoperative visual evoked potential has no monitoring of visual pathway function?
association with postoperative visual outcomes in Neurosurgery. 1987;21:709.
transsphenoidal surgery. Acta Neurochir (Wien). 33. Raudzens PA. Intraoperative monitoring of evoked
2012;154:1505–10. potentials. Ann N Y Acad Sci. 1982;388:308–26.
19. Ogawa Y, Nakagawa A, Washio T, Arafune T, 34. Cedzich C, Schramm J, Mengedoht CF, Fahlbusch
Tominaga T. Tissue dissection before direct manipu- R. Factors that limit the use of flash visual evoked
lation to the pathology with pulsed laser-induced potentials for surgical monitoring. Electroencephalogr
liquid jet system in skull base surgery: preservation Clin Neurophysiol. 1988;71:142–5.
of fine vessels and maintained optic nerve function. 35. Cedzich C, Schramm J. Monitoring of flash visual
Acta Neurochir (Wien). 2013;155:1879–86. evoked potentials during neurosurgical operations.
20. Kamio Y, Sakai N, Sameshima T, Takahashi G, Int Anesthesiol Clin. 1990;28:165–9.
Koizumi S, Sugiyama K, et al. Usefulness of intra- 36. Hamaguchi K, Nakagawa I, Hidaka S, Uesugi F, Kubo
operative monitoring of visual evoked potentials in T, Kato T. Effect of propofol on visual evoked poten-
transsphenoidal surgery. Neurol Med Chir (Tokyo). tials during neurosurgery. Masui. 2005;54:998–1002.
2014;54:606–11. 37. Nakagawa I, Hidaka S, Okada H, Kubo T, Okamura
21. Padalino DJ, Melnyk V, Allott G, Deshaies K, Kato T. Effects of sevoflurane and propofol on
EM. Electroretinography during embolization of an evoked potentials during neurosurgical anesthesia.
ophthalmic arteriovenous fistula. Surg Neurol Int. Masui. 2006;55:692–8.
2013;4:40. 38. Nau HE, Hess W, Pohlen G, Marggraf G, Rimpel J.
22. Goto T, Tanaka Y, Kodama K, Kusano Y, Sakai K, Evoked potentials in intracranial operations: current
Hongo K. Loss of visual evoked potential following status and our experiences [in German]. Anaesthesist.
temporary occlusion of the superior hypophyseal 1987;36:116–25.
68 S.C. Toleikis and J.R. Toleikis

39. Lorenz M, Renella RR. Intraoperative monitoring: 55. Zimmerer R, Schattmann K, Essig H, Jehn P,
visual evoked potentials in surgery of the sellar Metzger M, Kokemüller H, et al. Efficacy of trans-
region [in German]. Zentralbl Neurochir. 1989; cutaneous transseptal orbital decompression in
50:12–5. treating acute retrobulbar hemorrhage and a litera-
40. Chacko AG, Babu KS, Chandy MJ. Value of visual ture review. Craniomaxillofac Trauma Reconstr.
evoked potential monitoring during trans-sphenoidal 2014;7:17–26.
pituitary surgery. Br J Neurosurg. 1996;10:275–8. 56. Zhu Y, Song G, Tang D, Zhai X, Tian W. Monitor
41. Wiedemayer H, Fauser B, Armbruster W, Gasser T, visual function with flash visual evoked potential
Stolke D. Visual evoked potentials for intraopera- during orbital surgery [in Chinese]. Zhonghua Yan
tive neurophysiologic monitoring using total intra- Ke Za Zhi. 2000;36:445–8.
venous anesthesia. J Neurosurg Anesthesiol. 57. Nebbioso M, Lenarduzzi F, Pucci B, Plateroti AM,
2003;15:19–24. Rispoli E. Surgical management by means of elec-
42. Wiedemayer H, Fauser B, Sandalcioglu IE, troretinographic examination during extracorporeal
Armbruster W, Stolke D. Observations on intraop- circulation. Ann Ital Chir. 2012;83:523–8.
erative monitoring of visual pathways using steady-­ 58. Kamada K, Todo T, Morita A, Masutani Y, Aoki S,
state visual evoked potentials. Eur J Anaesthesiol. Ino K, et al. Functional monitoring for visual path-
2004;21:429–33. way using real-time visual evoked potentials and
43. Bergholz R, Lehmann TN, Fritz G, Rüther K. Fourier optic-radiation tractography. Neurosurgery.
transformed steady-state flash evoked potentials for 2005;57(1 Suppl):121–7.
continuous monitoring of visual pathway function. 59. Lozano A, Hutchison W, Kiss Z, Tasker R, Davis K,
Doc Ophthalmol. 2007;116:217–29. Dostrovsky J. Methods for microelectrode-guided
44. Hussain SS, Laljee HC, Horrocks JM, Tec H, Grace posteroventral pallidotomy. J Neurosurg.
AR. Monitoring of intra-operative visual evoked 1996;84:194–202.
potentials during functional endoscopic sinus sur- 60. Yokoyama T, Sugiyama K, Nishizawa S, Yokota N,
gery (FESS) under general anaesthesia. J Laryngol Ohta S, Uemura K. Visual evoked potentials during
Otol. 1996;110:31–6. posteroventral pallidotomy for Parkinson’s disease.
45. Neuloh G. Time to revisit VEP monitoring? Acta Neurosurgery. 1999;44:815–22. discussion 822–4.
Neurochir (Wien). 2010;152:649–50. 61. Cohen BA, Baldwin ME. Visual-evoked potentials
46. Miyake Y, Horiguchi M. Electroretinographic altera- for intraoperative neurophysiology monitoring:
tions during vitrectomy in human eyes. Graefes Arch another flash in the pan? J Clin Neurophysiol.
Clin Exp Ophthalmol. 1998;236:13–7. 2011;28:599–601.
47. Miyake Y, Yagasaki K, Horiguchi M. 62. Thirumala PD, Habeych ME, Crammond DJ, Balzer
Electroretinographic monitoring of retinal function JR. Neurophysiologic intraoperative monitoring of
during eye surgery. Arch Ophthalmol. 1991;109: olfactory and optic nerves. J Clin Neurophysiol.
1123–6. 2011;28:538–42.
48. Montezuma SR, Rizzo JF, Ziv OR. Combined vitrec- 63. Schumann P, Kokemüller H, Tavassol F, Lindhorst
tomy lens and contact electrode for erg recording D, Lemound J, Essig H, et al. Optic nerve monitor-
during surgery. Retina. 2002;22:828–9. ing. Craniomaxillofac Trauma Reconstr. 2013;
49. Kikuchi Y, Sasaki T, Matsumoto M, Oikawa T, 6:75–86.
Itakura T, Kodama N. Optic nerve evoked potentials 64. Moller A. Evoked potentials in intraoperative moni-
elicited by electrical stimulation. Neurol Med Chir toring. New York: Williams & Wilkins; 1988.
(Tokyo). 2005;45:349–55. discussion 354–5. 65. Inui K, Sannan H, Miki K, Kaneoke Y, Kakigi
50. Bošnjak R, Benedičič M. Direct epidural electrical R. Timing of early activity in the visual cortex as
stimulation of the optic nerve: a new method for revealed by simultaneous MEG and ERG record-
intraoperative assessment of function. J Neurosurg ings. Neuroimage. 2006;30:239–44.
Pediatr. 2008;109:647–53. 66. Tremblay F, Parkinson JE. Alteration of electroreti-
51. Tashiro C, Muranishi R, Gomyo I, Mashimo T, Tomi nographic recordings when performed under seda-
K, Yoshiya I. Electroretinogram as a possible moni- tion or halogenate anesthesia in a pediatric
tor of anesthetic depth. Graefes Arch Clin Exp population. Doc Ophthalmol. 2003;107:271–9.
Ophthalmol. 1986;224:473–6. 67. Holder GE. Electrophysiological assessment of
52. Yagi M, Tashiro C, Yoshiya I. Changes in the electro- optic nerve disease. Eye (Lond). 2004;18:1133–43.
retinogram during enflurane anesthesia [in Japanese]. 68. Chiappa K. Evoked potential in clinical medicine.
Masui. 1989;38:1438–43. New York: Raven; 1983.
53. Nogawa T, Katayama K, Okuda H, Uchida 69. Towle VL, Cakmur R, Cao Y, Brigell M, Parmeggiani
M. Changes in the latency of the maximum positive L. Locating VEP equivalent dipoles in magnetic
peak of visual evoked potential during anesthesia. resonance images. Int J Neurosci.
Nihon Geka Hokan. 1991;60:143–53. 1995;80(1–4):105–16.
54. Zimmerer R, Rana M, Schumann P, Gellrich 70. Nehamkin S, Windom M, Syed TU. Visual evoked
N-C. Diagnosis and treatment of optic nerve trauma. potentials. Am J Electroneurodiagnostic Technol.
Facial Plast Surg. 2014;30:518–27. 2008;48:233–48.
4 Visual-Evoked Potentials 69

71. Walsh P, Kane N, Butler S. The clinical role of nary bypass or circulatory arrest in neonates and
evoked potentials. J Neurol Neurosurg Psychiatry. infants. Can J Anaesth. 1993;40:298–307.
2005;76 Suppl 2:ii16–22. 86. Nenekidis I, Pournaras CJ, Tsironi E, Tsilimingas
72. Harding GF, Smith VH, Yorke HC. A contact lens pho- N. Vision impairment during cardiac surgery and
tostimulator for surgical monitoring. Electroencephalogr extracorporeal circulation: current understanding
Clin Neurophysiol. 1987;66(3):322–6. and the need for further investigation. Acta
73. American Electroencephalographic Society. Ophthalmol. 2012;90:e168–72.
Guidelines for intraoperative monitoring of sensory 87. Brandli A, Stone J. Remote ischemia influences the
evoked potentials. J Clin Neurophysiol. 1987; responsiveness of the retina: observations in the rat.
4:397–416. Invest Ophthalmol Vis Sci. 2014;55:2088–96.
74. Kriss A, Halliday AM, Halliday E, Pratt RT. Evoked 88. Møller AR, Burgess JE, Sekhar LN. Recording
potentials following unilateral ECT. II. The flash compound action potentials from the optic nerve in
evoked potential. Electroencephalogr Clin man and monkeys. Electroencephalogr Clin
Neurophysiol. 1980;48:490–501. Neurophysiol. 1987;67:549–55.
75. Noonan BD, Wilkus RJ, Chatrian GE, Lettich E. The 89. Curatolo JM, Macdonell RA, Berkovic SF, Fabinyi
influence of direction of gaze on the human electroret- GC. Intraoperative monitoring to preserve central
inogram recorded from periorbital electrodes: a study visual fields during occipital corticectomy for epi-
utilizing a summating technique. Electroencephalogr lepsy. J Clin Neurosci. 2000;7:234–7.
Clin Neurophysiol. 1973;35:495–502. 90. Tobimatsu S, Shima F, Ishido K, Kato M. Visual
76. Rubinstein MP, Harding GF. The visually evoked evoked potentials in the vicinity of the optic tract
subcortical potential: is related to the electroretino- during stereotactic pallidotomy. Electroencephalogr
gram? Invest Ophthalmol Vis Sci. 1981;21:335–44. Clin Neurophysiol. 1997;104:274–9.
77. Esakowitz L, Kriss A, Shawkat F. A comparison of 91. Benedičič M, Beltram M, Olup BD, Bošnjak R.
flash electroretinograms recorded from Burian Cortical potentials after electrical intraneural stimula-
Allen, JET, C-glide, gold foil, DTL and skin elec- tion of the optic nerve during orbital enucleation. Doc
trodes. Eye (Lond). 1993;7(Pt 1):169–71. Ophthalmol Adv Ophthalmol. 2012;125:195–202.
78. Gur M, Gath I. Time and frequency analysis of 92. Russ W, Kling D, Loesevitz A, Hempelmann G.
simultaneously recorded corneal and non-corneal Effect of hypothermia on visual evoked potentials
electroretinogram. J Biomed Eng. 1979;1:172–4. (VEP) in humans. Anesthesiology. 1984;61:207–10.
79. Hood DC, Bach M, Brigell M, Keating D, Kondo M, 93. Wongpichedchai S, Hansen RM, Koka B, Gudas VM,
Lyons JS, et al. ISCEV guidelines for clinical multi- Fulton AB. Effects of halothane on children’s electro-
focal electroretinography (2007 edition). Doc retinograms. Ophthalmology. 1992;99:1309–12.
Ophthalmol. 2008;116:1–11. 94. Raitta C, Karhunen U, Seppäläinen AM. Changes in
80. Nuwer MR, Dawson EC. Intraoperative evoked the electroretinogram and visual evoked potentials
potential monitoring of the spinal cord. A restricted during general anaesthesia using enflurane. Graefes
filter, scalp method during Harrington instrumenta- Arch Clin Exp Ophthalmol. 1982;218:294–6.
tion for scoliosis. Clin Orthop. 1984;183:42–50. 95. Iohom G, Collins I, Murphy D, Awad I, O’Connor
81. Martinez Piñeiro A, Cubells C, Garcia P, Castaño C, G, McCarthy N, et al. Postoperative changes in
Dávalos A, Coll-Canti J. Implementation of intraop- visual evoked potentials and cognitive function tests
erative neurophysiological monitoring during endo- following sevoflurane anaesthesia. Br J Anaesth.
vascular procedures in the central nervous system. 2001;87:855–9.
Interv Neurol. 2015;3:85–100. 96. Iohom G, Whyte A, Flynn T, O’Connor G, Shorten
82. Keenan NK, Taylor MJ, Coles JG, Prieur BJ, G. Postoperative changes in the full-field electroret-
Burrows FA. The use of VEPs for CNS monitoring inogram following sevoflurane anaesthesia. Eur
during continuous cardiopulmonary bypass and cir- J Anaesthesiol. 2004;21:272–8.
culatory arrest. Electroencephalogr Clin 97. Andréasson S, Tornqvist K, Ehinger B. Full-field
Neurophysiol. 1987;68:241–6. electroretinograms during general anesthesia in nor-
83. Reilly EL, Kondo C, Brunberg JA, Doty DB. Visual mal children compared to examination with topical
evoked potentials during hypothermia and pro- anesthesia. Acta Ophthalmol. 1993;71:491–5.
longed circulatory arrest. Electroencephalogr Clin 98. Tanskanen P, Kylmä T, Kommonen B, Karhunen
Neurophysiol. 1978;45:100–6. U. Propofol influences the electroretinogram to a
84. Markand ON, Warren CH, Moorthy SS, Stoelting lesser degree than thiopentone. Acta Anaesthesiol
RK, King RD. Monitoring of multimodality evoked Scand. 1996;40:480–5.
potentials during open heart surgery under hypo- 99. Chi OZ, McCoy CL, Field C. Effects of fentanyl
thermia. Electroencephalogr Clin Neurophysiol. anesthesia on visual evoked potentials in humans.
1984;59:432–40. Anesthesiology. 1987;67:827–30.
85. Burrows FA, Bissonnette B. Cerebral blood flow 100. Loughnan BL, Sebel PS, Thomas D, Rutherfoord
velocity patterns during cardiac surgery utilizing CF, Rogers H. Evoked potentials following diaze-
profound hypothermia with low-flow cardiopulmo- pam or fentanyl. Anaesthesia. 1987;42:195–8.
70 S.C. Toleikis and J.R. Toleikis

101. Sloan T. Anesthesia and intraoperative neurophysio- potential as a function of inspired nitrous oxide con-
logical monitoring in children. Childs Nerv Syst. centration. Electroencephalogr Clin Neurophysiol.
2009;26:227–35. 1984;57:178–83.
102. Uhl RR, Squires KC, Bruce DL, Starr A. Effect of 108. Hou WY, Lee WY, Lin SM, Liu CC, Susceto L, Sun
halothane anesthesia on the human cortical visual WZ, Lin SY. The effects of ketamine, propofol and
evoked response. Anesthesiology. 1980;53:273–6. nitrous oxide on visual evoked potentials during
103. Chi OZ, Field C. Effects of isoflurane on visual fentanyl anesthesia. Ma Zui Xue Za Zhi.
evoked potentials in humans. Anesthesiology. 1993;31:97–102.
1986;65:328–30. 109. Chi OZ, Ryterband S, Field C. Visual evoked poten-
104. Kameyama Y. Effect of isoflurane and sevoflurane tials during thiopentone-fentanyl-nitrous oxide
on evoked potentials and EEG. Jpn J Anesth. anaesthesia in humans. Can J Anaesth.
1994;43:657–64. 1989;36:637–40.
105. Sebel PS, Flynn PJ, Ingram DA. Effect of nitrous 110. Russ W, Luben V, Hempelmann G. Der EinfluB der
oxide on visual, auditory and somatosensory evoked Neuroleptanalgesie auf das visuelle evozierte
potentials. Br J Anaesth. 1984;56:1403–7. Potential (VEP) des Menschen. Anaesthesist. 1982;
106. Yamashiro H. Differentiation of brain stem anesthe- 31:575–8.
sia from high spinal anesthesia using auditory brain 111. Makela K, Harkainen K, Rorarius M, Jantti V.
stem response. Masui. 1990;39:1704–7. Suppression of F-VEP during isoflurane-induced
107. Fenwick PBC, Stone SA, Bushman J, Enderby EEG suppression. Electroencephalogr Clin
D. Changes in the pattern reversal visual evoked Neurophysiol. 1996;100:269–72.
 Deep Brain Stimulation
5
Jay L. Shils, Diana Apetauerova, Amal A. Mokeem,
and Jeffrey E. Arle

Key Learning Points

Introduction
• The role of the anesthesiologist in the DBS
case is critical to maintain patient comfort and Intraoperative monitoring (IOM) may generally
cardiovascular support in an awake patient be separated into two categories: (1) evaluating
without the use of common sedative real-time data to detect adverse changes in the ner-
medications. vous system, giving the surgical team a chance
• In complex movement disorder procedures, either to reverse or to stop what is causing the
the use of sedation is almost always necessary change; (2) evaluating real-time data to help deter-
and the fine balance between these medica- mine related physiologic localization or guidance
tions and the consciousness of the patient is for the surgical team through specific procedural
critical to allowing the physiologists to obtain steps of a particular surgery. In both categories,
the data necessary for optimal electrode the surgical, anesthetic, and neuromonitoring per-
placement. sonnel play a role in how the data are interpreted
and incorporated during the procedure. Deep
brain stimulation is challenging in that the patient
needs to be comfortable enough to be alert through
the complete procedure, yet not affect the proper-
J.L. Shils, PhD, D. ABNM, FASNM, FACNS (*) ties of the single unit recordings.
Department of Anesthesiology, Rush University
Deep brain stimulation (DBS) surgery, with
Medical Center, 1653 W. Congress Pkwy, Suite
1483 Jelke Building, Chicago, IL 60612, USA the use of microelectrode recording, for move-
e-mail: [email protected] ment disorders falls into the second category. For
D. Apetauerova, M.D. the anesthesiologist, the challenge is that most
Department of Neurology, Lahey Hospital and Health anesthetics used to minimize pain and sedate a
System, 41 Mall Road, Burlington, MA 01805, USA patient affect the firing patterns of the neurophys-
A.A. Mokeem, M.D. iologic data necessary to locate functional targets
Department of Neurosciences, MBC 76, King Faisal during the surgery, which in turn makes localiza-
Specialist Hospital & Research Center,
tion based on the firing patterns more difficult. In
3354, Riyadh 11211, Saudi Arabia
addition, the patient is likely to be fully awake
J.E. Arle, M.D., Ph.D.
and in their “worst” clinical state, which further
Department of Neurosurgery, Beth Israel Deaconess
Medical Center and Harvard University, complicates the situation. These issues make the
300 Brookline Avenue, Boston, MA 02215, USA communication between the anesthesiologist,

© Springer International Publishing AG 2017 71

A. Koht et al. (eds.), Monitoring the Nervous System for Anesthesiologists
and Other Health Care Professionals, DOI 10.1007/978-3-319-46542-5_5
72 J.L. Shils et al.

surgeon, and neurophysiologist critical, and not rant firing patterns caused by the disbalance of
something that should occur only in the morning the direct and indirect BG pathways. The exact
before the procedure starts. A dedicated member mechanism by which DBS affects this renormal-
of each of those three disciplines, who is always ization is unknown.
present, is preferable in making this procedure We present three different scenarios to help
successful in every case. illustrate important points. The first will involve
The most common movement disorders that are a classic parkinsonian patient undergoing a
treated with DBS are ones that are associated with straightforward placement of a DBS electrode in
some abnormality of the basal ganglia (BG). The the STN with minimal anesthetic intervention.
BG are a group of six nuclei in the brain with two The second involves a complex dystonic patient
principal input structures: the corpus ­ striatum who required continuous changes in anesthetic
(Striatum) and the subthalamic nucleus (STN), two management throughout the procedure. The third
output structures (internal segment of the globus involves a patient with a medication-induced
pallidus [GPi] and substantia nigra pars reticulata movement disorder that was on a continuous
[SNr]), and two intrinsic nuclei (external segment infusion of propofol prior to surgery to reduce the
of the globus pallidus [GPe] and substantia nigra adverse effects of the movement disorder. In
pars compacta [SNc]). The striatum receives excit- addition to these three examples, we also include
atory (glutamatergic) input from multiple areas of a basic surgical methodology common to all DBS
the cerebral cortex as well as feedback inhibitory procedures for movement disorders.
and excitatory input from the dopaminergic cells of
the SNc. One subset of these cells projects directly
to the GPi, forming the “direct pathway,” while Surgery
another subset projects to the GPe, the first relay
station of a complementary “indirect pathway,” Movement disorders surgery is performed stereo-
which passes through the STN before terminating tactically. This means the brain is placed in a
at GPi. The antagonistic actions of the direct and three-dimensional (3D)-coordinate space that
indirect pathways regulate the neuronal activity of allows accurate and reliable targeting of specific
GPi, which in turn provides inhibitory input to the anatomic areas. Historically, the most common
pedunculopontine nucleus (PPN) and the ventrolat- stereotactic systems consist of a frame that is
eral (VL) nucleus of the thalamus, which contains attached to the outer table of the patient’s skull
the sensory receiving area (ventral caudal nucleus followed by either a computed tomography (CT)
[VC]) and the cerebellar receiving area (ventralis scan or magnetic resonance imaging (MRI) scan.
intermedius nucleus [VIM]). The VL nucleus proj- A more recent technique uses a smaller stereotac-
ects back to the primary and supplementary motor tic platform (FHC Starfix system, Bodenheim,
areas, completing the cortico-­ ganglio-­
thalamo- ME; or Medtronic Nexframe, Minneapolis, MN)
cortical loop. The direct pathway inhibits GPi, where trajectory guidance is done using “frame-
resulting in a net disinhibition of the motor thala- less” stereotactic techniques. These new “frame-
mus and facilitation of the thalamocortical projec- less” techniques offer anesthesiology advantages
tions. The indirect pathway, via its serial of easy access to the patient’s airway. The stan-
connections, provides excitatory input to the GPi, dard frames make access to the patient’s airway
inhibiting the thalamocortical motor pathway. more difficult due to frame components making
In various movement disorders, the signaling emergent mask access difficult, so the use of a
between the complementary balance between the laryngeal mask airway (LMA) should be consid-
direct and indirect pathways is disrupted, which ered. Once the imaging has been performed, the
results in clinical symptoms. In Parkinson’s dis- surgeon calculates the target location in the 3D
ease (PD), the disruption results in a hypokinetic space. Due to anatomical variations, imaging dis-
clinical picture, whereas for dystonia the disrup- tortions, and functional neurophysiologic differ-
tion results in a hyperkinetic clinical picture. ences among patients, this initial targeting acts
DBS is thought to act by renormalizing the aber- only as a guide to reach the target structure. Once
5 Deep Brain Stimulation 73

targeting is complete, the surgeon creates a gery. Bladder catheters are placed in all patients
14-mm burr hole in the skull to accommodate the to save patients from worrying about voiding
electrode trajectory and readies the neurophysiol- during the procedure and because many move-
ogy recording equipment attached to the stereo- ment disorder patients can have difficulty mov-
tactic frame. ing after surgery. Because the patient is in a
Microelectrode recording (MER) is best per- modified sitting position with the head above the
formed with the patient’s full, nonmedicated chest, the occurrence of a venous air embolism
attention. After MER is performed and the appro- should always be considered a possibility; par-
priate functional target is located, the permanent ticularly with the burr hole allowing for surface
DBS electrode is placed and a second neurophys- vein exposure. Standard pulse oximetry and end-
iologic test using stimulation, which mimics the tidal CO2 monitoring are helpful in this regard.
DBS therapy, takes place. The wound is then Pulse oximetry monitoring, IV access, and an
closed and the controlling implantable pulse gen- arterial line (if used) should be placed on the
erator (IPG) is implanted. hand or foot that is on the same side as the sur-
The most serious complication during a DBS gery, so as to allow unhindered examination of
procedure is a fatal hemorrhage [1]. During a stan- the tremor, rigidity, and movement speed in the
dard DBS procedure, blood pressure (BP) control contralateral limb. If a venous air embolism is
is a critical anesthetic concern during both micro- suspected or detected, the subdural space should
recording and permanent lead placement. While be continually flushed with saline to avoid fur-
symptomatic hemorrhages may be readily ther air entering the venous system. All stereotac-
detected, asymptomatic hemorrhages may yet tic frames allow for the placement of endotracheal
presage clinical demise, so postoperative imaging tubes or laryngeal mask airways without remov-
is highly recommended [2]. To minimize the ing the frame. Specific wrenches for frame
chance of intraoperative hemorrhage, systolic BP removal and adjustment should be kept handy for
should be kept below 150 mmHg. If systolic pres- potential emergency use.
sure rises above this level, the procedure is halted Although sedatives and systemic analgesic
until the systolic BP returns below 150 mmHg. For anesthetics are categorically contraindicated prior
patients with PD there may be rebound hyperten- to and during the parts of the procedure when neu-
sion when dopaminergic agonists have been halted rophysiologic data are being acquired, there are
prior to the procedure, often making BP control times when both are beneficial and even necessary.
more challenging. Several standard BP control When the stereotactic frame is being placed on the
medications have been demonstrated to cause no patient’s head, a local anesthetic (e.g., lidocaine/
hindering effects on single-unit recordings. These bupivacaine) is injected subcutaneously at the
include most notably hydralazine, nimodipine, sites where the pins will be placed. At our center,
nitroglycerine, and sodium nitroprusside (authors’ we also use sedation (propofol boluses of
experience and Venkatraghavan [3]), although 20–50 mg/kg or constant infusions of ∼ 100 μg/kg/
there are no specific references in the literature min) during frame placement, incision, burr hole,
describing the effects in detail. The one report that and opening of the dura. Due to varying degrees of
discusses the effect of the β-blocker metoprolol tolerance in some patients, higher doses (as much
demonstrated a significant reduction in STN spik- as 250 μg/kg/min) may be needed (see example 3).
ing activity with a transient reduction in rigidity A second anesthetic that has been shown to be
after intravenous administration to reduce BP in acceptable during such procedures is dexmedeto-
three patients [4]. Some clinicians recommend midine (Dex). For these procedures, ideally it is
avoiding the use of β-blockers because these may important to use an anesthetic that is metabolized
reduce the ability to assess the movement disorder rapidly (propofol) and has minimal effects on sin-
during the electrode placement. gle unit recordings. Anesthetics may be used with-
Dehydration may be a concern in longer DBS out restrictions during wound closure. On
procedures. Intravenous (IV) fluids at mainte- occasion, it may also be necessary to use small
nance levels should be given throughout the sur- doses of anesthetics for brief periods to sedate
74 J.L. Shils et al.

patients between recording tracts. As described in Dex is a good choice for dystonic patients and has
the second case below, there are also special cir- been successfully used in pediatric patients [11]. A
cumstances in which anesthetics are needed even critical reason for choosing Dex is its minimal
while performing MER. effect on single-unit MER recording data, as one
Because anesthetics can alter neuronal firing can appreciate that neuronal activity is still readily
frequency [5] and impair patient assessment, the obtainable (Fig. 5.1). Dex reliably produces con-
use of gabaminergic sedative medications, even scious sedation while the patient remains respon-
in small doses, has been shown to affect the qual- sive and cooperative to verbal commands [2, 12].
ity of MER adversely [6]. Temporary modifica- One reason that Dex most likely does not affect
tion and suppression of parkinsonian tremor [7, BG recordings is that its effect is on the α2-
8] and increase in dyskinesias [9] have been adrenoreceptors in the locus coeruleus, a major
reported with the use of propofol and remifent- site of noradrenergic innervation in the central ner-
anil and thus, at our center, we require at least vous system, and not on the gabaminergic recep-
10 min for the propofol to wear off prior to tremor tors of the BG. The locus coeruleus has been
testing. It is unclear which general anesthetic implicated as a key modulator for a variety of criti-
agents allow the most effective MER because all cal brain functions, including arousal, sleep, and
of them affect recordings to some extent [6, 10] anxiety [13]. This, together with minimal respira-
and no studies comparing their effects on data tory depression, makes it an attractive agent to use
acquisition have been performed intraopera- in “awake” functional craniotomies. Low-dose
tively. An “awake” technique has obvious advan- infusion of this drug provides sedation from which
tages and most centers avoid anesthesia at least patients are easily arousable and cooperative to
during the recording and mapping phase of DBS verbal stimulation. Consequently, there are several
electrode placement in order to detect cellular reports on the successful use of the drug, both
activity and movement-related responses to alone [3, 14, 15] and in combination with intermit-
neurostimulation. Another common technique tent propofol [16]. Dex has also been shown to
used to improve intraoperative recordings and attenuate the hemodynamic and neuroendocrine
patient assessment is the withholding of antipar- responses to headpin insertion (the method for
kinsonian medications beginning the night before attaching the frame to the patient’s head) in
surgery; this can be unpleasant for the patient but patients undergoing craniotomy and to signifi-
is necessary to assure optimal DBS electrode cantly reduce the concomitant use of antihyperten-
placement. On the other end of the movement sive medication [17, 18]. It can theoretically
disorder spectrum are dystonic patients who usu- decrease cerebral blood flow via direct α2-mediated
ally do need sedation during the initial phase of vascular smooth muscle constriction and, indi-
surgery before neurophysiologic testing, and rectly, via effects on the intrinsic neural pathways
potentially even during the testing phases. modulating vascular effects. α2-Agonists have a
Ideally, any sedative effects should be readily more potent vasoconstrictor effect on the venous
reversible. Benzodiazepines and opioids with rather than on the arteriolar side of the cerebral
longer half-­lives should be avoided. Opioids can vasculature and can, therefore, decrease ICP. There
also cause agitation, muscular rigidity, sweating, is, so far, no evidence of adverse effects on cere-
and hyperpyrexia in combination with some PD bral hemodynamics associated with its use, even
medications (e.g., selegiline) [8]. in the setting of a compromised cerebral circula-
In PD patients, propofol has been used exten- tion; nor does Dex ameliorate the clinical signs of
sively, but its use requires vigilance to ensure Parkinson’s disease, such as tremor, rigidity, and
patient alertness. Erring on the side of less medica- bradykinesia. The pharmacologic profile of Dex
tion ensures that its effects wear off more quickly suggests that it may be an ideal sedative drug for
and decreases the risk of airway loss. The surgeon DBS implantation [19]. General anesthesia (GA)
may use extra local anesthetic with the scalp inci- is an option reserved for patients with severe dis-
sion to reduce surgical pain. Also, as stated earlier, orders for whom awake surgery would compro-
5 Deep Brain Stimulation 75

Fig. 5.1 Single-unit recordings under two different anes- using low-dose Dex (∼0.1 μg/kg/h); b and d were
thetics. Each tracing shows 1 s of data. Tracings recorded recorded using low-dose continuous infusions of propofol
from the same patient during a bilateral DBS STN proce- (<20 μg/kg/min). The recordings with propofol are much
dure (a and b). Tracings recorded during a GPi DBS in less robust than the ones during Dex infusions
two different patients (c and d). a and c were recorded

mise their safety. These include patients with their usual antiparkinsonian medication as
dystonia and related disorders who are on multiple soon as possible after the procedure to avoid
medications to prevent contractions and spasms. possible deterioration in neurologic function
Another group of patients are those that may not and respiratory muscle impairment.
be able to tolerate the procedure, have severe There is limited information on the incidence
respiratory conditions, or bad tics (such as patients of intraoperative anesthetic complications during
with Tourette syndrome) that preclude them from these procedures. A review of intraoperative anes-
remaining relatively motionless. thetic-related complications in a series of 158
After the DBS electrodes are inserted, cases of deep brain ablation or stimulation under
sedation can be increased and the frame sedation with propofol or Dex [20] found that
removed. Implantation of the pulse generator intraoperative events occurred in 6.96 % of cases.
and internalization of electrodes can be per- These events included coughing, sneezing, aspi-
formed either immediately or as second-stage ration, pulmonary edema, combative behavior
surgery under GA. Patients should receive and agitation/confusion, bronchospasm, angina,
76 J.L. Shils et al.

and intracranial hemorrhage. All of these have the asking the patient to move a particular body part
potential of moving the electrodes and cannula in while looking for changes in the firing pattern of
the brain and causing an intraparenchymal hem- the single unit under study. Kinesthetic testing is
orrhage. Yet, with an anesthesiologist who under- performed by moving an isolated joint and look-
stands the specific movement disorder of the ing for changes in the single unit’s firing rate.
patient, the surgical procedure, the need for qual-
ity, and who pays constant attention to the alert-
ness of the patient, these effects can be minimized.  arget Structures for the Case
T
In our experience, the only case of a complication Examples
was when the anesthesiologist was concentrating
on the infusion numbers and not the patient. For movement disorders surgery, there are three
common targets: (1) the ventral intermediate
nucleus (VIM) of the thalamus, (2) the internal
The MER Procedure globus pallidum (GPi), and (3) the STN. Each of
these targets is thought to best treat a specific
Due to accuracy problems with direct CT or MRI symptom of a movement disorder of a specific
targeting, visualization is generally only a first disease, although research is still underway to
step in locating the target, whereas MER gives a prove these hypotheses. Two of the three cases
much more detailed spatial and functional map. that will be presented in this chapter involve
Several excellent descriptions of the MER tech- placing DBS electrodes in the GPi and, for the
nique exist [21–32]. Understanding the type of other case, in the STN. These cases were chosen
neuronal activity from regions immediately adja- because they demonstrate the extremes of move-
cent to the intended target is critical to the suc- ment disorder surgery. VIM is the primary target
cess of the procedure. Following are two case for tremor-related diseases such as essential
examples (one fairly typical and the other unusual tremor and one can still use the lessons from the
and complex) illustrating the integration of diag- cases below in a VIM case because it is the
nosis, technique, and anesthetic management. patient that dictates the anesthetic intervention
During a MER recording session, not only are and not the disease. To get a better feel for the
recordings acquired from the target location but procedures, we have included a short description
also from structures located above and below that of what type of physiology is encountered to
location. These recordings from other structures show what can be lost if the anesthetic technique
are useful in helping to determine the correct sag- is not appropriately applied.
ittal, coronal, and axial positions of the target. At Although the complete GPi may be visualized
the time of the writing of this chapter, both on an MRI, the functional target location in the
patients are significantly improved and are still GPi is in the posterior and ventral region of the
benefiting from the procedure. In addition to nucleus [33–39]. As the microelectrode is lowered
looking for spontaneous single cell firings, it is into the brain along its recording trajectory toward
important to look for cells that respond to specific the target location, the firing patterns from three
types of evoked activity given that the basal gan- anatomical structures must be recognized to con-
glia is composed of segments that are nonsensory firm optimal placement: the striatum, the external
motor. Finding cells that respond to voluntary globus pallidum (GPe), and the internal GPi.
patient movement and cells that respond to limb Figure 5.2 shows the anatomy of the GPi and the
joint position (kinesthetic) are key in making surrounding structures. The recordings on the right
sure the electrode will be placed in the sensory show representative firing patterns from each area.
motor region of the specific target. Both volun- These are the critical physiologic markers in dif-
tary and kinesthetic cells will either increase or ferentiating the structures. During placement of
decrease their firing rate during the activity and the DBS electrode, proximity to certain structures
are only found in the sensory motor region of the must be avoided such as the optic tract and internal
specific nuclei. Voluntary testing is performed by capsule, which can render the therapy useless.
5 Deep Brain Stimulation 77

Fig. 5.2 A sagittal image, ∼21.5 mm from midline, Each trace represents single-unit activity as recorded from
through the GPi and associated anatomy. Traces to the a representative cell in that structure [68]
right are representative firing patterns from each area.

After the microelectrode mapping is per- the STN and the surrounding structures. The
formed, micro- or macroelectrode stimulation is waves on the right show representative firing pat-
performed prior to permanent DBS placement to terns from each area, which are the critical physi-
ensure that the electrode is at a safe distance from ologic markers in differentiating the structures. If
the internal capsule and the optic tract. If the the DBS electrode is placed too medial and pos-
patient is awake and alert, they can easily respond terior within the STN, it can activate the sensory
by indicating when they see flashing lights or thalamus and/or medial lemniscus; if it is placed
experience muscle contractions. If the patient too lateral and anterior, it can adversely affect the
cannot respond, they cannot indicate when the internal capsule and render the therapy useless.
optic tract is being activated or if they are having
muscle tightness. In these cases, EMG recordings
are used to indicate if muscles are being activated Cases and Disorders
by either direct stimulation of the motor fibers in
the internal capsule, which is an adverse effect, Case 1: Noncomplex (PD-STN): A 60-year-old
or the contractions are not a direct result of the male with a 10-year history of PD, well con-
stimulation but an indirect effect of stimulation at trolled on Sinemet until age 59.
the appropriate location in the GPi. Thus, it is Clinical symptoms of PD began on the patient’s
critical that no muscle relaxant be used. right side starting with tremors in the upper extrem-
The functional target in the STN is in the mid- ity and rigidity in both the upper and lower extrem-
dle of the structure between 10.5 and 13.0 mm ities. As the disease progressed, symptoms became
lateral from midline [28, 32, 40–42]. Once again, as problematic on the left side and started to affect
as the microelectrode is slowly lowered toward the patient’s ability to ambulate. One year prior to
its target location, the neuronal firing patterns surgery, levodopa-induced dyskinesias began in
from three anatomical structures must be recog- the head and face, followed by a progressive diffi-
nized to confirm optimal placement: the thala- culty swallowing. At the time of the surgery, the
mus, the STN, and the substantia nigra pars patient was on Stalevo (a mixture of carbidopa,
reticulata (SNr). Figure 5.3 shows the anatomy of levodopa, and entacapone) 37/5/150/200 q.i.d.,
78 J.L. Shils et al.

Fig. 5.3 A sagittal image, ∼12.5 mm from midline, Each trace represents single-unit activity as recorded from
through the STN and associated anatomy. Traces to the a representative cell in that structure. For the SNr, the
right are representative firing patterns from each area. recording is from multiple units [68]

Requip 4 mg t.i.d., Amantadine 100 mg t.i.d., and medication-induced dyskinesia, medication

Neurontin,300 mg t.i.d. Due to the medication-­ refractory tremor, or intolerance to medication.
induced symptom of dyskinesia and the minimal Levodopa-sensitive symptoms may be more likely
amount of benefit, surgical implantation of a DBS to respond to surgery [46], although in our experi-
electrode in the STN was then planned. ence, surgery in the STN and the GPi for PD has
demonstrated a benefit to dopa-induced symptoms
[41, 42]. Continued refinement of the knowledge
Parkinson’s Disease of BG circuitry and PD pathophysiology has nar-
rowed the focus of movement disorder surgery to
PD is a slowly progressive degenerative disorder three nuclei: (1) the thalamus, (2) GPi, and (3) the
of the BG. Nerve cells in SNc produce dopamine, STN. The STN is the preferred surgical target for
which is transported to the input of the BG (stria- DBS electrode placement in PD [41, 47–50].
tum). In PD, for reasons not yet understood, the Serious complications such as hemorrhagic bleed-
dopamine-producing nerve cells of the substantia ing associated with STN-DBS electrode place-
nigra die off. The clinical signs of tremor, bradyki- ment is relatively uncommon [51]. Hypertension
nesia, and rigidity do not fully become apparent must be treated prior to surgery because of the risk
until significant dopaminergic neuronal cells are of hemorrhage [52, 53]. PD patients commonly
lost [43–45]. Medications are the first line of treat- suffer from orthostatic hypotension, contributed to
ment to alleviate symptoms of PD, yet in many by the use of levodopa and dopamine agonists, as
patients who have been responding to medica- well as other autonomic disturbances [54–56].
tions, their symptoms usually begin to gradually Respiratory dysfunction is well known in PD [55].
worsen with time. As they become more pro- This includes an obstructive ventilation pattern,
nounced, patients may start to have difficulty dysfunction of upper airway musculature, rigidity,
walking, talking, or completing other simple tasks. bradykinesia, and dystonia of respiratory muscles
Surgery should be considered when the patient [57]. These problems are exacerbated by with-
develops moderate to severe motor fluctuation, drawal from antiparkinsonian medications.
5 Deep Brain Stimulation 79

Procedure and Decisions were given until the systolic BP dropped to

135 mmHg. Subsequent stimulation and record-
The patient arrived at the hospital on the morning ing required no changes or additions to the anes-
of surgery “off” of all PD medications from thesia. All of our PD patients have nasal cannula
7:00 p.m. of the night before. During frame O2 administered throughout the procedure and
placement, propofol was given in 20 mg boluses SpO2 monitoring.
for sedation and monitored by the anesthesiolo- The DBS electrode was placed in the first
gist. A foley catheter was also placed while the recording tract on the left side and the second
patient was sedated. The patient was then taken recording tract on the right side. These place-
to the CT scanner (an MRI was performed at an ments were chosen due to the number of kines-
earlier visit) with propofol given as needed to thetic cells and length of STN encountered. Once
keep the patient relaxed. After the CT, the patient the DBS electrodes were placed, they were tested
was brought to the OR, transferred to the bed and with an externalized stimulator (Medtronic Dual
positioned with the frame also locked to the bed. 7240 stimulator, Minneapolis, MN). Testing is
It is important that the patient feel comfortable performed in a sequential bipolar fashion (−0,+1:
with both their ability to breath as well as with −1,+2; −2,+3) using a pulse width of 60 μs and a
the position of their neck and back because they frequency of 180 Hz. Voltage is slowly raised to
will be locked in that position for the duration of 4 V. For this patient, no continuous adverse effects
the surgical procedure, which can be several were noted with stimulation up to 4 V. There were
hours. When the patient is comfortable, either a some transient sensory paresthesias in the arm
propofol infusion or bolus doses are given until that lasted for about 5–15 s. Transient adverse
the dura is opened. Once the dura is incised effects are acceptable since the device is never
(about 10–15 min before the MER is to start), the supposed to be turned off. We were able to get
propofol is stopped to allow the patient to be improvements in bradykinesia and rigidity at the
awake for testing. Recording tracts in this patient −1,+2 for the left side and both −0,+1 and −1,+2
included one on the left side and three on the on the right side. Postoperatively the patient was
right side. This difference in the number of improved by 72 % based on the unified Parkinson’s
recording tracts can be caused by potential asym- disease rating scale part III (a common motor
metries in anatomy, effects of nonlinear errors in classification system for PD patients).
imaging, or brain shift during the procedure. It is Case 2: Complex: A 14-year-old boy with
hard to pinpoint the exact reason, but in about methylmalonic acidemia (MMA) was diagnosed
15 % of our cases we find this discrepancy. Each at age 3 months. His condition was well con-
move requires the surgeon to remove the record- trolled and in good health until acquiring H1N1,
ing system and electrode from the head and place when he subsequently developed pancreatitis,
a new tract in the brain. Any time an electrode or sepsis, and in turn bilateral BG strokes.
cannula is placed in the brain, the chance for As a consequence of the strokes, he developed
hemorrhage increases. Thus, it is critical to keep spastic quadriparesis for which a baclofen pump
the BP below 150 mmHg systolic. On the left (a common treatment for spasticity) was placed
side, 4.9 mm of STN were encountered with four without benefit. He was admitted for worsening
kinesthetic cells. On the right side, the first tract dystonia, and resistance to multiple medical ther-
had 4.6 mm of questionable STN and no kines- apies. Due to the severity and worsening d­ ystonia,
thetic cells. The second tract had 5.7 mm of STN including fixed posturing and dynamic spasms, it
with three kinesthetic cells, and the third tract was decided to move forward with bilateral GPi
had 1.3 mm of STN with no kinesthetic cells. At stimulation. During the time between the baclofen
the start of the second tract on the right side, the trial and the decision to move forward with DBS
systolic BP increased above 150–165 mmHg. electrode implantation, the patient’s respiratory
Recordings were halted and 10 mg of labetolol status deteriorated somewhat as well.
80 J.L. Shils et al.

Dystonia workers all met to plan the pre-, intra-, and post-
surgical management of the patient. Each group
Dystonia refers to a syndrome of involuntary sus- discussed their particular needs for the case and
tained or spasmodic muscle contractions involv- the effects each would have on the patient. As
ing cocontraction of the agonist and the antagonist discussed previously, the most common issue
muscles [58–60]. The movements are usually facing the anesthesiologist in most DBS cases is
slow and sustained, and they often occur in a BP control. A plan was made to use Dex initially
repetitive and patterned manner. However, they and if that proved unacceptable, then propofol
can be unpredictable and fluctuate. The frequent would be used if the patient could tolerate it. Due
abnormal posturing and twisting can be painful to the underlying metabolic issues, it was ques-
and functionally disabling. Regardless of the tioned whether a controlled amount of Dex could
causes, the dystonic contractions can have a be given to allow for patient comfort and also the
chronic course and can lead to severe persistent ability to record. Also, one of the main problems
pain and disability. Because each type of dysto- with Dex is the potential for causing hypotension
nia is treated in a different manner, the distinction in a patient. This is not a major issue in PD
between the various types is therapeutically patients, where this effect is usually helpful, but
important [61–66]. On the basis of its clinical in children, it needs to be a consideration. Three
distribution, dystonia is classified as focal dysto- days prior to the procedure, the plan was to try
nia, segmental dystonia, multifocal dystonia, and wean the patient off of some medications,
generalized dystonia, and hemidystonia. including high doses of benzodiazepines, which
Systemic medications benefit about one-third proved unsuccessful.
of patients and consist of a wide variety of On the morning of surgery the patient was
options, including cholinergics, benzodiazepines, taken directly to the CT scanner and intubated.
antiparkinsonism drugs, anticonvulsants, Anesthesia included 3 mg of midazolam, 10 mg
baclofen pump, carbamazepine, and lithium [67]. of etomidate initially followed by another 5 mg,
Many patients with dystonia realize an inade- 12 mg of cisatracurium, and 100 μg of fentanyl.
quate response to those treatments [68]. For such The patient was maintained on sevoflurane
patients whose symptoms are sufficiently trou- (1.5 %) for the placement of the frame, and also
blesome, surgical treatment can be used to reduce while they were in the CT area. Prior to moving
symptoms and improve function. For dystonia, the patient to the CT area, Dex was started at
stimulation is primarily directed at the GPi, 0.7 μg/kg/h with remifentanil at 0.1 μg/kg/min.
which has been the most thoroughly studied Five minutes after this the sevoflurane was
stimulation site to date. stopped. This infusion continued up to the cre-
ation of the first burr hole in the operating room.
At this point the Dex was reduced to 0.1 μg/kg/h
Procedure and Decisions and the remifentanil was stopped. The purpose of
reducing the Dex at this time was to allow about
This particular case was one of the most complex 10–15 min to slightly awaken the patient for the
of all that we have experienced for DBS elec- recordings. Once the burr hole was created and
trode implantation. The patient was extremely the dura incised, the BP was confirmed to be
tenuous metabolically due to the MMA and 127/77 and the initial cannuli were inserted into
required anesthesia for the procedure due to the brain, followed by the microelectrode. At this
excessive movement. The anesthetics themselves point (with the patient still intubated yet alert and
can cause not only poor recordings but also able to follow simple commands) the patient
adverse metabolic effects. In this case, the sur- exhibited no spasms and even had his eyes open.
geon, neurophysiologist, anesthesiologist, criti- Once the electrode entered the brain, “burster”
cal care physicians, neurologists, and social cells were recorded indicating: (a) that the tip
5 Deep Brain Stimulation 81

was in the GPe, and (b) that the sedation level remifentanil was adjusted to 0.07 μg/kg/min for
(Dex at 0.1 μg/kg/h) permitted the ability to the closure of the first burr hole and the creation
record single units while still preventing exces- of the second burr hole. Once the second burr
sive movements. Recordings continued, moving hole was created, the Dex was reduced to 0.7 μg/
through GPe, through the laminae separating the kg/h and the remifentanil was stopped.
two structures and into GPi. Because the firing Nitroprusside was continued at 0.5 μg/kg/min.
rates of the GPe and the GPi are similar, the No further changes in anesthetic were needed
ability to record kinesthetic cells is critical to dis- during the recording or stimulation on this side.
tinguish between the two. Keeping all anesthetics The patient was able to open his eyes during the
constant, kinesthetic testing was done on all cells procedure. On this second side, 12 distinct GPi
encountered. Eight distinct single units were cells were recorded, and a small area of the inter-
located on the first side with kinesthetic activity nal laminae that separates the external and inter-
noted on three of them. Recording was stopped nal segments of the internal GPi was detected.
near the base of the GPi due to an increase in BP Kinesthetic activity was noted on five of those 12
to 168/120 mmHg, which was most likely due to cells. The optic tract was also noted about 1.7 mm
a continued wearing off of the Dex and the patient from the base of the GPi. Stimulation testing,
becoming more alert to his surroundings. similar to the first side, was performed with no
Advancing of the electrode stopped and two adverse events. At this point, sevoflurane was
6-mg doses of hydralazine and a 1.5-μg/kg/min added at 0.7 % for the remainder of the proce-
infusion of sodium nitroprusside was needed to dure, which included the implantation and tun-
bring the pressure down. The pressure eventually neling of the connecting wires and IPG.
stabilized to 114/50 mmHg. The Dex was Case 3: Complex: A 29-year-old male with
increased to 1 μg/kg/h and the remifentanil was mild cerebral palsy and severe status dystonicus,
increased to 0.1 μg/kg/min. Upon exiting GPi, a which developed after a surgical lysis of omental
border cell was noted and then the optic tract adhesions and a partial omentectomy.
2 mm below. Because it was impossible for the There were no complications during the
patient to describe muscle activity to the team omentectomy yet upon awakening from surgery,
during macrostimulation (stimulation testing the patient presented with intractable spasms
through the permanent DBS lead after it is placed and jerky movements, which were thought to be
to assure there are no significant adverse events), related to the opiate (fentanyl) or the scopol-
we used EMG recordings to assess muscle acti- amine the patient received during the procedure.
vation. No direct driving activity was noted at The patient had a similar reaction after an ear-
5 Hz, while a minor thumb contraction was noted lier surgery. At this time the patient was placed
at 130 Hz and 7.0 V. 5 Hz is used because it is the on Ativan (2 mg IV) and Benadryl (50 IV every
lowest output of the test stimulator and will give 6 h) with no change over the next couple of
muscle contractions that follow the stimulus days. At this time a diagnosis of myoclonic dys-
train. If this occurs, the muscle activity is related tonia was made. Due to the excessive and pain-
to direct activation of the internal capsule and ful dystonia and dystonic posturing, the patient
demonstrates that the electrode is too medial and was intubated and placed on a propofol drip
posterior. If no muscle activity is noted, then the (95 mg/h [25.13 μg/kg/min]; the patient weighed
contracture noted previously is most likely due to 63.1 kg), valproic acid (250 mg t.i.d.) and put
stimulation of the GPi. In our experience, this has back on Ativan and Benadryl on postoperative
been shown to be a positive indicator of electrode day 5. Attempts were made to reduce both the
placement, as is a contraction of the nasal labial propofol (40 mg/h (10.57 μg/kg/min)) and
fold, when stimulation is above 5.0 V. With this Ativan to levels to that which the patient could
information the electrode was then placed. The communicate; yet the painful dystonic cramping
Dex was then increased to 1.4 μg/kg/h and the would return. At 4 weeks postsurgery, Dex was
82 J.L. Shils et al.

started and the propofol was decreased with the Procedure and Decisions
hope of reducing the spastic movements. At this
point, the patient was extubated but continued Given the lack of a “best medical therapy” benefit,
to have the spastic movements, which eventu- bilateral GPi DBS surgery was performed 2 months
ally led to reintubation, being placed back on a after the onset of the intractable spasms and jerky
propofol drip (95 mg/h), and started on tetra- movements, which led to an improvement in symp-
benazine (50 mg IV q.6 h). The patient was toms. During the procedure, three MER tracts were
transferred to our institution for GPi DBS to performed (two on the right side and one on the left
treat the status dystonicus [69]. side). An average of 3.3 GPi cells were recorded per
stereotactic pass. This is somewhat less than the
normal 10+ cells that we find in the GPi in PD and
Status Dystonicus a DBS dystonia patient. The average GPi firing rate
of this patient was also significantly less: 34.3 +/−
Status dystonicus (SD) was first recognized by 16.5 Hz as compared to PD patients, which are in
Jankovic and Penn in 1982 and had been defined the range of 60–80 Hz. While for GPe, in this
as “increasingly frequent and severe episodes of patient, it was 44.5 +/− 16.6 Hz, which is lower than
generalized dystonia and rigidity, which may be found in PD, yet not as significant as was found in
refractory to standard drug therapy” [70]. This the GPi firing rate. Figure 5.4a shows the spike
condition had been labeled as “status dystonicus” activity in both the GPe and the GPi of this patient
or “dystonic storm.” The condition is quite rare, at different propofol concentration levels. This
with less than 40 episodes reported in the litera- reduced activity is similar to what is observed dur-
ture [71]. Patients with SD usually develop life-­ ing surgery in patients for dystonia when on propo-
threatening complications such as bulbar fol as compared with no propofol (Fig. 5.4b). The
weakness, compromising upper airway patency change in firing rate appears to be related to the
with the risk of pulmonary aspiration, progressive level of propofol, as seen in Fig. 5.4a. It is interest-
impairment of respiratory function leading to the ing to note that given the continuous propofol infu-
development of respiratory compromise, exhaus- sions over 2 months prior to surgery, the
tion and pain and metabolic derangements. intraoperative dose needed to be much higher for
Several drugs and surgical procedures have sedation when compared to other patients undergo-
been tried for SD with no consistent outcome. ing a variety of surgical procedures, and even at
Orally active medications have been tried [72], these high propofol concentrations the patient was
once again with no consistent outcome, and in still interacting in a meaningful way. Even with the
many cases the current literature favors using ability of the patient to interact with the surgical
intravenous agents for deep sedation [73]. Patients team, there was still a reduction in basal ganglia
with SD often develop metabolic complications activity (as measured by the GPi firing rate), which
such as rhabdomyolysis, which can lead to renal clinically manifests as a reduction in abnormal
failure, and bulbar and respiratory complications, movements. In a review article, Wilson et al. [74]
which require tracheal intubation. Other compli- found multiple instances of patients demonstrating
cations often seen include hyperpyrexia, muscle a propofol tolerance with continuous infusion, thus
exhaustion, pain, and dehydration. For all the requiring greater and greater doses for sedation and
above reasons, patients need to be treated in the reduction of pain over time. Propofol directly acti-
intensive care unit (ICU) setting. Patients com- vates GABAA receptors, which have the effect of
monly require deep sedation under muscle paraly- reducing the overall activity in the BG due to the
sis and assisted ventilation [71]. Intravenous receptors’ inhibitory nature. Additionally, propofol
infusion of midazolam and propofol may be used has been shown to affect other neuotranmitters and
in the management of SD. Second-­line strategies, neuromodulators, specifically cannabinoid recep-
especially in those with progressive disorders, tors, which are thought to be sedative sites of action
involve deep brain stimulation surgery. of anesthetics [75]. The variation in accommoda-
5 Deep Brain Stimulation 83

Fig. 5.4 Single-unit recordings from both the GPe and ings from the GPi in a patient with dystonia during a pro-
GPi during different infusion concentrations of propofol pofol infusion and with no infusion (b)
in a patient with status dystonicus (a). Single-unit record-

tion between these two receptors may account for lenge. One advantage in most movement disorder
the clinical differences (communication versus BG procedures is that very little sedative or global
activity reduction) noted in this patient. pain medication is needed, but because of the
highly sensitive nature of the tissue, BP control is
critical as is O2 saturation. Finally, due to the
Conclusion diversity of movement disorders, there may be
times where sedative anesthesia or even GA is
As in most other areas of intraoperative neuro- needed, and the procedures can only be performed
physiology, the trade-offs between anesthetics and by a close collaboration between the s­ urgeon, the
the collection of neurophysiology data are a chal- anesthesiologist, and the neurophysiologist.
84 J.L. Shils et al.

References1 niotomy without laryngeal mask. Arq Neuropsiquiatr.

2005;63:748–50.
15. Mack PF, Perrine K, Kobylarz E, Schwartz TH, Lien
1. Umemura A, Jaggi JL, Hurtig HI, Siderowf AD,
CA. Dexmedetomidine and neurocognitive test-
Colcher A, Stern MB, Baltuch GH. Deep brain stimu-
ing in awake craniotomy. J Neurosurg Anesthesiol.
lation for movement disorders: morbidity and mortal-
2004;16:20–5.
ity in 109 patients. J Neurosurg. 2003;98:779–84.
16. Souter MJ, Rozet I, Ojemann JG, Souter KJ, Holmes
2. Apetauerova D, Schirmer CM, Shils JL, Zani J,
MD, Lee L, Lam AM. Dexmedetomidine sedation
Arle JE. Successful bilateral deep brain stimulation
during awake craniotomy for seizure resection: effects
of the globus pallidus internus for persistent status
on electrocorticography. J Neurosurg Anesthesiol.
dystonicus and generalized chorea. J Neurosurg.
2007;19:38–44.
2010;113:634–8.
17. Rozet I, Muangman S, Vavilala MS, Lee LA, Souter
3. Venkatraghavan L, Luciano M, Manninen P. Anesthetic
MJ, Domino KJ, et al. Clinical experience with
management of patients undergoing deep brain stimu-
dexmedetomidine for implantation of deep brain
lator insertion. Anesth Analg. 2010;110:1138–45.
stimulators in Parkinson’s disease. Anesth Analg.
4. Coenen VA, Gielen FLH, Castro-Prado F, Abdel
2006;103:1224–8.
Rahman A, Honey CR. Noradrenergic modulation of
18. Uyar AS, Yagmurdur H, Fidan Y, Topkaya C, Basar
subthalamic nucleus activity in human: metoprolol
H. Dexmedetomidine attenuates the hemodynamic
reduces spiking activity in microelectrode recordings
and neuroendocrinal responses to skull-pin head-­
during deep brain stimulation surgery for Parkinson’s
holder application during craniotomy. J Neurosurg
disease. Acta Neurochir (Wien). 2008;150:757–62.
Anesthesiol. 2008;20:174–9.
5. Ruskin DN, Bergstrom DA, Kaneoke Y, Patel BN,
19. Rozet I. Anesthesia for functional neurosurgery: the
Twery MJ, Walters JR. Multisecond oscillations in
role of dexmedetomidine. Curr Opin Anaesthesiol.
firing rate in the basal ganglia: robust modulation
2008;21:537–43.
by dopamine receptor activation and anesthesia. *
20. Khatib R, Ebrahim Z, Rezai A. Anesthetic complica-
J Neurophysiol. 1999;81:2046–55.
tions during deep brain stimulation. Anesthesiology.
6. Hutchison WD, Lozano AM. Microelectrode record-
2004;101:A379.
ings in movement disorder surgery. In: Lozano AM,
21. Vitek JL, Bakay RA, Hashimoto T, Kaneoke Y,
editor. Movement disorders surgery. Basel: Karger;
Mewes K, Zhang JY, et al. Microelectrode-guided
2000. p. 103–17.
pallidotomy: technical approach and its applica-
7. Bohmdorfer W, Schwarzinger P, Binder S, Sporn
tion in medically intractable Parkinson’s disease.
P. Temporary suppression of tremor by remifentanil
J Neurosurg. 1998;88:1027–43.
in a patient with Parkinson’s disease during cataract
22. Bertrand C, Poirier L, Martinez N, Gauthier
extraction under local anesthesia [article in German].
C. Pneumotaxic localization, recording, stimulation,
Anaesthesist. 2003;52:795–7.
and section of basal brain structures in dyskinesia.
8. Burton DA, Nicholson G, Hall GM. Anaesthesia in
Neurology. 1958;8:783–6.
elderly patients with neurodegenerative disorders:
23. Bertrand G, Jasper H, Wong A. Microelectrode study
special considerations. Drugs Aging. 2004;21:229–42.
of the human thalamus: functional organization in the
9. Krauss JK, Akeyson EW, Giam P, Jankovic
ventro-basal complex. Confin Neurol. 1967;29:81–6.
J. Propofol-induced dyskinesias in Parkinson’s dis-
24. Albe-Fessard D, Arfel G, Guiot G, Derome P, Hertzog
ease. Anesth Analg. 1996;83:420–2.
E, Vourc’h G, et al. Electrophysiological studies of
10. Sanghera MK, Grossman RG, Kalhorn CG, Hamilton
some deep cerebral structures in man. J Neurol Sci.
WJ, Ondo WG, Jankovic J, et al. Basal ganglia neu-
1966;3:37–51.
ronal discharge in primary and secondary dystonia
25. Albe-Fessard D. Electrophysiological methods
in patients undergoing pallidotomy. Neurosurgery.
for the identification of thalamic nuclei. Z Neurol.
2003;52:1358–73.
1973;205:15–28.
11. Ard J, Doyle W, Bekker A. Awake craniotomy with
26. Sterio D, Berić A, Dogali M, Fazzini E, Alfaro G,
dexmedetomidine in pediatric patients. J Neurosurg
Devinsky O. Neurophysiological properties of pal-
Anesthesiol. 2003;15:263–6.
lidal neurons in Parkinson’s disease. Ann Neurol.
12. Bekker AY, Kaufman B, Samir H, Doyle W. The use
1994;35:586–91.
of dexmedetomidine infusion for awake craniotomy.
27. Alterman RL, Sterio D, Beric A, Kelly
Anesth Analg. 2001;92:1251–3.
PJ. Microelectrode recording during posteroventral
13. Berridge CW, Waterhouse BD. The locus coeruleus-­
pallidotomy: impact on target selection and complica-
noradrenergic system: modulation of behavioral state
tions. Neurosurgery. 1999;44:315–21.
and state-dependent cognitive processes. Brain Res *
28. Hutchison WD, Allan RJ, Opitz H, Levy R, Dostrovsky
Brain Res Rev. 2003;42:33–84.
JO, Lang AE, Lozano AM. Neurophysiological iden-
14. Almeida AN, Tavares C, Tibano A, Sasaki S, Murata
tification of the subthalamic nucleus in surgery for
KN, Marino Jr R. Dexmedetomidine for awake cra-
Parkinson’s disease. Ann Neurol. 1998;44:622–8.
29. Forster A, Eljamel MS, Varma TR, Tulley M, Latimer
1
Asterisks indicate key references. M. Audit of neurophysiological recording dur-
5 Deep Brain Stimulation 85

ing movement disorder surgery. Stereotact Funct Clinical predictive factors of subthalamic stimulation
Neurosurg. 1999;72:154–6. in Parkinson’s disease. Brain. 2002;125:575–83.
*
30. Zonenshayn M, Rezai AR, Mogilner AY, Beric A, 47. Benabid AL, Benazzouz A, Hoffmann D, Limousin
Sterio D, Kelly PJ. Comparison of anatomic and neu- P, Krack P, Pollak P. Long-term electrical inhibition
rophysiological methods for subthalamic nucleus tar- of deep brain targets in movement disorders. Mov
geting. Neurosurgery. 2000;47:282–92. Disord. 1998;13 Suppl 3:119–25.
31. Hardy J. Electrophysiological localization and identi- 48. Vesper J, Chabardes S, Fraix V, Sunde N, Østergaard
fication. J Neurosurg. 1966;24:410–4. K, Kinetra Study Group. Dual channel deep brain
32. Lozano A, Hutchison W, Kiss Z, Tasker R, Davis stimulation system (Kinetra™) for Parkinson’s
K, Dostrovsky J. Methods for microelectrode- disease and essential tremor: a prospective multi-
guided posteroventral pallidotomy. J Neurosurg. center open label clinical study. J Neurol Neurosurg
1996;84:194–202. Psychiatry. 2002;73:275–80.
33. Bakay RA, DeLong MR, Vitek JL. Posteroventral 49. Anderson VC, Burchiel KJ, Hogarth P, Favre J,
pallidotomy for Parkinson’s disease. J Neurosurg. Hammerstad JP. Pallidal vs subthalamic nucleus deep
1992;77:487–8. brain stimulation in Parkinson disease. Arch Neurol.
34. Baron MS, Vitek JL, Bakay RA, Green J, Kaneoke Y, 2005;62:554–60.
Hashimoto T, et al. Treatment of advanced Parkinson’s 50. Volkmann J, Allert N, Voges J, Weiss PH, Freund HJ,
disease by posterior GPi pallidotomy: 1-year results Sturm V. Safety and efficacy of pallidal or subthalamic
of a pilot study. Ann Neurol. 1996;40:355–66. nucleus stimulation in advanced PD. Neurology.
35. Sutton JP, Couldwell W, Lew MF, Mallory L, Grafton 2001;56:548–51.
S, DeGiorgio C, et al. Ventroposterior medial pallidot- 51. Blomstedt P, Hariz MI. Are complications less com-
omy in patients with advanced Parkinson’s disease. mon in deep brain stimulation than in ablative pro-
Neurosurgery. 1995;36(6):1112–6. cedures for movement disorders? Stereotact Funct
36. Lang AE, Lozano AM, Montgomery E, Duff J, Tasker Neurosurg. 2006;84(2–3):72–81.
R, Hutchinson W. Posteroventral medial pallidotomy 52. Alkhani A, Lozano AM. Pallidotomy for Parkinson
in advanced Parkinson’s disease. N Engl J Med. disease: a review of contemporary literature.
1997;337:1036–42. J Neurosurg. 2001;94:43–9.
37. Iacono RP, Shima F, Lonser RR, Kuniyoshi S, Maeda 53. Beric A, Kelly PJ, Rezai A, Sterio D, Mogilner A,
G, Yamada S. The results, indications, and physiol- Zonenshayn M, Kopell B. Complications of deep
ogy of posteroventral pallidotomy for patients with brain stimulation surgery. Stereotact Funct Neurosurg.
Parkinson’s disease. Neurosurgery. 1995;36:1118–25. 2001;77(1–4):73–8.
38. Laitinen LV, Bergenheim AT, Hariz 54. Mason LJ, Cojocaru TT, Cole DJ. Surgical inter-
MI. Ventroposterolateral pallidotomy can abolish all vention and anesthetic management of the patient
parkinsonian symptoms. Stereotact Funct Neurosurg. with Parkinson’s disease. Int Anesthesiol Clin.
1992;58:14–21. 1996;34:133–50.
39. Laitinen LV. Ventroposterolateral pallidotomy. 55. Nicholson G, Pereira AC, Hall GM. Parkinson’s dis-
Stereotact Funct Neurosurg. 1994;62:41–52. ease and anaesthesia. Br J Anaesth. 2002;89:904–16.
40. Vitek JL. Deep brain stimulation for Parkinson’s 56. Gross M, Bannister R, Godwin-Austen R. Orthostatic
disease. A critical re-evaluation of STN versus GPi hypotension in Parkinson’s disease. Lancet.
DBS. Stereotact Funct Neurosurg. 2002;78:119–31. 1972;1(7743):174–6.
41. Moreau C, Defebvre L, Destee A, Bleuse S, Clement 57. Vincken WG, Gauthier SG, Dollfuss RE, Hanson
F, Blatt JL, et al. STN-DBS frequency effects on freez- RE, Darauay CM, Cosio MG. Involvement of upper-­
ing of gait in advanced Parkinson disease. Neurology. airway muscles in extrapyramidal disorders. A cause
2008;71:80–4. of airflow limitation. N Engl J Med. 1984;311:438–42.
42. Brozova H, Barnaure I, Alterman RL, Tagliati 58. Lee KH. Oromandibular dystonia. Oral Surg Oral
M. STN-DBS frequency effects on freezing of Med Oral Pathol Oral Radiol Endod. 2007;104:491–6.
gait in advanced Parkinson disease. Neurology. 59. Farmer SF, Sheean GL, Mayston MJ, Rothwell JC,
2009;72(8):770. Marsden CD, Conway BA, et al. Abnormal motor unit
43. Brotchie J, Fitzer-Attas C. Mechanisms compensat- synchronization of antagonist muscles underlies path-
ing for dopamine loss in early Parkinson disease. ological co-contraction in upper limb dystonia. Brain.
Neurology. 2009;72(7 Suppl):S32–8. 1998;121(Pt 5):801–14.
44. Chesselet MF. Dopamine and Parkinson’s dis- 60. Gracies J-M, Simpson DM. Spastic dystonia. In: Brin
ease: is the killer in the house? Mol Psychiatry. MF, Comella C, Jankovic J, editors. Dystonia etiol-
2003;8:369–70. ogy, clinical features, and treatment. Philadelphia:
45. Ekesbo A, Rydin E, Torstenson R, Sydow O, Lippincott Williams & Wilkins; 2004. p. 195–212.
Låengström B, Tedroff J. Dopamine autorecep- 61. Zhang JG, Zhang K, Wang ZC, Ge M, Ma Y. Deep
tor function is lost in advanced Parkinson’s disease. brain stimulation in the treatment of secondary dysto-
Neurology. 1999;52:120–5. nia. Chin Med J (Engl). 2006;119:2069–74.
46. Welter ML, Houeto JL, Tezenas du Montcel S, 62. Katsakiori PF, Kefalopoulou Z, Markaki E, Paschali
Mesnage V, Bonnet AM, Pillon B, Arnulf I, et al. A, Ellul J, Kagadis GC, et al. Deep brain stimulation
86 J.L. Shils et al.

for secondary dystonia: results in 8 patients. Acta ders. In: Deletis V, Shils JL, editors. Neurophysiology
Neurochir (Wien). 2009;151:473–8. in neurosurgery: a modern intraoperative approach.
63. Sani S, Ostrem JL, Shimamoto S, Levesque N, Starr New York: Elsevier; 2002. p. 405–48.
PA. Single unit “pauser” characteristics of the globus 69. Apetauerova D, Schirmer CM, Shils JL, Zani J, Arle
pallidus pars externa distinguish primary dystonia JE. Successful bilateral deep brain stimulation of the
from secondary dystonia and Parkinson’s disease. globus pallidus for persistent status dystonicus and
Exp Neurol. 2009;216:295–9. generalized chorea. J Neurosurg. 2010;113:634–8.
64. Woehrle JC, Blahak C, Kekelia K, Capelle HH, 70. Manji H, Howard RS, Miller DH, Hirsch NP, Carr L,
Baezner H, Grips E, et al. Chronic deep brain stim- Bhatia K, et al. Status dystonicus: the syndrome and
ulation for segmental dystonia. Stereotact Funct its management. Brain. 1998;121:243–52.
Neurosurg. 2009;87:379–84. 71. Mariotti P, Fasano A, Contarino MF, Della Marca G,
65. Schneider SA, Klein C. PINK1 type of young- Piastra M, Genovese O, et al. Management of status
onset Parkinson disease. In: Pagon RA, Bird TC, dystonicus: our experience and review of the litera-
Dolan CR, Stephens K, editors. GeneReviews ture. Mov Disord. 2007;22:963–8.
[Internet]. Seattle: University of Washington; 2010. 72. Teive HA, Munhoz RP, Souza MM, Anoniuk SA,
p. 1993–2010. Santos ML, Teixeira MJ, et al. Status dystonicus: study
66. Raymond D, Bressman SB. Early-onset primary dys- of five cases. Arq Neuropsiquiatr. 2005;63:26–9.
tonia (DYT1). In: Pagon RA, Bird TC, Dolan CR, 73. Vaamonde J, Narbona J, Weiser R, Garcia MA,
Stephens K, editors. GeneReviews [Internet]. Seattle: Branna T, Obeso JA. Dystonic storms: a practi-
University of Washington; 2010. p. 1993–9. cal management problem. Clin Neuropharmacol.
67. Albanese A, Barnes MP, Bhatia KP, Fernandez-­ 1994;17:334–47.
Alvarez E, Filippini G, Gasser T, et al. A systematic 74. Wilson C, Cannin P, Caravati M. Clin Toxicol (Phila).
review on the diagnosis and treatment of primary 2010;48:165–70.
(idiopathic) dystonia and dystonia plus syndromes: 75. Patel S, Wohlfeil ER, Radmacher DJ, Carrier EJ,
report of an EFNS/MDS-ES Task Force. Eur J Neurol. Perry LJ, Kundu A, et al. The general anesthetic
2006;13:433–44. Propofol increases brain N-arachidonylethanolamine
68. Shils JL, Tagliati M, Alterman RL. Neurophysiological (anandamide) content and inhibits fatty acid amide
monitoring during neurosurgery for movement disor- hydrolase. Br J Pharmacol. 2003;139:1005–13.
 Monitoring of Spinal Cord
Functions 6
Sumihisa Aida, Tatsuro Kohno, and Koki Shimoji

Key Learning Points

Introduction
• The methods to record the spinal cord
potentials Spinal and cardiovascular surgeries impart
• Segmental nerve stimulation mechanical [1] or ischemic [2] stress to the spinal
• Transcranial electric or magnetic stimulation cord. To prevent spinal cord injury due to such
• Spinal cord stimulation surgical stresses, the intraoperative monitoring of
• Heterosegmental nerve stimulation spinal cord function is important. Until the 1960s,
• The techniques to introduce the catheter elec- there were no available methods of spinal cord
trodes into the epidural space function monitoring. In the late 1960s, a novel
• To identify the components of segmental spi- tool for monitoring spinal cord function became
nal cord potentials—initial spikes, negative available; it recorded somatosensory-evoked
waves, slow positive waves potentials (SSEPs) from the scalp. In this tech-
• Conducting (conductive) spinal cord potentials nique, the electrical potentials evoked by periph-
• Heterosegmental spinal cord potentials—slow eral nerve stimulation are recorded on an
positive waves electroencephalogram (EEG). The short-latency
component of SEPs may indicate cervical spinal
cord function in waveforms (latencies and ampli-
S. Aida, M.D., Ph.D. (*) tudes) [3].
Geriatric Health Service Facility,
Since their introduction, SEPs have been used
Izumi, 5-35-2 Nishiarai, Adachi-ku, Tokyo, Japan
e-mail: [email protected] in clinical monitoring, diagnosis and investiga-
tions, as well as in animal experiments [4]. In the
T. Kohno, M.D., Ph.D.
Division of Anesthesiology, Niigata University 1980s, SSEPs were also adopted for the monitor-
Graduate School of Medical and Dental Sciences, ing of spinal cord function during spinal [5] or
1-757 Asahi-machi, Niigata, Japan cardiovascular [6] surgery. The recording of
e-mail: [email protected]
small electrical changes, such as SSEPs, requires
K. Shimoji, M.D., Ph.D. the close attachment of electrodes to spinal cord
Niigata University Graduate School of Medicine,
tissues. Commonly, electrodes are placed on the
Niigata, Japan
scalp close to the sensory cortex. These poten-
Standard Medical Information Center, NPO, Pain
tials recorded from the scalp or even the cervical
Control Institute, Inc., 45-304, Yarai-cho,
Shinjuku-ku, Tokyo, Japan skin surface may not reflect the spinal cord func-
e-mail: [email protected] tion because they interact with brainstem potentials

© Springer International Publishing AG 2017 87

A. Koht et al. (eds.), Monitoring the Nervous System for Anesthesiologists
and Other Health Care Professionals, DOI 10.1007/978-3-319-46542-5_6
88 S. Aida et al.

or far-field potentials. Thus, segmental or regional both electrical activities (SSEPs and SCPs) when
electric changes in the spinal cord cannot be monitoring spinal cord function, particularly
recorded from the body surface, because the spi- when surgical manipulations involve the sensory
nal cord is situated deep in the body, and its activ- spinal tracts (see Appendix).
ities are obscured by other electrical activities Thoracic and lumbar aortic occlusion has fre-
such as electroencephalogram (EEG), electromyo- quently been demonstrated to result in ischemia
gram (EMG), and electrocardiogram (ECG) [7]. of the ventral two-thirds of the spinal cord [16].
During the same time, another novel technique As such, there was also a great demand from sur-
of recording SCPs directly from the spinal cord by geons and anesthesiologists for the monitoring of
inserting electrodes into the dorsal epidural space spinal motor function. In the late 1990s, transcra-
was developed by Shimoji et al. [8]. A catheter nial motor-evoked potentials (tc-MEPs) were uti-
electrode was placed percutaneously in the same lized to monitor spinal motor function. Using this
manner as for a continuous epidural block. From monitoring technique, impulses generated by
the epidural electrodes, spinal cord field potentials transcranial stimulation of giant pyramidal cells
were recorded [9]. In addition, the spinal cord was descend the pyramidal tract to the spinal moto-
stimulated electrically using the same epidural neurons, producing skeletal muscle contractions.
electrodes utilized for pain management (spinal In this way, the stimulation technique and the
cord stimulation [SCS]) [10, 11]. recording of tc-MEPs are simple. Both electric
Intrathecal electrodes were utilized by and magnetic stimulation are currently used in
Magladery et al. [12] initially, but the risks asso- clinical practice (electric tc-MEPs [17] and mag-
ciated with electrode insertion raised great con- netic tc-MEPs [18, 19]). The motor responses
cern. As a result, intrathecal electrodes were not resulting from transcranial stimulation are
used clinically except in those cases in which sur- recorded as EMG responses.
gical manipulations were carried out directly on Following the development of tc-MEPs,
the cord [13]. With the idea that epidural record- transcranial stimulation was used to evoke SCPs
ing can minimize the risks compared with those (transcranially evoked SCP [tc-SCP]) in the
of intrathecal recording, Shimoji et al. [9, 14] early 2000s. Either transcranial electric (electric
further developed techniques to exclude contami- tc-­SCP) [17] or magnetic (magnetic tc-SCP)
nation by ECG activity when recording evoked [18, 19] stimulation is applied to giant pyrami-
SCPs (discussed later). Following these develop- dal cells and the responses can be recorded
ments, recording and stimulation with epidural directly from the spinal cord epidurally [15].
electrodes have been used in routine clinical The tc-SCPs provide highly precise data,
monitoring, therapeutics, and investigations, as because spinal motor function can be directly
well as in animal experiments at various insti- monitored.
tutes [15].
The acquisition of SSEPs requires a some-
what lengthy period of time from a few to several  ecording Evoked SCPS
R
minutes, because the repetitive recording (50– for Intraoperative Monitoring
200 times) of waveforms is needed for amplifica-
tion by computed averaging [3–6]. Both SSEP A catheter equipped with several Ag-AgCl elec-
and SCP recordings are easily interfered with by trodes, designed by Shimoji et al. [8], is inserted
the noise from electrocoagulation or other elec- into the epidural space, using the same technique
trical sources of noise during surgery. SSEPs are as that for the continuous epidural block, at a level
cerebral electrical changes evoked by sensory corresponding to the spinal segment to be moni-
nerve stimulation, and evoked SCPs are spinal tored. The catheter with platinum–iridium elec-
electrical changes evoked by sensory nerve or trodes (Medtronic™ Inc., Minneapolis, MN),
spinal cord stimulation (SCP). Therefore, it is which was developed for SCS, is also available for
advantageous to acquire simultaneous records of SCP recording. Because evoked SCPs are easily
6 Monitoring of Spinal Cord Functions 89

Fig. 6.1 Recording methods of spinal cord potentials manner, evoked SCPs are recorded in the electrically
(SCPs). A catheter electrode is inserted into the dorsal epi- silent phase (0.5–0.7 s) between T and P waves [9]. (a) A
dural space using the same technique as that for the con- catheter (1 m diameter) equipped with several Ag-AgCl
tinuous epidural block at a level corresponding to the electrodes (A-1), and platinum–iridium electrodes (A-2),
spinal segment to be monitored [8]. To exclude ECG con- which are designed for spinal cord stimulation, are also
tamination, stimulation pulses are triggered by a QRS available. (b) An illustration of a catheter electrode
component of ECG with a delay of 0.3–0.5 s, when the inserted into the epidural space. (c) The relation between
triggered pulses coincide with the end of a T wave. In this ECG and electrical stimulation [14, 15]

affected by ECG artifacts, the exclusion of ECG The waveforms of the segmental SCPs
contamination is very important. To exclude such recorded in the cervical and lumbosacral enlarge-
contamination, a peripheral nerve is stimulated by ments are very similar. Evoked SCPs can be
square pulses triggered by a QRS component of recorded at the level of the cervical enlargement
ECG with a delay of 0.3–0.5 s, when the triggered by the stimulation of a nerve trunk in the lower
pulses coincide with the end of a T wave. Thus, limb (ascending evoked SCPs; see Fig. 6.2b).
evoked SCPs are recorded on the electrically silent The waveforms of SCPs recorded at the lumbar
phase (0.5–0.7 s) between T and P waves (Fig. 6.1) enlargement by epidural stimulation of the cervi-
[8, 20–23]. cal spinal cord are similar to those of the segmen-
The peripheral nerve trunk is electrically stim- tally evoked SCPs (descending evoked SCPs; see
ulated to evoke SCP responses. For monitoring of Fig. 6.2a, c) [8, 9, 10, 14, 15, 22, 23].
cervical spinal cord activity, epidural catheter The summated electrical potentials travel
electrodes are inserted into the cervical epidural along the spinal cord in response to SCS in the
space close to the cervical enlargement. Next, the epidural space of the upper or lower spinal seg-
brachial plexus or the radial, ulnar, or median ments, cauda equina, or from peripheral nerve
nerve is stimulated (segmentally evoked SCPs, stimulation. Thus, the human SCPs can be
Fig. 6.2a). To monitor lumbar spinal cord func- recorded from the cervical epidural space in
tion, the recording electrodes are placed close to response to cauda equina stimulation at the L3–4
the lumbar enlargement, and the common tibial or vertebral level or SCS by an electrode situated in
peroneal nerve is stimulated (segmentally evoked the epidural space close to the lumbar enlarge-
SCPs, Fig. 6.2a, c) [7, 9, 12]. ment (ascending SCPs; see Fig. 6.2b, c) and vice
90 S. Aida et al.

Fig. 6.2 Recording of spinal cord potentials (SCPs). A tal evoked SCPs recorded from the cervical enlargement
peripheral nerve trunk is electrically stimulated (see (c) by stimulation of the ulnar nerve (a). (b) Ascending
Fig. 6.1). Recording electrodes are inserted into the epi- evoked SCPs recorded from the cervical enlargement (c)
dural space to monitor the dorsal spinal activity. by stimulation to the common tibial nerve (b). (c)
Dysfunction of the spinal cord is reflected as abnormali- Segmental evoked SCPs recorded from the lumbar
ties in the waveforms, such as prolongation of peak laten- enlargement (d) by stimulation at the common tibial nerve
cies and depression or augmentation of amplitudes. The (b). P1: action potential of spinal nerve roots. N1: synchro-
waveforms of the segmental SCPs recorded in the cervical nized activities of interneurons. P2: primary afferent depo-
and lumbosacral enlargements are very similar to each larization (PAD). Sometimes, P2 splits into two
other. In the ascending evoked SCPs, complex positive components, first (P2f) and second (P2s) components.
waves (C1, C2, and C3) are recorded, but N1 and P2 N-dip: negative dip driving P2f and P2s [8, 22, 23]
waves are hardly noticed. (a) The waveforms of segmen-

versa (SCS from cervical epidural space and such as corkscrew-­shaped electrodes are used for
recording at the lumbar enlargement level) stimulation so as to avoid scalp burns. For mag-
(descending SCPs) [8, 10, 15]. netic tc-MEPs, a magnetic coil is used for stimula-
tion. For both forms of stimulation, EMG (CMAP)
responses are usually recorded from the abductor
 ecording Electric tc-MEPS
R pollicis brevis muscle (upper limb) or the tibialis
and Magnetic tc-MEPS anterior muscle (lower limb) using needle or sur-
for Intraoperative Monitoring face electrodes [24–26].
In response to pulse train stimulations, giant
In transcranial electrical stimulation, electrodes pyramidal cells are directly depolarized generat-
are placed at C3 or C4 on the scalp, and a train of ing D waves [27–29]. However, during magnetic
square wave stimulation pulses (with a pulse dura- stimulation, interneurons are depolarized first
tion of 0.02–0.2 ms and an interpulse interval of and the firing of giant pyramidal cells follows
0.2 ms) are applied at an intensity of 250–1000 V sequentially generating several I waves at 1.5- to
[24–26]. Because of the use of high-voltage stimu- 2.0-ms intervals [28, 29]. Both D and I waves
lation to obtain electric tc-­MEPs, durable electrodes descend the pyramidal tract as a group of several
6 Monitoring of Spinal Cord Functions 91

Fig. 6.3 Recording of transcranially stimulated motor- is stimulated magnetically at the parietal cranium (m)
evoked potentials (tc-MEPs). The brain is stimulated and recorded from the abductor pollicis brevis muscle
electrically or magnetically, and the evoked EMG (m-emg). (b) Electric tc-MEP, which is stimulated electri-
(motor-evoked potential [MEP]) is recorded from the cally at the contralateral scalp, C3 or C4 (e), and recorded
abductor pollicis brevis muscle (upper limb) or the tibialis from the abductor pollicis brevis muscle (e-emg)
anterior muscle (lower limb). (a) Magnetic tc-MEP, which

spiky waves (multiple descending volleys), and electric and magnetic stimulation are described
the spinal motor neurons are stimulated by these for the respective tc-MEPs. However, the meth-
piston-like multiple descending volleys, resulting ods of recording the tc-SCP responses are the
in summation of excitatory postsynaptic poten- same as those described for the segmental SCPs
tials (EPSP) [18, 29–31]. Thus, the spinal motor [8–10, 17, 18].
neurons are excited with a few milliseconds of When acquiring magnetic tc-SCP responses,
delay (Fig. 6.3) [18, 19, 28–33]. the generation of multiple descending volleys in
Convulsions are occasionally induced during the pyramidal tract as well as the process of elec-
brain stimulation for a clinical examination in the tric summation in the dorsal horn are clearly
arousal state [32], but they are rare during stimu- depicted [18, 19, 27, 29]. These phenomena are
lation for intraoperative monitoring under gen- never evident when acquiring magnetic tc-MEP
eral anesthesia [33, 34]. Surgical procedures on responses because only evoked EMG activity can
the spine or spinal cord should be performed be observed [14, 18, 19]. Therefore, magnetic tc-­
under general anesthesia with adequate monitor- SCPs may provide fine data quality when moni-
ing of spinal cord function [35]. toring spinal motor function (Fig. 6.4).

Recording tc-SCPS  nesthetics Used in Spinal Cord

A
for Intraoperative Monitoring Monitoring

Transcranial stimulation also results in electrical Many general anesthetics, especially inhalation
changes in the SCP recordings, which appear to anesthetics, suppress spinal electrical activity,
provide better evidence of spinal motor function thereby reducing the amplitudes and prolonging the
than tc-MEPs [18, 19, 27–29]. The methods of latencies of SSEPs, SCPs [36–39], and tc-­MEPs.
92 S. Aida et al.

Fig. 6.4 Recording of transcranially stimulated evoked tract. The volleys result in the summation of excitatory
spinal cord potentials (tc-SCPs). The brain is stimulated postsynaptic potentials (EPSPs). (a) Magnetic tc-SCP, which
magnetically or electrically, and the resulting spinal cord is stimulated magnetically at the parietal cranium (m) and
potentials (SCPs) are recorded with epidural electrodes. Both recorded from the spinal cord (scp). (b) Electric tc-SCP,
forms of stimulation evoke a group of several spiky waves which is stimulated electrically at the contralateral scalp, C3
(multiple descending volleys) descending the pyramidal or C4 (e), and recorded from the spinal cord (scp)

Therefore, the use of intravenous anesthetics, aorta is occluded during a surgical procedure,
including low-dose propofol, ketamine, fentanyl, blood flow in the distal region of the occlusion
and remifentanil, are usually recommended for becomes very small, and blood is pooled in the
surgeries that use spinal cord monitoring. Also, proximal region, where 30 % of the circulating
anesthesia can suppress spinal electrical activity, blood volume is distributed [43]. Therefore,
leading to incorrect judgments [19, 23, 24, 40, when intravenous anesthetics are continuously
41]. In light of this, simultaneous monitoring of infused using a syringe pump, the concentration
the bispectral index (BIS) is recommended for of an anesthetic in the proximal region of the aor-
maintenance of adequate anesthesia depth (40– tic occlusion elevates two- to fourfold [44]. This
60 % in BIS) [42]. Administration of a muscle elevation results in a deeper anesthetic state than
relaxant is desirable when recording SSEPs and was assumed before the aorta occlusion, leading
SCPs because the use of relaxants diminishes the to misjudgments. The monitoring of BIS is there-
presence of muscle contractions and noise due to fore helpful in realizing correct judgments in
EMG artifact. Muscle relaxants, however, inter- such cases [45].
fere with MEP monitoring.
Intravenously administered anesthetics, which
are mixed with venous blood, stream into the Case Studies (Clinical Applications)
heart (right atrium). After circulating through the
lungs, the anesthetics return to the heart (left Spinal cord monitoring by SCP, tc-MEP, or tc-­
atrium) with minimal loss or consumption. Thus, SCP is utilized during spinal surgery to assess the
most of the anesthetics enter the aorta. When the force on spinal neurons caused by traction, the
6 Monitoring of Spinal Cord Functions 93

Amplitude reduction in the SCP due to ischemia

can occur with or without prolongation of latency.
Based on findings obtained from SCP monitor-
ing, a nonharmful traction force on the spine can
be determined. Consequently, a spinal cord injury
was able to be prevented in the following case
involving open traction and fixation of the spine
to correct idiopathic scoliosis (Fig. 6.5) [46].
The surgical manipulation during resection of
spinal cord tumors causes direct and/or indirect
mechanical stress or injury to spinal neurons,
which can be determined by using tc-MEP
monitoring. When spinal cord damage occurs,
the amplitudes of the tc-MEP responses drop and
the stimulation thresholds elevate. Based on
observations of the tc-MEPs, alterations to the
tumor resection can be recommended so as to
avoid postsurgical motor dysfunction (Fig. 6.6).
When amplitude reductions exceed 50 % during
the resection, the surgical approach or the resec-
tion size needs to be reconsidered: should the
resection be continued, should a smaller resec-
tion be performed, should the angle or direction
of the section be changed, or should the resection
be discontinued altogether? Also, based on the
final findings from the tc-MEP monitoring, the
Fig. 6.5 Clinical application of SCP for spine traction. level of postsurgical motor dysfunction can be
Open traction and fixation of the thoracic spine to correct predicted [15].
idiopathic scoliosis in a 26-year-old female. Anesthesia Aortic surgery usually requires aortic cross-­
was maintained with fentanyl and ketamine. The cauda
equina (L4) was supramaximally stimulated by electric clamping, which often results in spinal cord isch-
pulses. Ascending SCPs were recorded from the posterior emia [47, 48]. Because of the prompt response to
epidural space at the C7 spinal cord level. Immediately ischemia, SCPs are monitored during aortic sur-
after traction of the spine at 15 kg, the amplitudes of C1, gery to assess the severity of ischemic stress to the
C2, and C3 were reduced to 47.0, 47.5, and 49 %, respec-
tively. These reductions in amplitude recovered within spinal cord. Sequential changes in segmental
15 min after the traction force was reduced to 12 kg. In SCPs during surgery for aortic aneurysm are
this case, the latency did not change. Reprinted with per- shown in Fig. 6.7. Ischemia due to aortic cross-­
mission from Fujioka et al. [46] clamping caused a rapid decrease in the amplitude
and an increase in the latency of the responses.
stress or injury caused by surgical manipulation, The waveforms recovered quickly after declamp-
the stress or damage due to ischemia, and hypo- ing. However, the duration of ischemia that is tol-
function under conditions of hypothermia. erable or reversible when a reduction/abolition of
Representative cases of each clinical application the SCP occurs without any resulting postsurgical
are presented below. neurological symptoms is still not clear (see
Traction force on the spine and spinal cord Fig. 6.7).
produces immediate spinal cord ischemia as a Thus, test cross-clamping before dividing an
result of blood vessels being stretched. The force aneurysm is recommended (i.e., SCP is closely
can reduce the amplitudes of SCPs by over 50 %. observed for 15 min after aortic cross-clamping).
94 S. Aida et al.

Fig. 6.6 Changes in tc-MEPs during spinal cord tumor rior (TA) muscles was recorded. During tumor resection,
resection. A 38-year-old male patient with a spinal cord the amplitudes decreased, but they recovered immediately
tumor in the T5–6 area underwent tumor resection. after the surgical approach was altered. The tumor resec-
Anesthesia was maintained by TIVA. While monitoring tion was continued from the other direction. Thus, post-
tc-MEPs, electric stimulation was performed with a five-­ surgical motor weakness was not evident. Reprinted with
pulse train (50–100 μs of duration, 2-ms intervals, 600 V) permission from Fukaya et al. [15]
on the scalp (C3 and C4), and an EMG of the tibialis ante-

If a greater than 50 % reduction of the amplitude to a drop in body temperature with characteristic
is observed, the surgeons should make repeated responses in the waveform under such conditions
short-term releases of the clamp during the sur- being prolongation of the latency, widening of
gery so as to avoid long-term ischemia or should the duration, and augmentation of the amplitude
change to another bypass route [49]. This test can (Fig. 6.8) [49].
also be modified when monitoring via tc-MEPs or Interestingly, the change in the amplitude was
tc-SCPs. On the other hand, when no or only min- biphasic when the body temperature was lowered
imal recovery in the SCP is noted during the final even further: the amplitude gradually increased
observation, motor dysfunction corresponding to until around 30 °C, and then began to decrease
the amplitude decrease can be predicted. Because under deeper hypothermia. The precise mecha-
there are concerns that heparinization during aor- nisms of the biphasic response are still not clear.
tic surgery causes an epidural hematoma, an epi- Under profound hypothermia (below 20 °C), the
dural catheter electrode should be inserted at least amplitude decreases or disappears. Under even
1 h before the surgery and extracted 1 day later lower body temperature of approximately 10 °C,
when the effect of the heparin has dissipated. the N1 wave is split into two peaks (not illus-
Cardiovascular surgery is usually performed trated in Fig. 6.7) [49]. These changes recover to
under moderate or deep hypothermia, which the baseline level immediately upon rewarming.
makes interpretation of waveform changes more Thus, body temperature measurements are indis-
complicated. The waveform of the SCP under pensable for the interpretation of the waveform
moderate hypothermia responds very sensitively changes in the SCP monitoring.
6 Monitoring of Spinal Cord Functions 95

Fig. 6.7 Effects of aortic cross clamping on segmental (13:05) after cross-clamping (12:35) and reappeared
SCPs. A 71-year-old man underwent abdominal aortic (14:12) after the declamping (13:19), with the time lag
surgery. Segmental SCPs were recorded from the epidural between its reappearance and the declamping being
space at the T12/L1 vertebral level, and the common tibial 53 min. The durations of aortic cross-clamping and that of
nerve was stimulated supramaximally at the popliteal the disappearance of SCP were 44 and 67 min, respec-
fossa. Anesthesia was maintained mainly with fentanyl tively. There were no neurological sequelae after surgery.
and midazolam, and blood cooling was initiated by cardio Reprinted with permission from Kondo et al. [49], with
pulmonary bypass. The N1 wave disappeared 30 min modification

Conclusion Appendix: Techniques

and Physiology
In conclusion, besides basic routine monitoring
of SSEPs and tc-MEPs [47], combined monitor- Introduction
ing of other parameters such as tc-MEP or tc-­SCP
and MEPs may provide an accurate monitoring The development of the catheter electrode has
of spinal cord function during surgical manipula- made it possible to stimulate the spinal cord from
tions of the spine or nearby structures depending the epidural space for pain management [53] and
on individual surgeries [50–52]. to record epidurally human spinal cord potentials
Recording of the spinal cord potentials recorded for monitoring spinal cord function during an
from the epidural space may add more precise mon- operation [8].
itoring of spinal cord functions in certain cases.
96 S. Aida et al.

Fig. 6.8 Waveform changes of descending SCPs under and rectal body temperature was measured. Subsequent
moderate hypothermia. A 56-year-old patient underwent serial changes of N1 and N2 waves were superimposed:
thoracoabdominal aortic surgery under hypothermic car- (a) before cooling; (b) SCPs during cooling process; (c)
diopulmonary bypass. Descending SCPs were recorded SCPs during rewarming process. Prolongation of the
from the epidural space at the T12/L1 vertebral level, and latency, widening of the duration, and augmentation of the
the spinal cord was supramaximally stimulated at the amplitude are noted under moderate hypothermia.
C6/7 vertebral level with epidurally inserted electrodes. Reprinted with permission from Kondo et al. [49]
Anesthesia was maintained with fentanyl and midazolam,

The Catheter Electrodes tials, measurement of epidural pressure and epi-

dural tissue blood flow, epidural spinal cord
Human spinal cord potentials (SCPs) can be stimulation, and epidural injection of drugs [55].
recorded from the epidural space using the same Accurate insertion of the catheter electrode at
catheter used for continuous epidural block. The the required site in the epidural space is critical
epidural catheter electrode can be made simply for these applications. We have been using three
by insertion of a stainless steel wire through the methods to determine the proper placement of
epidural catheter approximately 5 mm beyond its the catheter electrodes in the posterior epidural
tip (Fig. A6.1a). This simple catheter has been space: (1) epidural electrical stimulation test, (2)
used for the recording of spinal cord potentials in recording of the spinal cord potentials evoked by
patients during surgical operations or for the stimulation of the segmental, heterosegmental
stimulation of the spinal cord in patients with nerves, or dorsal cord [14, 56, 57], and (3) image
various spinal cord diseases [54]. An epidural examination such as X-ray, MRI, or CT scan.
catheter with three orifices on the side and three When the catheter electrodes are situated in
platinum wire electrodes was developed in our the posterior epidural space on the mid-line,
laboratory for multiple applications (Fig. A6.1b, stimulation through the catheter electrodes produces
c), including monitoring of spinal cord poten- the bilateral twitches of the segmental muscles.
6 Monitoring of Spinal Cord Functions 97

Fig. A6.1 Three types of epidural catheter electrodes are eter (b) that may also have three orifices on the side to
shown. A stainless steel wire can be placed through an measure epidural pressure and epidural tissue blood flow,
epidural catheter approximately 5 mm beyond its tip (a). epidural spinal cord stimulation, and epidural injection of
Three platinum wire electrodes can be placed on the cath- drugs (c)

When it produces unilateral muscle twitches in is inserted into the epidural space using the para-
the same spinal segment, the electrodes might be median approach, with the bevel parallel to the
situated laterally in the epidural space. By this sagittal plane and targeting the predicted seg-
stimulation test, you can verify the spinal seg- ment. When the tip of the Tuohy needle is located
ment position and the laterality of the catheter in the epidural space, the direction of the bevel is
electrode in the epidural space. When the cathe- adjusted, and the catheter electrode is inserted
ter electrode is situated in the anterior epidural approximately 5 cm into the epidural space. The
space, the polarity of the segmental SCPs is tip of the catheter electrode and the skin surface
reversed as expected. Laterality of the catheter electrode are connected, respectively, to the neg-
electrodes in the epidural space can also be deter- ative and positive outlets of an electrical nerve
mined by the waveform characteristics of the stimulator [57]. Using these electrodes, record-
SCPs. When the catheter electrodes are situated ings can be made at several spinal levels
ipsilateral to the stimulated peripheral nerves and (Fig. A6.2).
close to the roots, the recorded initial positive
spikes and P2 wave are larger than those recorded
contralateral to the nerves.  rigins of Each Component
O
The procedure used to introduce the catheter of the Segmental SCPs
electrodes into the epidural space is the same as
that used to place catheters for continuous epi- The initially positive spike, P1, of the segmental
dural anesthesia [53]. The patients are placed in SCPs is believed to be a reflection of the extracel-
the lateral position and flexed to open the inter- lular events associated with the action potential
spaces of the vertebral column. After making a propagation through the roots into the spinal cord
skin wheal aseptically and injecting 0.5–1.0 % [14, 58–62]. The generation of an action potential
lidocaine (5 mL), a 16- to 18-gauge Tuohy needle at a node of Ranvier creates a positive ­capacitive
98 S. Aida et al.

a median N stim

Erb’s point stim

epidural epidural scalp

(T10 - L1) (C5 - T1) (C3' / C4')

b 6 6

1
2
9

1
2
3 8
4 5
5 7
3
7 4

Fig. A6.2 Recording of the human SCPs. The recording electrodes are placed at various levels of the spine (a) into the
epidural space (b)

current, which is conducted electronically down efflux from the axon into the extracellular space.
the axon and the nearby tissues. This current is The additional positivity in the extracellular
responsible for the initial positivity of the tripha- space is recorded from the cord dorsum as the
sic spikes (Fig. A6.3a). The capacitative current second positive component of the triphasic spikes
is also responsible for depolarizing the cell mem- (Fig. A6.3b) [63].
brane at the next node to threshold, thereby initi- In addition, heterosegmental nerve stimula-
ating the production of an action potential here. tions also produce a slow positive potential (het-
The rising phase of the action potential is gener- erosegmental slow positive [HSP] wave) in
ated by an influx of Na + into the axon from the cervical and lumbar enlargements in animal [58,
extracellular space. The loss of Na + causes the 60, 61, 64] and man during wakeful state [61].
extracellular space to become negatively charged; The negative waves, N1, of the segmental
this event is recorded as the negative component SCPs (Fig. A6.3c) are thought to be reflections of
of the triphasic spikes (Fig. A6.3b). The falling changes in the extracellular environment produced
phase of an action potential is caused by a K+ by activity of dorsal horn interneurons (see also
6 Monitoring of Spinal Cord Functions 99

Fig. A6.3 Origins of each component of the segmental space, which is recorded from the cord dorsum as the sec-
SCPs. The initial positivity of the triphasic spike is ond positive component of the triphasic spike (b). The
believed to be a reflection of the extracellular events asso- negative waves, N1 (c), are thought to be reflections of
ciated with the propagation of action potentials through changes in the extracellular environment produced by
the roots down the axon into the spinal cord (a). The rising activity of dorsal horn interneurons. The slow positive
phase of the action potential is generated by an influx of wave, P2, of the segmental SCPs (d) has been demon-
Na + into the axon from the extracellular space resulting in strated as the extracellular manifestation of the process of
the extracellular space becoming negatively charged; this primary afferent depolarization. Positive ionic current
event is recorded as the negative component of the tripha- leaves the extracellular space at excited axon-axonal syn-
sic spike (b). The falling phase of an action potential is apses (sinks) resulting in the dorsal most portion of the
due to a K+ efflux from the axon into the extracellular spinal cord becoming positively charged (d)

Fig. A6.4). When the interneurons are synapti- The slow positive wave, P2, of the segmental
cally activated, positive ionic current leaves the SCPs (Fig. A6.4d) has been demonstrated as the
extracellular space at the synapses (sinks) and extracellular manifestation of the process of ­primary
reappears along the ventrally projecting axons of afferent depolarization (PAD) just as observed
the cells (sources). Thus, the dorsal horn takes on in the spinal animals (Fig. A6.4) [14, 65, 66].
a negative charge and the ventral horn takes on a Positive ionic current leaves the extracellular
positive charge [62, 65, 66]. space at excited axo-axonal synapses (sinks) and
100 S. Aida et al.

3. Ikuta T, Furuta N, Kihara S, Okura M, Nagamine I,

Feed-back relay nuclei Nakayama H, et al. Differences in waveforms of cere-
in brain bral evoked potentials among healthy subjects,
schizophrenics, manic-depressives and epileptics.
J Med Invest. 2007;54:303–15.
4. Vodusek DB. Interventional neurophysiology of the
sacral nervous system. Neurophysiol Clin.
N1-dip 2001;31:239–46.
5. Grundy EL, Nash Jr CL, Brown RH. Arterial pressure
manipulation alters spinal cord function during correc-
tion of scoliosis. Anesthesiology. 1981;54:249–53.
P2S 6. Carenini L, Botacchi E, Camerlingo M, Mamoli
N1α A. Considerations after intraoperative monitoring of
P1 somatosensory evoked potentials during carotid end-
arterectomy. Ital J Neurol Sci. 1989;10:315–20.
N1β 7. Takada T, Denda S, Baba H, Fujioka H, Yamakura T,
Fujihara H, et al. Somatosensory evoked potentials
recorded from the posterior pharynx to stimulation of
P2F the median nerve and cauda equina. Electroencephalogr
P2F
Clin Neurophysiol. 1996;100:493–9.
P2S
8. Shimoji K, Higashi H, Kano T. Epidural recording of
P1
spinal electrogram in man. Electroencephalogr Clin
N-dip Neurophysiol. 1971;30:236–9.
9. Shimoji K. Origins and properties of spinal cord
evoked potentials. In: Dimitrijevic MR, Halter JA,
editors. Atlas of human spinal cord potentials. Boston:
Butterworth-Heinemann; 1995.
10. Aida S, Shimoji K. Descending pathways in spinal
cord stimulation and pain control. In: Bountra C,
N1 α,β Munglani R, Schmidt WK, editors. Pain: current
understanding, emerging therapies, and novel
Fig. A6.4 Proposed origins of each component of the approaches to drug discovery. New York: Marcel
segmental SCPs. The negative waves, N1, are thought to Dekker; 2003. p. 101–17.
be reflections of changes in the extracellular environment 11. Shimoji K, Kitamura H, Ikezono E, Shimizu H,
produced by activity of dorsal horn interneurons. The Okamoto K, Iwakura Y. Spinal hypalgesia and analge-
positive wave, P2, of the segmental SCPs has been demon- sia by low-frequency electrical stimulation in the epi-
strated as the extracellular manifestation of the process of dural space. Anesthesiology. 1974;41:91–4.
primary afferent depolarization and/or intracellular hyper- 12. Magladery LW, Porter WE, Park AM, Teasdal
polarization [68] RD. Electroencephalogical studies of nerve reflex
activity in normal man. IV. The two-neuron reflex and
identification of certain action potentials from spinal
roots and cord. Bull Johns Hopkins Hosp.
reappears along the primary afferents (sources).
1951;88:499–519.
Thus, the dorsal most portion of the spinal cord 13. Fujioka H, Shimoji K, Tomita M, Denda S, Hokari T,
becomes positively charged (Fig. A6.3d) [21, 23, Tohyama M. Effects of dorsal root entry zone lesion
67–72]. Another component, inhibitory postsyn- on spinal cord potentials evoked by segmental,
ascending and descending volleys. Acta Neurochir
aptic potential (IPSP), might be involved in the P2
(Wien). 1992;117:135–42.
wave [68, 69]. 14. Shimoji K, Matsuki M, Shimizu H. Wave-form char-
acteristic and special distribution of evoked spinal
electrogram in man. J Neurosurg. 1977;46:304–13.
15. Fukaya C, Katayama Y. Spinal cord tumor. In: Shimoji
References K, Willis Jr WD, editors. Evoked spinal cord poten-
tials. Tokyo: Springer; 2006. p. 143–9.
1. Hicks JM, Singla A, Shen FH, Arlet V. Complications 16. Becske T, Nelson PK. The vascular anatomy of the
of pedicle screw fixation in scoliosis surgery: a vertebro-spinal axis. Neurosurg Clin N Am. 2009;
systematic review. Spine. 2010;35:E465–70. 20:259–64.
2. Mel MW, Wynn MM, Reeder SB, Tefera G, Hoch JR, 17. de Haan P, Kalkman CJ, de Mol BA, Ubags LH,
Acher CW. A new intercostal artery management Veldman DJ, Jacobs MJ. Efficacy of transcranial
strategy for thoracoabdominal aortic aneurysm repair. motor-evoked myogenic potentials to detect spinal
J Surg Res. 2009;154:99–104. cord ischemia during operations for thoracoabdominal
6 Monitoring of Spinal Cord Functions 101

aneurysms. J Thorac Cardiovasc Surg. 1997;113: Willis WD, editors. Evoked spinal cord potentials.
87–100. Tokyo: Springer; 2006. p. 105–11.
18. Tobita T, Denda S, Takada T, Endoh H, Baba H, 31. Di Lazzaro V, Thickbroom GW, Pilato F, Profice P,
Yamakura T, et al. Effects of fentanyl on spinal cord Dileone M, Mazzone P, et al. Direct demonstration of
potentials and electromyogram evoked by transcranial the effects of repetitive paired-pulse transcranial mag-
magnetic stimulation in man. In: Hashimoto I, Kakigi netic stimulation at I-wave periodicity. Clin
R, editors. Recent advances in human neurophysiol- Neurophysiol. 2007;118:1193–7.
ogy. Amsterdam: Excepta Medica; 1998. p. 1034–7. 32. Davis SF, Altstadt T, Flores R, Kaye A, Oremus
19. Ubags LH, Kalkman CJ, Been HD, Koelman JH, G. Report of seizure following intraoperative moni-
Ongerboer de Visser BW. A comparison of myogenic toring of transcranial motor evoked potentials.
motor evoked responses to electrical and magnetic Ochsner J. 2013;13:558–60.
transcranial stimulation during nitrous oxide/opioid 33. MacDonald DB. Safety of intraoperative transcranial
anesthesia. Anesth Analg. 1999;88:568–72. electrical stimulation motor evoked potential moni-
20. Tomita M, Shimoji K, Denda S, Tobita T, Uchiyama toring. J Clin Neurophysiol. 2002;19:416–29.
S, Baba H. Spinal tracts producing slow components 34. Legat AD. Current practice of motor evoked potential
of spinal cord potentials evoked by descending vol- monitoring: results of a survey. J Clin Neurophysiol.
leys in man. Electroencephalogr Clin Neurophysiol. 2002;19:454–60.
1996;100:68–73. 35. Cheng JS, Ivan ME, Stapleton CJ, Quinones-Hinojosa
21. Shimoji K, Matsuki M, Ito Y, Masuko K, Maruyama A, Gupta N, Auguste KI. Intraoperative changes in
M, Iwane T, et al. Interactions of cord dorsum poten- transcranial motor evoked potentials and somatosen-
tial. J Appl Physiol. 1976;40:79–84. sory evoked potentials predicting outcome in children
22. Shimizu H, Shimoji K, Maruyama Y, Matsuki M, with intramedullary spinal cord tumors. J Neurosurg
Kuribayashi H, Fujioka H. Human spinal cord poten- Pediatr. 2014;13:591–9.
tials produced in lumbosacral enlargement by descend- 36. Shimoji K, Kano T, Nakashima H, Shimizu H. The
ing volleys. J Neurophysiol. 1982;48:1108–20. effects of thiamylal sodium on electrical activities of
23. Maruyama Y, Shimoji K, Shimizu H, Kuribayashi H, the central and peripheral nervous systems in man.
Fujioka H. Human spinal cord potentials evoked by Anesthesiology. 1974;40:234–40.
different sources of stimulation and conduction 37. Tobita T, Okamoto M, Shimizu M, Yamakura T,
velocities along the cord. J Neurophysiol. Fujihara H, Shimoji K, Baba H. The effects of isoflu-
1982;48:1098–107. rane on conditioned inhibition by dorsal column stim-
24. Kawanishi Y, Munakata H, Matsumori M, Tanaka H, ulation. Anesth Analg. 2003;97:436–41.
Yamashita T, Nakagiri K, et al. Usefulness of tran- 38. Kohno T, Kumamoto E, Baba H, Ataka T, Okamoto
scranial motor evoked potentials during thoracoab- M, Shimoji K, Yoshimura M. Actions of midazolam
dominal aortic surgery. Ann Thorac Surg. on GABAergic transmission in substantia gelatinosa
2007;83:456–61. neurons of adult rat spinal cord slices. Anesthesiology.
25. Kawaguchi M, Sakamoto T, Inoue S, Kakimoto M, 2000;92:507–15.
Furuya H, Morimoto T, et al. Low dose propofol as a 39. Shimoji K, Fujiwara N, Fukuda S, Denda S, Takada T,
supplement to ketamine-based anesthesia during Maruyama Y. Effects of isoflurane on spinal inhibi-
intraoperative monitoring of motor-evoked potentials. tory potentials. Anesthesiology. 1990;72:851–7.
Spine. 2000;25:974–9. 40. Kano T, Shimoji K. The effects of ketamine and neu-
26. Chaudhary K, Speights K, McGuire K, White AP. roleptanalgesia on the evoked electrospinogram and
Trans-cranial motor evoked potential detection of electromyogram in man. Anesthesiology. 1974;
femoral nerve injury in trans-psoas lateral lumbar 40:241–6.
interbody fusion. J Clin Monit Comput. 41. Scheufler K-M, Zentner J. Total intravenous anesthe-
2015;29(5):549–54. sia for intraoperative monitoring of the motor path-
27. Ellen R. Grass Lecture: motor evoked potential moni- ways: an integral view combining clinical and
toring. Am J Electroneurodiagnostic Technol. experimental data. J Neurosurg. 2002;96:571–9.
2004;4:223–43. 42. Johanscn JW, Sebcl PS. Development and clinical
28. Novak K, de Camargo AB, Neuwirth M, Kothbauer application of electroencephalographic bispectrum
K, Amassian VE, Deletis V. The refractory period of monitoring. Anesthesiology. 2000;93:1336–44.
fast conducting corticospinal tract axons in man and 43. Criou A, Monchi M, Joly LM, Bellenfant F, Claessens
its implications for intraoperative monitoring of YE, Thébert D, et al. Noninvasive cardiac output
motor evoked potentials. Clin Neurophysiol. monitoring by aortic blood flow determination: evalu-
2004;115:1931–41. ation of the Sometec Dynamo-3000 system. Crit Care
29. Deletis V, Sala F. Intraoperative neurophysiological Med. 1998;26:2066–72.
monitoring of the spinal cord during spinal cord and 44. Kakinohana M, Nakamura S, Fuchigami T, Miyata Y,
spine surgery: a review focus on the corticospinal Sugahara K. Influence of the descending thoracic
tracts. Clin Neurophysiol. 2008;119:248–64. aortic cross clamping on bispectral index value and
30. Tobita T, Shimoji K. Transcranial magnetically plasma propofol concentration in humans.
evoked SCPs [TCM-Evoked SCPs]. In: Shimoji K, Anesthesiology. 2006;104:939–43.
102 S. Aida et al.

45. Dahaba AA. Different conditions that could results in in feedback inhibitory spinal cord potentials in
the bispectral index indicating an incorrect hypnotic ketamine-­anaesthetized rats. Br J Anaesth. 1996;76:
state. Anesth Analg. 2006;101:765–73. 258–65.
46. Fujioka H, Shimoji K, Tomita M, Denda S, Takada H, 59. Shimoji K, Sato Y, Denda S, Takada T, Fukuda S,
Homa T, et al. Spinal cord potential recordings from Hokari T. Slow positive dorsal cord potentials acti-
the extradural space during scoliosis surgery. Br vated by heterosegmental stimuli. Electroencephalogr
J Anaesth. 1994;73:350–6. Clin Neurophysiol. 1992;85:72–80.
47. Piñeiro AM, Cubells C, Garcia P, Castaño GC, 60. Shimoji K, Fujiwara N, Denda S, Tomita M, Toyama
Dávalos A, Coll-Cantia J. Implementation of intraop- M, Fukuda S. Effects of pentobarbital on heteroseg-
erative neurophysiological monitoring during endo- mentally activated dorsal root depolarization in the
vascular procedures in the central nervous system. rat. Investigation by sucrose-gap technique in vivo.
Interv Neurol. 2015;3:85–100. Anesthesiology. 1992;76:958–66.
48. Kim ST, Paeng SH, Jeong DM, Lee KS. Usefulness of 61. Shimoji K, Tomita M, Tobita T, Baba H, Takada T,
intraoperative monitoring during microsurgical Fukuda S, et al. Erb’s point stimulation produces slow
decompression of cervicomedullary compression positive potentials in the human lumbar spinal cord.
caused by an anomalous vertebral artery. J Korean J Clin Neurophysiol. 1994;11:365–74.
Neurosurg Soc. 2014;56:513–6. 62. Shimoji K. Overviews of human (evoked) spinal cord
49. Kondo K, Harada H, Kaneko S, Tyama K, Kano potentials (SCPs): recording methods and terminol-
T. Intraoperative monitoring of the conductive evoked ogy. In: Shimoji K, Willis Jr WD, editors. Evoked
spinal cord potentials under deep hypothermia. spinal cord potentials—an illustrated guide to physi-
J Electrodiag Spinal Cord. 1996;18:160–2. ology, pharmacology, and recording techniques.
50. Costa P, Peretta P, Faccani G. Relevance of intraop- Tokyo: Springer; 2006. p. 40–9.
erative D wave in spine and spinal cord surgeries. 63. Yates BJ, Thompson FJ, Mickle JP. Origin and prop-
Eur Spine J. 2013;22:840–8. erties of spinal cord field potentials. Neurosurgery.
51. Kim SM, Eur Kim SH, Seo DW, Lee KW. 1982;11:439–50.
Intraoperative neurophysiologic monitoring: basic 64. Besson JM, Rivot JP. Spinal interneurones involved in
principles and recent update. J Korean Med Sci. presynaptic controls of supraspinal origin. J Physiol.
2013;28:1261–9. 1973;230:235–54.
52. Pastorelli F, Di Silvestre M, Plasmati R, Michelucci 65. Schmidt RF. Presynaptic inhibition in the vertebrate cen-
R, Greggi T, Morigi A, et al. The prevention of neural tral nervous system. Ergeb Physiol. 1971;63:20–101.
complications in the surgical treatment of scoliosis: 66. Beal JE, Applebaum AE, Forman RD, Willis WD.
the role of the neurophysiological intraoperative mon- Spinal cord potentials evoked by cutaneous afferents
itoring. Eur Spine J. 2011;20 Suppl 1:S105–14. in the monkey. J Neurophysiol. 1977;40:199–211.
53. Shimoji K, Higashi H, Kano T, Asai S, Morioka 67. Shimoji K, Kano T, Higashi H, Morioka T, Henschel
T. Electrical management of intractable pain. Masui EO. Evoked spinal electrograms recorded from epi-
(Jpn J Anesthesiol). 1971;20:44–7. dural space in man. J Appl Physiol. 1972;33:
54. Shimoji K, Matsuki M, Shimizu H, Iwane T, 468–71.
Takahashi R, Maruyama M, Masuko K. Low-­ 68. Tanaka E, Tobita T, Murai Y, Okabe Y, Yamada A,
frequency, weak extradural stimulation in the man- Kano T, et al. Thiamylal antagonizes the inhibitory
agement of intractable pain. Br J Anaesth. effects of dorsal column stimulation on dorsal
1977;49:1081–6. horn activities in humans. Neurosci Res. 2009;64:
55. Shimoji K, Fujiwara N, Sato Y. Development of the 391–6.
system and electrode for simultaneous measurements 69. Bernhard DG. The spinal cord potentials in leads from
of regional blood flow, oxygen pressure and electrical the cord dorsum in relation to peripheral source of
activity [in Japanese]. Byotai Seiri. 1986;11:876–80. afferent stimulation. Acta Physiol Scand. 1953;29
56. Kano T, Shimoji K. Influence of anesthesia on intra- Suppl 106:1–29.
operative monitoring of SCEPs. In: Dimitrijevic MR, 70. Eccles JC, Schmidt R, Willis WD. Pharmacological
Halter JA, editors. Atlas of human spinal cord evoked studies on presynaptic inhibition. J Physiol.
potentials. Boston: Butterworth-Heinemann; 1995. 1963;168:500–30.
p. 97–106. 71. Shimoji K, Ito Y, Ohama K, Sawa T, Ikezono
57. Hayatsu K, Tomita M, Fujihara H, Baba H, Yamakura E. Presynaptic inhibition in man during anesthesia
T, Taga K, et al. The placement of the epidural cathe- and sleep. Anesthesiology. 1975;43:388–91.
ter at the predicted site by electrical stimulation test. 72. Maruyama Y, Shimoji K, Shimizu H, Sato Y,
Anesth Analg. 2001;93:1035–9. Kuribayashi H, Kaieda R. Effects of morphine of
58. Denda S, Shimoji K, Tomita M, Baba H, Yamakura T, human spinal cord and peripheral nervous activities.
Masaki H, et al. Central nuclei and spinal pathways Pain. 1980;8:63–73.
 Electromyography
7
J. Richard Toleikis

of the spine, it has become the standard of care

Introduction for the surgical management of spinal defor-
mity, degenerative spinal disease, and traumatic
Intraoperative neurophysiologic monitoring insults to use pedicle screws to hold rods in
(IOM) using either electrically elicited trig- place for the purpose of segmental transpedicu-
gered or mechanically elicited spontaneous lar fixation. Although the scope of their usage is
electromyographic (EMG) activity has become increasing, these screws are most commonly
widely used for the preservation of neurologic utilized in the most caudal aspects of the spine
function during various surgical procedures. Its from L2 to S1, where their usage has resulted in
use has a long history dating back to the 1960s primarily placing nerve root rather than spinal
when it was first used for the preservation of cord (which ends in the conus medullaris at
facial nerve function [1, 2]. Its use at that time about L1–L2) [24] function at risk.
was prompted by the high incidence of loss of It is now widely recognized that the use of
such function during surgical procedures monitoring can improve surgical outcomes by
involving tumors of the acoustic nerve. As tech- several means. It can help to reduce the risk of
niques evolved for performing surgery in the surgically induced injuries and can help to prop-
cerebellopontine angle, on large skull base erly identify specific neural structures. As a
tumors, and for other intra- and extra-cranial result, the scope of its use during various neuro-
procedures, monitoring techniques for assessing surgical and orthopedic surgical procedures has
and preserving the function of the auditory significantly expanded. The EMG monitoring
nerve and other cranial nerves also evolved [3– techniques that are used for this purpose provide
23]. Much the same can be said regarding the an inexpensive and effective means for assessing
advent of the revolutionary use of metallic, functional integrity such that, if needed, surgical
internal fixation devices for the treatment of spi- intervention and correction can occur in a timely
nal deformity. In the lumbosacral and increas- fashion so as to preserve function. These tech-
ingly in the thoracolumbar and cervical regions niques that utilize EMG responses to protect
neurologic function when brain, spinal cord, cra-
nial nerve, cauda equina, and nerve root function
J.R. Toleikis, Ph.D. (*) are at risk, include cranial nerve monitoring,
Department of Anesthesiology, Rush University brainstem and cortical motor strip mapping,
Medical Center, 1653 West Congress Parkway,
Suite 739 Jelke, Chicago, IL 60612-3833, USA
nerve root monitoring for surgeries in the region
e-mail: [email protected] of the cauda equina, nerve root monitoring during

© Springer International Publishing AG 2017 103

A. Koht et al. (eds.), Monitoring the Nervous System for Anesthesiologists
and Other Health Care Professionals, DOI 10.1007/978-3-319-46542-5_7
104 J.R. Toleikis

pedicle screw placements, H-reflex testing, and However, the way in which these contractions are
transcranial motor-evoked responses. The latter detected has changed. Initially, an observer
technique is only mentioned here and will be fur- detected the muscular contractions of the facial
ther discussed in Chap. 2 (“Transcranial Motor-­ musculature [1, 2]. Later, mechanotransducers
Evoked Potentials”) and elsewhere. [25, 26] were used for this purpose but it was
soon found that these contractions could be
detected by simply recording the EMG activity
 eneral Principle of EMG
G from facial muscles [8, 9, 27]. As a result, during
Monitoring a surgical procedure, the general principle for
monitoring cranial nerve function involves the
The methodology for protecting neurologic func- use of a handheld monopolar electrical stimulat-
tion during various neuro- and orthopedic surgi- ing electrode to probe the surgical field and sin-
cal procedures using EMG techniques is gle or pairs of needle electrodes that are used to
relatively simple and straightforward. The brain, record EMG activity from muscle groups inner-
spinal cord, cranial nerves, and spinal nerve roots vated by cranial nerves at risk [28] (Fig. 7.1,
contain motor pathways that innervate various Table 7.1). During procedures involving tumor
muscle groups. By placing recording electrodes removal, short constant-voltage [28] or constant-­
in these muscles and with controlled usage of current pulses are typically used intermittently
neuromuscular junction blocking agents, EMG for stimulation and recordings of spontaneous or
activity can be acquired when these pathways are free-running EMG activity are continuously
irritated or stimulated due to surgical manipula- monitored. Stimulus pulse durations are usually
tion or activated by means of an external electri- 50–100 μs, with a rate of stimulation of 3–5
cal stimulus. pulses per second [10]. The intensity of the stim-
The monitoring of cranial nerve function is ulus is an important consideration. When using
dependent on being able to record EMG activity. monopolar stimulation, nerves that are located
Such activity cannot be recorded if the patient is within its sphere of influence will be activated;
pharmacologically relaxed. Hence, if muscle the more intense the stimulus, the larger the
relaxants are given, they should be short acting diameter of its sphere of influence. In this way, it
and should be allowed to wear off or reversed is possible to identify regions of a tumor where
before monitoring is attempted. To avoid there is no motor component of a cranial nerve
extended paralysis as a result of the use of neuro- present so that large portions of the tumor can be
muscular blocking agents, there should be good quickly resected with a low risk of causing any
communication between all members of the sur- permanent neurologic injury. A stimulus intensity
gical and monitoring teams regarding the use of that is too low can result in unintended surgical
relaxants and when the use of monitoring will be nerve injury because the nerve may not react to
needed. In addition, a means of accurately assess- the stimulus intensity and may be part of tissue
ing the degree of muscle relaxation should be that is resected. A stimulus intensity that is too
used in order to validate the monitoring findings high may give a false indication that the nerve is
when relaxants are utilized. located close to the stimulator and may slow
down and lead to incomplete tumor removal. At
the same time, the same methodology can be
 ranial Nerve Monitoring
C used to identify the anatomic location of cranial
Technique nerves in order to avoid surgical manipulation
and preserve functional integrity. Initially keep-
The technique that was initially used for monitor- ing the stimulus intensity high will protect the
ing cranial nerve function in the 1960s involved a nerve from damage, and once a response is
handheld stimulation electrode to elicit muscle obtained, the intensity should be reduced in order
contractions. The methodology for eliciting these to determine its exact location. Therefore, when
contractions has not changed to any great extent. using constant-current stimulation, initially the
7 Electromyography 105

Fig. 7.1 Placement of electrodes for

recording EMG responses from the
extraocular muscles (CNs III, IV, VI),
facial muscles (CN VII), masseter
muscle (CN V), trapezius muscle (CN
XI), and the tongue (CN XII). Cranial
nerve IX is monitored by either
placing needles or an adhesive surface
electrode in or on the soft palate.
Cranial nerve X can be monitored by
placing needle electrodes in the vocal
folds, percutaneous needle electrodes
in the larynx muscles, or by using a
tracheal tube with built-in surface
electrodes. Also included are
electrodes for recording auditory
brainstem responses (ABRs) or
brainstem auditory evoked potentials
(BAEPs) and visual evoked potentials
(VEPs). Also shown are insert
earphones for presenting a click
stimulus to the inner ear and eliciting
the ABRs and a contact lens with
light-emitting diodes for stimulating
the eye and eliciting the VEPs. As a
result, the function of cranial nerves
VIII and II can be assessed as well.
Reprinted from Moller et al. [28]

Table 7.1 Muscle groups commonly used to assess cra- flat or quiet, indicating that the nerve has not
nial nerve function based on their innervations become activated as a result of mechanical stimu-
Cranial nerve Muscle group lation. Manipulation of a nerve may result in peri-
III Medial rectus ods of activation. The length of time a nerve is
IV Superior oblique activated depends on the degree of nerve irritation
V Masseter, temporalis [4, 29]. Short periods of activation generally cor-
VI Lateral rectus relate with no permanent injury. Frequent or sus-
VII Orbicularis oris, orbicularis oculi tained periods of activation have a greater
IX Soft palate likelihood of being associated with a postoperative
X Vocal cords
neurologic deficit [30]. This type of sustained acti-
XI Trapezius
vation has been given the name “A-train” activity,
XII Tongue
the occurrence of which has been associated with
postoperative paresis in patients operated on for
stimulus may be as high as a few milliamps, but vestibular schwannoma [11, 13, 31] and microvas-
once a response has been obtained, 0.1–0.2 mA cular decompression for trigeminal neuralgia [12,
may be sufficient to elicit a robust EMG response 27, 28]. The condition of a nerve prior to manipu-
when cranial nerves are directly stimulated [10]. lation also plays a role in determining how it will
As below, continuous monitoring of spontane- react to mechanical stimulation. Manipulation of
ous or free-running EMG activity provides a normal healthy nerves will generally produce little
means for assessing if an injury may occur to a or no activation, whereas nerves that are already
cranial nerve. Normally, the activity should remain slightly injured will tend to react more strongly
106 J.R. Toleikis

when manipulated and may act as impulse genera- cranial or auditory nerve is clearly at risk. Unlike
tors of spontaneous EMG activity even when no the other cranial nerves where the monitoring is
manipulation is taking place [4, 30]. However, dependent upon recording EMG activity, the
severely injured nerves may not react to mechani- auditory nerve consists of sensory pathways and,
cal stimulation at all and therefore the absence of as a result, monitoring of auditory nerve function
EMG activity does not guarantee that a nerve has is dependent on assessing sensory pathway func-
not sustained an injury. As a result, it is important tion using brainstem ABRs. This monitoring
to use electrical stimulation to validate that an technique is thoroughly discussed in Chap. 3
injury has not occurred. If the responses that result (“Auditory Evoked Potentials”).
have latencies or amplitudes that are prolonged or
diminished relative to baseline responses, these
are indications that an injury has occurred. Cranial Nerves III, IV, and VI

Cranial nerves III, IV, and VI innervate the extra-

 onitoring Specific Cranial Nerve
M ocular muscles. Hence, as is the case when moni-
Function toring facial and trigeminal nerve function, if
needle electrodes are placed in or in close prox-
Cranial Nerves V and VII imity to these muscles, cranial nerve function can
be monitored by recording the EMG activity
Historically, monitoring cranial nerve function from the muscles that these nerves innervate [27,
began with attempts to assess and preserve the 28]. Cranial nerve III function can be monitored
motor function of the facial (VIIth cranial nerve) by recording from the medial rectus muscle.
and trigeminal (Vth cranial nerve) nerves during Similarly, cranial nerve IV function is monitored
operations for removal of large acoustic or other by recording from the superior oblique muscle.
skull base tumors [1, 2]. The facial nerve pro- Recordings from the lateral rectus muscle are
vides innervation to the orbicularis oculi and used to monitor cranial nerve VI function [21].
orbicularis oris muscles, whereas the trigeminal Because of space limitations, only single elec-
nerve innervates the muscles of mastication trodes are placed in each of these muscles with a
(masseter and temporalis muscles). Therefore, if reference electrode placed contralateral to the
pairs of needle electrodes are placed in these side of surgery.
muscles, they provide a means for recording
EMG activity during tumor removal and other
surgical procedures [12–15, 19, 20, 32]. If each Cranial Nerves IX, X, XI, and XII
pair is connected to a different recording channel,
the recordings provide a means for differentiating As was the case with monitoring cranial nerve III,
between which muscle group is being activated IV, and VI function using the EMG activity from
and hence which cranial nerve is being irritated or extraocular muscles, the function of the IX, X, XI,
stimulated. Electrical stimulation of the trigemi- and XII cranial nerves is assessed by monitoring
nal nerve will result in responses with a peak the EMG activity from muscles innervated by the
latency less than 6 ms, whereas the responses elic- motor components of these cranial nerves [15–20,
ited by facial nerve stimulation will have peak 33]. Although these cranial nerves also contain
latencies greater than 8 ms [30]. This provides sensory and autonomic components, it is assumed
another means for differentiating which cranial that the EMG activity associated with these nerves
nerve is being activated. represents the condition of the entire nerve and
During operations for removal of large acous- not just the motor component. The motor compo-
tic tumors or during operations in the cerebello- nent of the IXth cranial nerve is monitored by
pontine angle, the function of the component recording the EMG activity from the soft palate
branches (acoustic and vestibular) of the VIIIth using needle electrodes [9, 27]. Although it is
7 Electromyography 107

possible to monitor the motor function of the Xth stimulation provides a means for continuous
cranial nerve (the vagus nerve) using needle elec- monitoring of cortibulbar pathway function. The
trodes placed in the vocal folds, it is difficult to technique involves the use of the same standard
place these electrodes [9]. As a result, recording scalp stimulation sites and parameters that are
of vagal nerve EMG activity is typically done routinely utilized to elicit transcranial motor-­
using a tracheal tube with attached electrodes that evoked potentials (tcMEPs) (see Chap. 2,
are able to make contact with the vocal cords. “Transcranial Motor-Evoked Potentials”).
Monitoring of recurrent laryngeal nerve (a branch However, rather than recording these MEP
of the vagus nerve) function during surgery is a responses from muscles in the extremities, the
common application of this technique [16, 17]. recordings are acquired from the same muscles in
Monitoring the motor function of the spinal acces- the face and head that are routinely used to moni-
sory or XIth cranial nerve is relatively easy. This tor triggered and spontaneous EMG activity
is done by placing needle electrodes in the trape- when assessing cranial nerve function in the tra-
zius muscle and recording the ongoing EMG ditional manner as discussed earlier. These
activity. Finally, monitoring of the XIIth cranial responses are known as corticobulbar tract
or hypoglossal nerve is done by placing needle motor-­evoked potentials (CBT-MEPs). The suc-
electrodes in the tongue [9, 28], although other cessful use of this corticobulbar technique during
techniques have been shown to be effective as cerebellopontine angle (CPA) tumor removal
well [34]. The hypoglossal nerve is very small but when facial nerve function is at risk or for other
very important. Therefore, if there is any question skull base surgeries when the function of other
regarding its location or identity, electrical stimu- cranial nerves is at risk, including the vagal nerve
lation can elicit EMG responses from the tongue pathways, has been reported [35–38].
or other appropriate muscles. When monitoring
the function of all of these cranial nerves, caution
must be taken with regard to the stimulation inten- Brainstem Mapping
sity that is used for eliciting EMG responses from
the muscles that they innervate. If the intensity is Brainstem mapping is a neurophysiologic tech-
too high, the resulting responses may be exces- nique for locating the cranial nerve motor nuclei
sive and possibly injurious. (CMN) on the floor of the fourth ventricle. This
Although the amplitude of the EMG responses technique has proven to be valuable for preventing
resulting from electrical stimulation of the vari- damage to the CMN during surgery for the removal
ous cranial nerves may vary, the responses are of tumors and other pathologies located in and
typically monitored by viewing their traces on an around the brainstem [20, 39–43]. Occasionally,
oscilloscope-type display with each trace dedi- the surgeon is confronted with having to make a
cated to a particular cranial nerve. In addition, the decision regarding the safest approach to reach
surgeon may want to receive auditory feedback structures below the brainstem surface. When this
of the stimulus and any elicited EMG activity approach involves the floor of the fourth ventricle,
each time they stimulate. Therefore, most moni- stimulation of the floor at various locations using a
toring equipment comes equipped with a speaker handheld stimulation probe can result in EMG
system for this purpose. responses from various muscle groups that are
However, the use of electrical stimulation to innervated by cranial motor nerves of the head.
test cranial nerve function is necessarily only These muscle responses can provide information
intermittent and, in addition, sometimes a nerve regarding the location of various CMN and also
will be inaccessible to test because of the size of safe entry points to the brainstem. The motor
a tumor. Hence, even when a nerve is available nuclei are generally located near specific anatomic
for testing, this method cannot provide the landmarks such as the facial colliculus and striae
desired ongoing functional assessment. In such medullares. However, even in normal patients,
cases, the use of multipulse transcranial electrical visualization of these landmarks is not always
108 J.R. Toleikis

apparent; when these landmarks become distorted Anesthetic Management

due to the presence of a tumor, the locations of
these nuclei become even more problematic. This Because the EMG responses used for mapping
is when this technique is of particular value. purposes result from the activation of lower
However, because it is only used intermittently motor neurons, the anesthetics used for general
during tumor resection to localize and confirm the anesthesia have little or no effect on these
location of CMN, it is only a mapping and not a responses. Because lower motor neurons are
monitoring technique. Unlike monitoring tech- being stimulated via the motor nuclei of cranial
niques, it is not used continuously to assess func- nerves or their intramedullary roots, as long as
tion and validate the integrity of neural pathways. muscle relaxants are not utilized during the map-
Therefore, any neural damage that might occur ping and monitoring period, any type of anes-
during brainstem tumor resection would not be thetic management is compatible with brain stem
preventable using this technique. mapping.

Mapping Technique Motor Strip Mapping

Stimulation of the floor of the fourth ventricle is Identification of the motor cortex is typically per-
performed using a handheld monopolar stimula- formed by placing a recording grid of electrodes
tion probe with a tip that allows for very focal directly on the cortical surface of the pre- and
stimulation. The anode or return electrode is gen- post-central gyri. The contralateral median or
erally placed at Fz, near the front of the scalp. ulnar nerves are then stimulated and the resulting
Stimulation consists of 0.2-ms pulses presented somatosensory-evoked potentials (SSEPs) can be
at a frequency of 4 Hz. The stimulation intensity recorded from each of the grid electrodes. At the
is kept low and generally begins at 1.5– transition between the sensory and motor cortex,
2.0 mA. Once muscle responses are obtained, the the elicited responses reverse polarity or undergo
intensity is gradually reduced in order to estab- a phase reversal. In this way, in most instances,
lish stimulation thresholds. This is generally the location of the motor cortex can easily be
between 0.3 and 2.0 mA [43]. identified. However, with distortion of the corti-
As is the case with cranial nerve monitoring, cal anatomy that can occur as a result of the pres-
recording electrodes are inserted into the appro- ence of a brain tumor, these results can become
priate muscle groups. These are placed in the unreliable. Phase reversal may not be obtained
extraocular muscles for mapping the cranial and other techniques may have to be used. One of
nerve III, IV, and VI motor nuclei. For the CMN these techniques utilizes direct electrical stimula-
VII, they are placed into the orbicularis oris and tion of the cortical surface to elicit motor
oculi muscles. For mapping the cranial nerve IX responses that can be visualized or to elicit EMG
and X motor nuclei, the electrodes are inserted responses that can be recorded from various mus-
into the soft palate and posterior pharyngeal wall cle groups of the limbs, face, or trunk [44].
using direct laryngoscopy. Alternatively, a tra- Location of the lesion within or adjacent to the
cheal tube with attached electrodes can be used. motor cortex will dictate what muscle groups to
For the CMN XI, electrodes are placed in the tra- focus on. Subdermal needle electrodes are then
pezius muscle, and for the CMN XII, electrodes placed within the appropriate muscles.
are inserted into the lateral aspect of the tongue. The surgeon uses a handheld stimulator to
The EMG responses that result from stimulation excite the cortex. The electrical stimulation tech-
of these nuclei are typically several hundred nique relies on the activation of the cortical cir-
microvolts in amplitude and can be recorded cuitry by means of a train of bipolar, biphasic
using a time base of 20 ms and a filter bandwidth short-duration (1 ms) rectangular electrical
between 50 and 2000 Hz. pulses. These pulses are applied at a rate of
7 Electromyography 109

50–60 Hz. The intensity of the stimulus is nor- actions by themselves could affect postoperative
mally kept low (3–5 mA) and is gradually outcomes. As a result, fluoroscopy has largely
increased in 1- to 2-mA steps until motor been relied on to detect misplaced screws. The
responses are elicited. The duration of the stimu- incidence of screws associated with neurologic
lus train is typically applied for a few seconds in functional impairment has been reported to range
order to detect any elicited motor responses in the from 1 % to more than 11 % [46–49]. Improved
spontaneous EMG activity that is being observed. imaging technology, including intraoperative
The surgeon notes the location of these occur- computed axial tomography (CAT) scans and ste-
rences and in this way is able to determine what reotactic imaging technology may help to reduce
tissue can be removed while still preserving func- the incidence of misplaced screws. However, in
tion. Chapter 9 (“Cortical Mapping”) provides a the meantime, electrophysiologic techniques have
more comprehensive discussion of this topic. evolved for monitoring and assessing screw place-
ments that rely on the use of both spontaneous and
triggered EMG activity.
 echniques for Assessing Nerve
T
Root Function and Pedicle Screw
Placement Methodology

The use of pedicle screws for spinal stabilization is As indicated earlier, the purpose of monitoring dur-
becoming increasingly more common. However, ing surgical procedures involving pedicle screw
proper placement of pedicle screws such that they placements is to protect and preserve nerve root
do not irritate or injure nerve roots requires that the function. Because all nerve roots consist of both
surgeon doing the screw placement be very expe- sensory and motor fibers, the monitoring tech-
rienced and knowledgeable about the anatomic niques that can be used to assess nerve root func-
characteristics of all aspects of the spine. Although tion can involve the acquisition of either sensory or
a surgeon will rely on anatomic landmarks and motor responses. The sensory responses that are
fluoroscopy for accurate placement, the placement mediated by a single nerve root are known as der-
is still largely done blindly. Ideally, screw place- matomal evoked responses and can be elicited by
ment should result in the screws being placed electrically stimulating a specific body surface area
within the pedicles with about 1 mm of bone known as a dermatome. These responses are medi-
between the lateral and medial walls of the pedicle ated by the dorsal column pathways and can be
and with no breaches of the pedicle walls. recorded from the scalp, much like SSEPs [50].
However, when significant deformity is present, Conversely, the responses that are mediated by the
even a skilled surgeon can misplace screws. Nerve motor components of a single nerve root result
roots tend to position themselves near the medial from either direct or indirect mechanical or electri-
and inferior aspects of the pedicles as they exit the cal stimulation of these motor components and
spinal canal through the spinal foramen. If screws consist of EMG activity from a group of muscles
are misplaced such that they protrude from the known as a myotome. Myotomes are the motor
pedicle wall in either of these areas, they can cause complement to dermatomes and myotomal distri-
nerve root irritation or injury. The current litera- butions as with dermatomal distributions can be
ture indicates that the incidence of screw place- quite variable between individuals. Whereas a
ments that result in cortical perforations of the myotome is a group of muscles that receives inner-
pedicle wall ranges between 5.4 and 40 % [45]. vation from a specific spinal nerve root, most mus-
Such events might go undetected because most cles receive their efferent innervation from several
surgeons are reluctant to visualize and validate nerve roots. Like the afferent innervation for der-
screw placements unless such actions are war- matomes, the amount and type of efferent innerva-
ranted. To do so would require multiple laminoto- tion to a muscle will vary between individuals.
mies, which is time-consuming; in addition, these Typically, during pedicle screw placements, EMG
110 J.R. Toleikis

activity is recorded from several muscle groups function, EMG monitoring for pedicle screw
using either surface or subdermal needle electrodes placement and for surgical procedures in the
placed over or into these muscles. The selection of region of the cauda equina consists of detecting
what muscle groups to monitor is dictated by which two types of activity: spontaneous or free-run-
spinal nerve roots are at risk for irritation or injury. ning activity and triggered EMG activity. The
Although muscles typically receive their innerva- activity of interest in both cases will consist of
tion from nerve roots associated with more than compound muscle action potentials (CMAPs),
one spinal level, one spinal level generally predom- which are elicited either as a result of mechani-
inates. The selection of what muscles to use for cal or electrical stimulation and are recorded
monitoring purposes is based on knowledge of this from appropriate muscle groups (muscles that
innervation. In most cases, recordings are made are innervated by nerve roots at risk for injury)
from muscles of the lower extremities because using pairs of needle or surface electrodes.
screw placement is generally done in the lumbosa- However, it has been reported that intramuscu-
cral area of the spine where lumbar or sacral nerve lar electrodes are preferred over surface elec-
roots are at risk. However, the mechanical advan- trodes for obtaining these recordings [66]. When
tages of pedicle fixation are not restricted to the monitoring both types of activity, it is assumed
lumbosacral spine. As a result, the use of pedicle that the muscles are sufficiently recovered from
screw placement in both the thoracic [51–60] and any muscle relaxants used so that activity can be
cervical [18, 61] regions of the spine has gained elicited when stimulation occurs. For spontane-
popularity. This has occurred despite the known ous activity, CMAP activity can generally be
risks to the spinal cord, nerve roots, and major detected when train-of-four (TOF) testing pro-
blood vessels in these regions. In such cases, myo- duces only one twitch but having more than one
genic activity can be recorded from muscles inner- twitch is desirable, because the amplitude of the
vated by cervical and/or thoracic nerve roots. Using CMAPS is reduced by muscle relaxants and
this approach, monitoring of thoracic screw place- small CMAPs may be missed if a patient is too
ment was first reported in 2001 [62]. However, this relaxed. In most cases, nerve root irritation is an
type of monitoring has had mixed success for tho- infrequent occurrence. As a result, the tracings
racic screw placements [51, 54, 55, 57–59, 62]. In of spontaneous EMG activity are generally flat
particular, the ability to detect medially misplaced and consist of little or no CMAP activity. When
thoracic screws has been very variable. As with the activity is present, it is generally associated with
placement of lumbosacral instrumentation, nerve nerve root decompression (Fig. 7.2). Depending
root injury can occur as a result of inferior, supe- on the activated nerve root, the muscle groups
rior, or lateral pedicle screw placement. However, a being monitored, and the placement of the
thoracic screw that is misplaced too medially can recording electrodes on or in these muscle
result in a spinal cord injury. To avoid such occur- groups will determine how many muscle groups
rences, a technique has been developed that utilizes are activated when stimulation occurs. This is
multipulse stimulation of the spinal cord, in partic- true for triggered EMG activity as well.
ular the corticospinal tract, to elicit EMG responses Not all nerve roots react in the same manner
from lower extremity musculature when a medial when irritation occurs. Normal nerve roots and
breach occurs [63]. This technique has been shown irritated or regenerating nerves in continuity
to successfully detect such occurrences [64, 65]. react differently to mechanical forces. When
Because a single nerve root typically inner- such forces are statically or rapidly applied to
vates more than one muscle, a choice between normal nerve roots, they result in no elicited
several muscles is possible when selecting a activity or only short-duration trains of CMAP
muscle for monitoring purposes. A list of those activity [68]. However, when similar forces are
muscles that are commonly used for recording applied to nerve roots that are already irritated
purposes and their innervation appears in or injured, they result in periods of firing of
Table 7.2. As with monitoring cranial nerve long duration. In cases of preexisting nerve root
7 Electromyography 111

Table 7.2 Muscle groups commonly used to assess ped- has been reported to correlate with postoperative
icle screw placements based on their innervation
motor deficits [69, 70]. The presence of con-
Spinal Nerve root tinuous activity has been used as a means for
region innervation Muscle groups
determining where spinal decompression is
Cervical C2, C3, C4 Trapezius,
sternocleidomastoid
needed [71] and even the adequacy of decom-
C5, C6 Biceps, deltoid
pression [72]. However, the presence of con-
C6, C7 Triceps, flexor carpi tinuous EMG activity in patients undergoing
radialis surgery for tethered cord syndrome has been
C8, T1 Abductor pollicis brevis, reported to be a poor predictor of postoperative
Abductor digiti minimi outcome, with a sensitivity of 100 % but only a
Thoracic T1, T2, T3, Intercostals specificity of 19 %, and suggests that continu-
T4
ous EMG monitoring and the understanding of
T5, T6 Upper rectus abdominis,
Intercostals
its significance remains an evolving technique
T7, T8 Middle rectus abdominis, [73]. When the significance of EMG activity is
Intercostals questionable, the combined acquisition of
T9, T10, Lower rectus abdominis, MEPs and spontaneous EMG activity has been
T11 Intercostals reported to provide a means for corroborating
T12 Inferior rectus abdominis, the significance of the EMG activity. The pres-
Intercostals
ence of sustained or the sudden loss of EMG
Lumbar L1 Psoas
activity is an indication for acquiring an MEP
L2, L3 Adductor magnus
[74]. The complementary use of both of these
L3, L4 Vastus medialis
monitoring modalities has been shown to be a
L4, L5 Anterior tibialis
L5, S1 Peroneus longus
means for detecting and minimizing both nerve
Sacral S1, S2 Medial gastrocnemius root and spinal cord injuries [74–80].
S2, S3, S4 External anal sphincter During spinal fusion procedures involving
pedicle screws, triggered as well as spontaneous
EMG activity are utilized for monitoring pur-
irritation, recordings of the activity will often poses. Triggered activity is elicited in two ways:
consist of low-­amplitude low-frequency activ- either through direct stimulation of nerve roots at
ity even prior to mechanical irritation. Short risk for threshold determination purposes [45] or
aperiodic bursts of activity are common. via indirect nerve root stimulation when pedicle
Although attention should be paid to these, they screws are stimulated in order to assess their
are rarely indicative of neural injury. As indi- placement [47, 49, 52, 53]. Ideally, these two
cated earlier, they are often associated with stimulation techniques should be used in conjunc-
nerve root decompression and result from tug- tion with one another. Direct stimulation is used
ging and displacement, irrigation, electrocau- to determine if a nerve root has an elevated stimu-
tery, metal-to-metal contact, or application of lation threshold; generally, the result of chronic
soaked pledgets. However, long trains of activ- nerve root compression as is the case when a
ity may be indicative of neural injury and are radiculopathy is present. The indirect stimulation
causes for concern and alarm. They are com- technique relies on data obtained from the stimu-
monly related to sustained traction and com- lation of healthy, normal functioning nerve roots
pression and the more sustained the activity, the and is used for assessing the placement of pedicle
greater the likelihood of nerve root damage. screws [47]. These data indicate that normal nerve
When such activity occurs, it is imperative that roots have an average stimulation threshold of
the surgeon be notified so that corrective mea- about 2 mA and that under these circumstances,
sures can immediately be taken. The presence an indirect stimulation warning threshold of about
of sustained spontaneous activity during both 10 mA is adequate for detecting misplaced pedi-
spinal and intracranial tumor removal surgeries cle screws [47]. However, if a direct stimulation
112 J.R. Toleikis

Fig. 7.2 Spontaneous EMG activity elicited from the left mechanical irritation of a nerve root that is already irri-
anterior tibialis muscle as a result of mechanical irritation tated, or a nerve root injury. Such activity warrants notifi-
of the left L5 nerve root. If the activity is of short duration cation of the surgeon and should be avoided in order to
(<1 s), it is generally considered not to be significant and minimize any chance of a postoperative deficit. Reprinted
rarely results in postoperative sequelae. However, if the from Toleikis et al. [67]; with permission
activity is of a longer duration, it may be caused either by

threshold is elevated (above 2 mA), the indirect nerve roots. In addition, stimulation thresholds
stimulation detection threshold needs to be ele- may be elevated when assessing patients with
vated as well in order to compensate. It has been metabolic disorders such as diabetes. The obvious
reported that the direct stimulation thresholds for way to avoid such findings is to directly stimulate
some chronically compressed nerve roots are as each nerve root at risk in order to ensure that it is
high as 20 mA [81]. If such elevated thresholds functioning normally before pedicle screw stimu-
are not taken into account, they can lead to false- lation. If decompression is already being per-
negative findings, e.g., a screw that is reported to formed, this may seem like a reasonable thing to
be adequately placed when in fact it is not. The do. However, if this is not the case, routine lami-
physiologic factors that can contribute to false- notomies to explore each nerve root are time-con-
negative findings largely pertain to the health sta- suming and not without some degree of risk.
tus of the nerve roots that are involved because Therefore, most surgeons choose not to directly
the criteria for assessing pedicle screw place- stimulate each nerve root at risk prior to pedicle
ments are based on results involving healthy screw stimulation testing. In cases when patients
7 Electromyography 113

do exhibit signs of nerve root malfunction, it is intensity is gradually increased starting at or near
strongly recommended that direct nerve root zero until either CMAPs are elicited from one or
stimulation be utilized in order to establish stimu- more monitored muscle groups or a preset inten-
lation thresholds. sity limit is reached. The lack of elicited CMAP
Although most surgeons prefer to use indirect responses at intensities of 20–30 mA is probably
stimulation for testing the placements of pedicle a sufficient indication that a breach has not
screws, the technique can also be used to test for occurred even if nerve roots have elevated stimu-
breaches in the pedicle wall by stimulating the lation thresholds due to factors such as chronic
markers that are placed in the spinal pedicles. nerve root compression. If EMG responses are
Fluoroscopy is generally used by surgeons to elicited at stimulation intensities lower than a
identify the potential trajectory of pedicle screws, predetermined “warning threshold,” the course
the taps, and/or the pedicle screw holes prior to of action is to advise the surgeon to examine the
screw placements. Stimulation is typically per- pedicle hole or the pedicle screw placement. Just
formed using a ball-tipped probe with a needle as the stimulation parameters may vary among
electrode used as an anode to complete the return groups, so too do the “warning thresholds.” Some
current path (Fig. 7.3). The same assessment cri- have used stimulus intensities of 10 mA or
teria are used to determine if the pedicle wall has greater, whereas others have used intensities of
been breached regardless of whether a marker, 8 mA or less as “warning thresholds” [84, 87–
tap, screw, or hole is stimulated. Although the 90]. As indicated earlier, normal nerve roots have
technique is generally the same regardless of a stimulation threshold of about 2 mA [47]. This
who is providing monitoring services, there is threshold depends on both the amplitude of the
some variability in the stimulation parameters pulsatile stimulus as well as the stimulus pulse
used [47–49, 51–53, 67, 81–86]. Stimulation width. Therefore, different “warning thresholds”
rates have ranged from 1 to 5 Hz with pulse dura- may result from different stimulus parameters.
tions of 50–300 μs. Typically, the stimulation As indicated in Fig. 7.3, stimulation current can

Fig. 7.3 The stimulation technique used to assess pedicle through the pedicle screw and the breach in the pedicle
screw placements that can also be used to test markers, the wall, is expected to excite the nerve root resulting in trig-
tap, and pedicle screw holes. The stimulation current can gered EMG responses at a low stimulus intensity from
take many pathways as it returns to the anodal electrode those muscles that receive innervation from that nerve root.
that is placed in muscle tissue. The current will follow those However, if the screw is in contact with fluid or tissue, cur-
pathways that provide the least resistance. In this case, the rent shunting can occur and only a fraction of the applied
pedicle screw has broken through the wall of the pedicle stimulus current will pass through the screw, resulting in
and is situated very close to an emerging nerve root. The elevated stimulation thresholds and possible false-negative
electrical current, following the path of least resistance findings. Reprinted from Toleikis [94]; with permission
114 J.R. Toleikis

take many pathways but ultimately it will follow indicate when a screw has breached a pedicle wall
those that provide the least resistance. When a such that its placement should not be ignored. The
pedicle is intact, the pathway through bone is lower this “warning threshold” is set, the greater
generally one of high resistance depending on the likelihood that a screw with a lower stimula-
bone density. However, when there is a breach of tion threshold is misplaced and consideration
a medial or inferior pedicle wall, the fluid and tis- should be given to having it removed. The results
sue outside of the wall is likely to provide a path of other studies have suggested that a “warning
of least resistance. Now the distance from a stim- threshold” should be set to at least 8 mA. However,
ulation probe or a screw becomes a factor. The it is more likely that only screws with stimulus
intensity of the current that is present to excite a threshold intensities less than 5 mA pose a real
proximal nerve root will obey Coulomb’s Law danger for nerve root injury [67, 88, 90, 92, 93].
(E = K[Q/r2]), where E = stimulating current pres- Even in cases when radiographs are suggestive of
ent at a nerve root, K = a constant, Q = applied adequate screw placement, low stimulation
stimulating current, and r = pedicle screw to thresholds with resulting visual inspection or a
nerve root distance, and will depend on the postoperative CT have revealed that screw place-
inverse square of the distance between the nerve ments needed to be revised (Fig. 7.4).
root and the pedicle screw [91]. When testing a Another factor that can contribute to stimula-
screw placement, the current intensity that is tion thresholds being artificially elevated is cur-
needed to elicit a CMAP response simply pro- rent shunting, which results from fluid or tissue
vides an indication of whether a breach is likely being in contact with the screws, thus allowing
to have occurred and if it has, whether the posi- current to take an alternate additional pathway
tion of the screw could potentially cause nerve when screws are stimulated [94]. In such cases,
root injury. A “warning threshold” of 10 mA is only a fraction of the current that is being applied
often used to assess whether such a breach has actually passes down the screw and is available to
occurred; it is then verified either visually, or via excite a nerve root if the screw placement has
palpation or fluoroscopy [47, 67]. In one study resulted in a breach of the pedicle wall (see
[67], it was found that on visual inspection by the Fig. 7.3). This is another way in which false-­
operating surgeon, screws that require more than negative findings can occur. In addition, some
7 mA of stimulation to elicit a CMAP response screws are made from a material such as titanium
were associated with no breach, a cracked pedi- alloy, which is not a good current conductor, or
cle, or a slight medial exposure of one or two are coated with a material such as hydroxyapatite,
threads of the pedicle screw. In the surgeons’ which is also not a good conductor and, as a result,
judgment, in most cases, such findings posed no stimulation testing can result in false-­ negative
threat to nerve root functional integrity or of neu- findings [95, 96]. Some screws are constructed in
ral injury and the screws were generally left in such a fashion that they do not allow current flow.
place. It was only those screws with stimulation It has been reported that if the top of polyaxial
thresholds of 5 mA or less that were typically pedicle screws are stimulated rather than the stem
removed or redirected. Those screws with stimu- of the screw, this can result in elevated stimula-
lation thresholds in the 5–7 mA range were tion thresholds [97]. In both instances, stimulation
equally likely to be left in place or removed. It is will result in current taking an alternate pathway
noteworthy that none of the screws that were left to the anode rather than through the screw and can
in place (except one) resulted in new postopera- result in false-negative findings.
tive deficits, including two screws with thresholds More recently, surgical procedures involving a
of 4 and 5 mA—an indication that low stimula- lateral transpsoas approach to achieve interbody
tion thresholds, unless they are far below a “warn- fusion in the lumbar spine using either the extreme
ing threshold,” are unlikely to be associated with lateral interbody fusion (XLIF) (Nuvasive) or
new postoperative neurologic findings. Therefore, direct lateral interbody fusion (DLIF) (Medtronic
a “warning threshold” is only that and is meant to Sofamor-Danek) minimally invasive techniques
7 Electromyography 115

Fig. 7.4 Triggered EMG response elicited from the left tive of a possible breach of the pedicle wall. On visual
anterior tibialis muscle as a result of stimulation of the left L5 inspection of the screw placement, it was found to be located
screw at a stimulus intensity of 4.3 mA. The stimulus inten- in the spinal canal and thus removed. No postoperative defi-
sity was below the 10-mA warning threshold that is sugges- cits resulted. Reprinted from Toleikis [94]; with permission

have become an increasingly popular method to neural structures at risk. As dilation and dissection
treat spinal disease. However, both techniques occur, the ­instrumentation for doing so is electri-
require dissection and dilation through the ilio- fied such that proximity to neural structures is
psoas muscle, which places the lumbosacral identified via triggered EMG responses that result
plexus at risk for injury. While this is occurring, from the application of Coulomb’s law. The close-
the visibility of the neural structures at risk is min- ness of instrumentation to neural structures is
imal. As a result, emphasis has been placed on the determined by the intensity of the stimulation cur-
importance of performing these procedures with rent that is required to elicit triggered EMG
the aid of intraoperative monitoring so as to iden- responses. As instrumentation approaches these
tify the proximity of nerves to the muscle dissec- structures, decreasing amounts of current are
tion plane. To do so, spontaneous and triggered required to elicit CMAP responses. The use of this
EMG monitoring are used often in conjunction technique provides a means for avoiding neural
with other monitoring modalities such as MEPs structures as the instrumentation passes through
and SSEPs. EMG activity is recorded from mus- the iliopsoas muscle on its way to the spine.
cle groups that receive their innervations from However, use of the technique has had mixed
116 J.R. Toleikis

results [98–100], with reports of new postopera- An accurate assessment of the degree of muscle
tive motor neuropraxia despite its use [98, 100]. relaxation is essential. An effective means for
These results appear to be associated with pro- making the assessment is to use the TOF tech-
longed retraction time and suggest that this is a nique. For the hands, the ulnar nerve can be stimu-
predictor of declining nerve integrity [100]. lated at the wrist and CMAPs can be elicited and
recorded from the adductor digiti minimi or thenar
muscles. For the legs, the peroneal nerve can be
Anesthetic Management stimulated at the fibular head and CMAPs can be
recorded from the anterior tibialis muscle.
The degree of muscle relaxation is the only Although the anesthesiologist often has access to
anesthetic factor of concern when myogenic the hands and can make an assessment of the
activity is used for monitoring purposes, and it degree of relaxation, TOF testing for monitoring
cannot be overstated how important it is to have purposes should be done by the person providing
the patient adequately un-relaxed when pedicle the monitoring rather than the anesthesiologist for
screw testing is being performed. When moni- several reasons. First, the anesthesiologist typi-
toring nerve root function using spontaneous cally uses a small portable battery-driven device to
and/or triggered myogenic activity from appro- perform a TOF assessment. Depending on how
priate muscle groups, it is imperative that these these devices are used, findings can sometimes be
muscle groups be reactive to changes in nerve erroneous. Second, the anesthesiologist’s assess-
root function that result from a mechanical ment of TOF testing is a subjective one. It is based
insult such as traction or compression of the on visible twitches from muscle groups that the
nerve roots. The myogenic responses that result anesthesiologist has access to, e.g., either the hand
from such an insult are significantly affected by or facial muscles. However, it is unlikely that these
the degree of muscle relaxation. In addition, muscles will be relaxed to the same degree as leg
when electrical stimulation is used to either muscles, which are the muscles of interest when
directly or indirectly stimulate nerve roots, the pedicle screw stimulation is performed. In addi-
degree of muscle relaxation can significantly tion, the device that the anesthesiologist uses has
influence the stimulation thresholds at which an output stimulation intensity of 80 mA—greater
responses are elicited. Therefore, it would be than that typically used by monitoring personnel
ideal if no muscle relaxants were used when with their devices. As a result, the perception that
monitoring myogenic activity and, in fact, some a patient is sufficiently un-relaxed for testing pur-
neurophysiologists insist on patients being poses based on the responses from face or hand
totally un-relaxed when monitoring. However, muscles may be false and may lead to false-nega-
in many clinical settings, this degree of relax- tive findings. Finally, it is appropriate that the per-
ation may be difficult if not impossible to son providing the monitoring be responsible for
achieve because some surgeons may feel that it ­guaranteeing that the test results are accurate by
compromises their ability to adequately per- doing their own TOF testing of leg musculature.
form surgery. What generally occurs in these Their machine may provide more accurate results
situations is that muscle relaxation is used dur- than that of the anesthesiologist. However, their
ing the exposure period when there is frequent findings can also be erroneous due to technical and
use of the Bovie knife, which makes continuous other factors. Therefore, it is useful for the person
monitoring difficult. In addition, there is a low providing the monitoring and the anesthesiologist
risk of neurologic injury during this time. The to periodically compare relaxant findings.
muscle relaxation is then allowed to wear off so The issue of how relaxed a patient must be in
that the patient is sufficiently un-relaxed, per- order to accurately assess both spontaneous and
mitting monitoring during the periods of sur- triggered myogenic activity remains controversial
gery that are associated with a higher risk of [101]. Chapter 19 (“Anesthesia Management and
injury. Intraoperative Electrophysiological Monitoring”)
7 Electromyography 117

provides a much fuller discussion of muscle relax- However, relaxant levels become a lot more strin-
ants. Clearly, 100 % blockade is not suitable for gent when triggered EMG is used for assessment
making these assessments. On the other hand, purposes. If a patient is too relaxed when testing
although it may be ideal from a monitoring per- is performed, the stimulation thresholds are likely
spective to insist on no blockade for making these to be artificially elevated, which can lead to false-­
assessments, it may not be necessary or even fea- negative findings. It has been reported that the
sible. A surgeon may feel that at least partial mus- minimal criterion for making such assessments is
cle relaxation is essential for adequate exposure the presence of a fourth twitch when TOF testing
and accurate screw placement. From personal is performed [67, 102]. Using an amplitude ratio
experience, as indicated earlier, spontaneous of the fourth to the first twitch to determine ade-
EMG activity can be elicited and observed when quacy of relaxation criteria for accurately assess-
one twitch out of a TOF or 90 % neuromuscular ing pedicle screw placements, our findings
blockade is present but, for accuracy purposes, it suggest that this ratio must be 0.1 or higher [94]
is preferable that patients be less relaxed. (Fig. 7.5). Direct electrical stimulation of nerve

Fig. 7.5 Train-of-four testing. The right side of the figure can result in elevated stimulation thresholds and false-nega-
indicates four EMG responses from the right anterior tibialis tive findings during screw testing. For accurate results, the
muscle that resulted from four 2 Hz 0.3-ms pulse stimuli patient should have at least four twitches present, with a
being presented to the right peroneal nerve at the head of the twitch four to twitch one amplitude ratio of at least 0.1–0.2.
fibula at an intensity of 40 mA. Excessive relaxation levels Reprinted from Toleikis et al. [67]; with permission
118 J.R. Toleikis

roots is another means for determining whether tonia distal to the lesion. The hyporeflexia may
the degree of muscle relaxation is sufficient for be due to an increase in the efficacy of presynap-
accurate assessments. If a healthy nerve root is tic inhibition because this type of spinal cord
stimulated using a constant current stimulation insult disrupts the suprasegmental influences that
intensity between 2 and 4 mA and no myogenic mediate presynaptic inhibition. This hyperpolar-
responses are elicited, it is likely that the degree of ization of motor neurons is thought to be caused
muscle relaxant is contributing to elevated stimu- by decreased suprasegmental facilitation of
lation thresholds and potentially incorrect moni- motor neurons, which normally keeps them
toring findings. slightly hypopolarized. Over time, the presynap-
tic inhibition subsides, resulting in enhanced spi-
nal reflexes.
H-Reflex Testing H-reflex responses reflect the level of excita-
tion of a large percentage of the motor neuron
Two techniques have been traditionally used to pool in the spinal gray matter [104, 105]. As a
monitor spinal cord function: SSEPs and TcMEPs. result, these responses may be more helpful than
These techniques are a means for assessing long SSEPs in detecting intraoperative spinal cord
tract sensory and motor function. Although pres- ischemic insults because gray matter excitability
ervation of this function is necessary in order to is depressed more than white matter excitability
interact with our environment in a normal fash- when such an insult occurs. In animals, the dorsal
ion, the monitoring techniques that are used to horn potentials that result from postsynaptic gray
preserve this function are not sufficient to ensure matter activity disappear within 3–5 min when
that complex coordinated motor function remains cessation of spinal cord perfusion occurs,
intact [103]. Simply preserving the ability to walk whereas SSEPs mediated by the posterior col-
does not guarantee that the ability to dance or to umn white matter tracts remain for 12–15 min
perform other complex motor functions has been [106]. A thorough discussion of this monitoring
preserved. Such functions rely on and are con- modality appears in Chap. 8, “The Use of Reflex
trolled by groups of electrically coupled spinal Responses for IOM.”
cord central pattern generators (CPGs). The com-
ponents that contribute to these generators consist
of descending and propriospinal pathways in addi- H-Reflex Response Acquisition
tion to peripheral input and segmental interneu-
rons. It is the interneurons that integrate and The H-reflex response is a CMAP recorded from
summate the excitatory and inhibitory effects of muscle, which results from what is believed to be
the other components and determine their level of electrical activation of a monosynaptic reflex.
excitation. This level of excitation is reflected in The first part of the reflex pathway is the afferent
the activity that can be elicited and recorded from or sensory portion and involves activation of
muscle groups that they innervate. If there is a dis- large 1a nerve fibers which originate in muscle.
ruption in the input to the interneurons, it will be This activation typically results from stimulation
reflected in the activity of the muscles that rely on of the posterior tibial nerve at the popliteal fossa.
the descending influences of these interneurons. Because the nerve is made up of both sensory and
H-reflex muscle responses are monosynaptic motor components, stimulation results in both
and result from processing at a single segment of orthodromic and antidromic sensory and motor
the spinal cord. They are of short latency, short activity, respectively. With supramaximal stimu-
duration, simple configuration, and high ampli- lation, the antidromic motor activity ascends to
tude. If an acute transection of the spinal cord the spinal cord and invades motor neurons in the
were to occur, the result would be spinal shock, ventral gray matter, which are activated. This
which is characterized by complete paralysis, results in orthodromic motor conduction down
hyporeflexia, loss of sensation, and muscle hypo- the same motor fibers to muscle, which produces
7 Electromyography 119

Fig. 7.6 Monitoring with H-reflex responses. H-reflex SSEPs and MEPs represents the gold standard for monitor-
responses were elicited continuously using 0.1-Hz single-­ ing neural function during spinal surgery. However, the abil-
pulse stimulation (1000 μs duration, 2–35 mA) of the poste- ity to reliably acquire H-reflex responses when SSEPs and
rior tibial nerves at the popliteal fossae and recorded MEPs could not be obtained makes them a potentially useful
bilaterally from the medial gastrocnemius muscles. Use of complement to or substitute for these latter modalities

an F response. The orthodromic sensory activity occur, it is the result of orthodromic conduction
is mediated by the 1a nerve fibers and ascends to from the stimulation site to the muscle. Hence, it
the spinal cord where it enters the dorsal horn via will have a very short latency compared to the
the dorsal roots and then synapses with motor H-reflex response (Fig. 7.6). The H-reflex
neurons. Activation of these motor neurons then response usually reaches its peak amplitude at or
results in efferent orthodromic conduction just before the M-wave appears. Further increases
through motor fibers that are located in the same in stimulation intensity result in larger amplitude
homologous spinal segment as the 1a afferent M responses and smaller amplitude H responses
fibers, which then activates muscles in the legs. until no further increases in the amplitude of the
H-reflex responses can be recorded from several M responses occur and the H responses have
muscles in the adult. However, the reflex is most largely disappeared only to be replaced by the F
often acquired from the gastrocnemius muscle response. Once an H-reflex response has been
following stimulation of the tibial nerve at the acquired and the stimulus intensity is kept con-
popliteal fossa. In order to excite the fast-­ stant for each subsequent stimulus, the responses
conducting, low-threshold 1a fibers, long dura- will be short latency, short duration responses
tion 1000-μs low-intensity pulses are presented at with a simple configuration and constant ampli-
a low stimulation rate (0.1–0.5 Hz) to the periph- tudes. In order to determine if the CMAP response
eral nerve which, as indicated earlier, consists of is an H-reflex response, the amplitude of the
both sensory and motor fibers. The intensity of response should exceed the amplitude of the
the stimulation is gradually increased from 0 mA M-wave, and the waveform configuration and
and because 1a sensory fibers are activated before latency should remain constant each time a stim-
motor fibers, the H-reflex appears before what is ulation pulse is presented. Intraoperative normal
known as the M-wave. When the M-wave does parameters have not been established for the
120 J.R. Toleikis

latency of these responses, which are age and 7. Prass RL, Kinney SE, Hardy RW, Hahn JF, Lüders
H. Acoustic (loudspeaker) facial EMG monitoring:
height dependent. However, in a clinical setting,
II. Use of evoked EMG activity during acoustic
the mean latency has been found to be neuroma resection. Otolaryngol Head Neck Surg.
28.9 ± 2.7 ms in awake human [103]. In the oper- 1987;97:541–51.
ating room, these onset latencies are likely to be 8. Yingling CD, Gardi JN. Intraoperative monitor-
ing of facial and cochlear nerves during acoustic
greater due to decreased limb temperatures.
neuroma surgery. Otolaryngol Clin North Am.
1992;25:413–48.
9. Yingling C, Gardi J. Intraoperative monitoring in
Anesthetic Management skull base surgery. In: Jackler R, Brachmann D,
editors. Neurotology. St. Louis: Mosby Year Book;
1994. p. 967–1002.
The amplitudes of both the H-reflex and M-wave 10. Selesnick SH, Goldsmith DF. Issues in the optimal
responses may be reduced because of the use of selection of a cranial nerve monitoring system. Skull
neuromuscular junction blocking agents. Like Base Surg. 1993;3:230–9.
11. Romstöck J, Strauss C, Fahlbusch R. Continuous
any of the other monitoring modalities that rely
electromyography monitoring of motor cranial
on the presence of either mechanically elicited nerves during cerebellopontine angle surgery.
spontaneous or electrically elicited triggered J Neurosurg. 2000;93:586–93.
myogenic activity to make assessments of neuro- 12. Prell J, Rampp S, Rachinger J, Scheller C, Naraghi
R, Strauss C. Spontaneous electromyographic activ-
logic function, controlled usage of these agents is
ity during microvascular decompression in trigemi-
essential in order to avoid false-negative monitor- nal neuralgia. J Clin Neurophysiol. 2008;25:225–32.
ing findings. The H-reflex can be monitored using 13. Prell J, Rachinger J, Scheller C, Alfieri A, Strauss
maintenance anesthetic management consisting C, Rampp S. A real-time monitoring system for the
facial nerve. Neurosurgery. 2010;66:1064–73.
of less than a 0.5 MAC of any inhalational agent
14. Chiara J, Kinney G, Slimp J, Lee GS, Oliaei S,
and a continuous infusion of any narcotic. This Perkins JA. Facial nerve mapping and monitoring
anesthetic management will also usually allow for in lymphatic malformation surgery. Int J Pediatr
other monitoring modalities such as SSEPs, Otorhinolaryngol. 2009;73:1348–52.
15. Dillon FX. Electromyographic (EMG) neuromoni-
MEPs, and spontaneous and triggered EMG to be
toring in otolaryngology-head and neck surgery.
used in conjunction with H-reflex monitoring. Anesthesiol Clin. 2010;28:423–42.
16. Dimopoulos VG, Chung I, Lee GP, Johnston KW,
Kapsalakis IZ, Smisson HF, et al. Quantitative esti-
mation of the recurrent laryngeal nerve irritation by
References employing spontaneous intraoperative electromyo-
graphic monitoring during anterior cervical discec-
1. Hilger JA. Facial nerve stimulator. Trans Am Acad tomy and fusion. J Spinal Disord Tech. 2009;22:1–7.
Ophthalmol Otolaryngol. 1964;68:74–6. 17. Genther DJ, Kandil EH, Noureldine SI, Tufano
2. Rand RW, Kurze TL. Facial nerve preservation by RP. Correlation of final evoked potential amplitudes
posterior fossa transmeatal microdissection in total on intraoperative electromyography of the recur-
removal of acoustic tumors. J Neurol Neurosurg rent laryngeal nerve with immediate postoperative
Psychiatry. 1965;28:311–6. vocal fold function after thyroid and parathyroid
3. Al-Mefty O, Holoubi A, Rifai A, Fox surgery. JAMA Otolaryngol Head Neck Surg.
JL. Microsurgical removal of suprasellar meningio- 2014;140:124–8.
mas. Neurosurgery. 1985;16:364–72. 18. Holdefer RN, Heffez DS, Cohen BA. Utility of
4. Prass RL, Lüders H. Acoustic (loudspeaker) facial evoked EMG monitoring to improve bone screw
electromyographic monitoring: Part 1. Evoked elec- placements in the cervical spine. J Spinal Disord
tromyographic activity during acoustic neuroma Tech. 2013;26:E163–9.
resection. Neurosurgery. 1986;19:392–400. 19. Jahangiri FR, Minhas M, Jane J. Preventing lower
5. Sekhar LN, Møller AR. Operative management of cranial nerve injuries during fourth ventricle tumor
tumors involving the cavernous sinus. J Neurosurg. resection by utilizing intraoperative neurophysiolog-
1986;64:879–89. ical monitoring. Neurodiagn J. 2012;52:320–32.
6. Harner SG, Daube JR, Ebersold MJ, Beatty 20. Sala F, Manganotti P, Tramontano V, Bricolo A,
CW. Improved preservation of facial nerve function Gerosa M. Monitoring of motor pathways during
with use of electrical monitoring during removal of brain stem surgery: what we have achieved and what
acoustic neuromas. Mayo Clin Proc. 1987;62:92–102. we still miss? Neurophysiol Clin. 2007;37:399–406.
7 Electromyography 121

21. San-juan D, Barges-Coll J, Gómez Amador JL, Díaz electrical stimulation during skull base surgery. Clin
MP, Alarcón AV, Escanio E, et al. Intraoperative mon- Neurophysiol. 2005;16:588–96.
itoring of the abducens nerve in extended endonasal 36. Deletis V, Fernandez-Conejero I, Ulkatan S,
endoscopic approach: a pilot study technical report. Costantino P. Methodology for intraoperatively elic-
J Electromyogr Kinesiol. 2014;24:558–64. iting motor evoked potentials in the vocal muscles
22. Skinner SA. Neurophysiologic monitoring of the by electrical stimulation of the corticobulbar tract.
spinal accessory nerve, hypoglossal nerve, and Clin Neurophysiol. 2009;120:336–41.
the spinomedullary region. J Clin Neurophysiol. 37. Morota N, Ihara S, Deletis V. Intraoperative neuro-
2011;28:587–98. physiology for surgery in and around the brainstem:
23. Son BC, Lee SW, Kim S, Hong JT, Sung JH, Yang role of brainstem mapping and corticobulbar tract
S-H. Transzygomatic approach with intraopera- motor-evoked potential monitoring. Childs Nerv
tive neuromonitoring for resection of middle cra- Syst. 2010;26:513–21.
nial fossa tumors. J Neurol Surg B Skull Base. 38. Deletis V, Fernandez-Conejero I, Ulkatan S, Rogic
2012;73:28–35. M, Carbo EL, Hiltzik D. Methodology for intra-
24. Lin N, Bebawy JF, Hua L, Wang BG. Is spinal anaes- operative recording of the corticobulbar motor
thesia at L2–L3 interspace safe in disorders of the evoked potentials from cricothyroid muscles. Clin
vertebral column? A magnetic resonance imaging Neurophysiol. 2011;122:1883–9.
study. Br J Anaesth. 2010;105:857–62. 39. Katsuta T, Morioka T, Fujii K, Fukui M. Physiological
25. Sugita K, Kobayashi S. Technical and instrumental localization of the facial colliculus during direct sur-
improvements in the surgical treatment of acoustic gery on an intrinsic brain stem lesion. Neurosurgery.
neurinomas. J Neurosurg. 1982;57:747–52. 1993;32:861–3. comment 863.
26. Silverstein H, Smouha E, Jones R. Routine identi- 40. Strauss C, Romstöck J, Nimsky C, Fahlbusch
fication of the facial nerve using electrical stimula- R. Intraoperative identification of motor areas
tion during otological and neurotological surgery. of the rhomboid fossa using direct stimulation.
Laryngoscope. 1988;98:726–30. J Neurosurg. 1993;79:393–9.
27. Moller A. Intraoperative neurophysiologic monitor- 41. Morota N, Deletis V, Epstein FJ, Kofler M, Abbott
ing. Luxembourg: Harwood Academic; 1995. R, Lee M, et al. Brain stem mapping: neurophysi-
28. Moller A. Intraoperative monitoring of evoked ological localization of motor nuclei on the floor of
potentials: an update. In: Wilkins R, Rengachery S, the fourth ventricle. Neurosurgery. 1995;37:922–9.
editors. Neurosurgery update 1: diagnosis, opera- discussion 929–30.
tive technique, and neuro-oncology. New York: 42. Morota N, Deletis V, Lee M, Epstein FJ. Functional
McGraw-­Hill; 1990. p. 169–76. anatomic relationship between brain-stem
29. Daube J. Intraoperative monitoring of cranial tumors and cranial motor nuclei. Neurosurgery.
motor nerves. In: Schramm J, Moller A, editors. 1996;39:787–93. discussion 793–4.
Intraoperative neurophysiologic monitoring in neu- 43. Morota N, Deletis V, Epstein FJ. Brainstem map-
rosurgery. Heidelberg: Springer; 1991. p. 246–67. ping. In: Neurophysiology in neurosurgery. San
30. Moller A. Monitoring and mapping the cranial Diego, CA: Academic; 2002. p. 319–35.
nerves and the brainstem. In: Deletis V, Shils J, edi- 44. Maertens de Noordhout A, Born JD, Hans P,
tors. Neurophysiology in neurosurgery. San Diego, Remacle JM, Delwaide PJ. Intraoperative locali-
CA: Academic; 2002. p. 291–318. sation of the primary motor cortex using single
31. Prell J, Rampp S, Romstöck J, Fahlbusch R, Strauss electrical stimuli. J Neurol Neurosurg Psychiatry.
C. Train time as a quantitative electromyographic 1996;60:442–4.
parameter for facial nerve function in patients 45. Gertzbein SD, Robbins SE. Accuracy of pedicular
undergoing surgery for vestibular schwannoma. screw placement in vivo. Spine. 1990;15:11–4.
J Neurosurg. 2007;106:826–32. 46. Calancie B, Lebwohl N, Madsen P, Klose
32. Lu AY, Yeung JT, Gerrard JL, Michaelides EM, KJ. Intraoperative evoked EMG monitoring in an
Sekula RF, Bulsara KR. Hemifacial spasm and neu- animal model. A new technique for evaluating pedi-
rovascular compression. ScientificWorldJournal. cle screw placement. Spine. 1992;17:1229–35.
2014;2014:349319. 47. Calancie B, Madsen P, Lebwohl N. Stimulus-evoked
33. Karlikaya G, Citçi B, Güçlü B, Türe H, Türe U, EMG monitoring during transpedicular lumbosacral
Bingöl CA. Spinal accessory nerve monitoring spine instrumentation. Initial clinical results. Spine.
in posterior fossa surgery. J Clin Neurophysiol. 1994;19:2780–6.
2008;25:346–50. 48. Clements DH, Morledge DE, Martin WH, Betz
34. Holdefer RN, Kinney GA, Robinson LR, Slimp RR. Evoked and spontaneous electromyography
JC. Alternative sites for intraoperative monitoring of to evaluate lumbosacral pedicle screw placement.
cranial nerves X and XII during intracranial surger- Spine (Phila Pa 1976). 1996;21:600–4.
ies. J Clin Neurophysiol. 2013;30:275–9. 49. Darden BV, Wood KE, Hatley MK, Owen JH,
35. Dong CCJ, MacDonald DB, Akagami R, Westerberg Kostuik J. Evaluation of pedicle screw insertion
B, AlKhani A, Kanaan I, et al. Intraoperative facial monitored by intraoperative evoked electromyogra-
motor evoked potential monitoring with transcranial phy. J Spinal Disord. 1996;9:8–16.
122 J.R. Toleikis

50. Toleikis JR, Carlvin AO, Shapiro DE, Schafer 63. Donohue ML, Murtagh-Schaffer C, Basta J, Moquin
MF. The use of dermatomal evoked responses during RR, Bashir A, Calancie B. Pulse-train stimulation
surgical procedures that use intrapedicular fixation for detecting medial malpositioning of thoracic ped-
of the lumbosacral spine. Spine. 1993;18:2401–7. icle screws. Spine. 2008;33:E378–85.
51. Shi Y, Binette M, Martin WH, Pearson JM, Hart 64. Calancie B, Donohue ML, Harris CB, Canute GW,
RA. Electrical stimulation for intraoperative evalua- Singla A, Wilcoxen KG, et al. Neuromonitoring with
tion of thoracic pedicle screw placement. Spine. pulse-train stimulation for implantation of thoracic
2003;15:595–601. pedicle screws: a blinded and randomized clinical
52. Rodriguez-Olaverri JC, Zimick NC, Merola A, De Blas study. Part 1. Methods and alarm criteria. J Neurosurg
G, Burgos J, Piza-Vallespir G, et al. Using triggered Spine. 2014;20:675–91.
electromyographic threshold in the intercostal muscles 65. Calancie B, Donohue ML, Moquin RR. Neuro­
to evaluate the accuracy of upper thoracic pedicle monitoring with pulse-train stimulation for implan-
screw placement (T3–T6). Spine. 2008;33:E194–7. tation of thoracic pedicle screws: a blinded and
53. Norton JA, Hedden DM. Monitoring placement of randomized clinical study. Part 2. The role of feed-
high thoracic pedicle screws by triggered electromy- back. J Neurosurg Spine. 2014;20:692–704.
ography of the intercostal muscles. Can J Surg. 66. Skinner SA, Transfeldt EE, Savik K. Surface elec-
2009;52:E47–8. trodes are not sufficient to detect neurotonic dis-
54. De Blas G, Barrios C, Regidor I, Montes E, Burgos J, charges: observations in a porcine model and clinical
Pizá-Vallespir G, et al. Safe pedicle screw placement review of deltoid electromyographic monitoring
in thoracic scoliotic curves using t-EMG: stimulation using multiple electrodes. J Clin Monit Comput.
threshold variability at concavity and convexity in 2008;22:131–9.
apex segments. Spine. 2012;37:E387–95. 67. Toleikis JR, Skelly JP, Carlvin AO, Toleikis SC,
55. Lewis SJ, Lenke LG, Raynor B, Long J, Bridwell Bernard TN, Burkus JK, et al. The usefulness of
KH, Padberg A. Triggered electromyographic electrical stimulation for assessing pedicle screw
threshold for accuracy of thoracic pedicle screw placements. J Spinal Disord. 2000;13:283–9.
placement in a porcine model. Spine. 2001;26:2485– 68. Howe JF, Loeser JD, Calvin WH. Mechanosensitivity
9. discussion 2490. of dorsal root ganglia and chronically injured axons:
56. Montes E, De Blas G, Regidor I, Barrios C, Burgos a physiological basis for the radicular pain of nerve
J, Hevia E, et al. Electromyographic thresholds after root compression. Pain. 1977;3:25–41.
thoracic screw stimulation depend on the distance of 69. Holland NR, Kostuik JP. Continuous electromyo-
the screw from the spinal cord and not on pedicle graphic monitoring to detect nerve root injury during
cortex integrity. Spine J. 2012;12:127–32. thoracolumbar scoliosis surgery. Spine.
57. Raynor BL, Lenke LG, Kim Y, Hanson DS, Wilson-­ 1997;22:2547–50.
Holden TJ, Bridwell KH, et al. Can triggered elec- 70. Gläsker S, Pechstein U, Vougioukas VI, Van Velthoven
tromyograph thresholds predict safe thoracic pedicle V. Monitoring motor function during resection of
screw placement? Spine. 2002;27:2030–5. tumours in the lower brain stem and fourth ventricle.
58. Regidor I, de Blas G, Barrios C, Burgos J, Montes E, Childs Nerv Syst. 2006;22:1288–95.
García-Urquiza S, et al. Recording triggered EMG 71. Jimenez JC, Sani S, Braverman B, Deutsch H, Ratliff
thresholds from axillary chest wall electrodes: a new JK. Palsies of the fifth cervical nerve root after cervi-
refined technique for accurate upper thoracic (T2– cal decompression: prevention using continuous
T6) pedicle screw placement. Eur Spine intraoperative electromyography monitoring.
J. 2011;20:1620–5. J Neurosurg Spine. 2005;3:92–7.
59. Samdani AF, Tantorski M, Cahill PJ, Ranade A, 72. Chappuis JL, Johnson G. Using intraoperative electro-
Koch S, Clements DH, et al. Triggered electromyog- physiologic monitoring as a diagnostic tool for deter-
raphy for placement of thoracic pedicle screws: is it mining levels to decompress in the cervical spine: a
reliable? Eur Spine J. 2011;20:869–74. case report. J Spinal Disord Tech. 2007;20:403–7.
60. Silverstein JW, Mermelstein LE. Utilization of para- 73. Paradiso G, Lee GYF, Sarjeant R, Hoang L,
spinal muscles for triggered EMG during thoracic Massicotte EM, Fehlings MG. Multimodality intra-
pedicle screw placement. Am J Electroneurodiagnostic operative neurophysiologic monitoring findings dur-
Technol. 2010;50:37–49. ing surgery for adult tethered cord syndrome:
61. Djurasovic M, Dimar JR, Glassman SD, Edmonds analysis of a series of 44 patients with long-term
HL, Carreon LY. A prospective analysis of intraop- follow-up. Spine. 2006;31:2095–102.
erative electromyographic monitoring of posterior 74. Skinner SA, Transfeldt EE, Mehbod AA, Mullan JC,
cervical screw fixation. J Spinal Disord Tech. Perra JH. Electromyography detects mechanically-­
2005;18:515–8. induced suprasegmental spinal motor tract injury:
62. Danesh-Clough T, Taylor P, Hodgson B, Walton review of decompression at spinal cord level. Clin
M. The use of evoked EMG in detecting misplaced Neurophysiol. 2009;120:754–64.
thoracolumbar pedicle screws. Spine (Phila Pa 75. Bose B, Sestokas AK, Schwartz DM. Neuro­
1976). 2001;26:1313–6. physiological detection of iatrogenic C-5 nerve
7 Electromyography 123

deficit during anterior cervical spinal surgery. 88. Raynor BL, Lenke LG, Bridwell KH, Taylor BA,
J Neurosurg Spine. 2007;6:381–5. Padberg AM. Correlation between low triggered
76. Mok JM, Lyon R, Lieberman JA, Cloyd JM, Burch electromyographic thresholds and lumbar pedicle
S. Monitoring of nerve root injury using transcranial screw malposition: analysis of 4857 screws. Spine.
motor-evoked potentials in a pig model. Spine. 2007;32:2673–8.
2008;33:E465–73. 89. Parker SL, Amin AG, Farber SH, McGirt MJ,
77. Skinner SA, Transfeldt EE. Electromyography in the Sciubba DM, Wolinsky J-P, et al. Ability of electro-
detection of mechanically induced spinal motor tract myographic monitoring to determine the presence of
injury: observations in diverse porcine models. malpositioned pedicle screws in the lumbosacral
J Neurosurg Spine. 2009;11:369–74. spine: analysis of 2450 consecutively placed screws.
78. Macdonald DB, Stigsby B, Al Homoud I, Abalkhail J Neurosurg Spine. 2011;15:130–5.
T, Mokeem A. Utility of motor evoked potentials for 90. Lee CH, Kim HW, Kim HR, Lee CY, Kim JH, Sala
intraoperative nerve root monitoring. J Clin F. Can triggered electromyography thresholds assure
Neurophysiol. 2012;29:118–25. accurate pedicle screw placements? A systematic
79. Bhalodia VM, Schwartz DM, Sestokas AK, review and meta-analysis of diagnostic test accu-
Bloomgarden G, Arkins T, Tomak P, et al. Efficacy racy. Clin Neurophysiol. 2015;126:2019–25.
of intraoperative monitoring of transcranial electri- 91. Urmey WF. Using the nerve stimulator for periph-
cal stimulation-induced motor evoked potentials and eral or plexus nerve blocks. Minerva Anestesiol.
spontaneous electromyography activity to identify 2006;72:467–71.
acute-versus delayed-onset C-5 nerve root palsy dur- 92. Skinner SA, Rippe DM. Threshold testing of lumbo-
ing cervical spine surgery: clinical article. sacral pedicle screws: a reappraisal. J Clin
J Neurosurg Spine. 2013;19:395–402. Neurophysiol. 2012;29:493–501.
80. Fotakopoulos G, Alexiou GA, Pachatouridis D, 93. Nichols GS, Manafov E. Utility of electromyogra-
Karagiorgiadis D, Konitsiotis S, Kyritsis AP, et al. phy for nerve root monitoring during spinal surgery.
The value of transcranial motor-evoked potentials J Clin Neurophysiol. 2012;29:140–8.
and free-running electromyography in surgery for 94. Toleikis J. Neurophysiological monitoring during
cervical disc herniation. J Clin Neurosci. pedicle screw placement. In: Deletis V, Shils J, edi-
2013;20:263–6. tors. Neurophysiology in neurosurgery. San Diego,
81. Holland NR, Lukaczyk TA, Riley LH, Kostuik CA: Academic; 2002. p. 231–64.
JP. Higher electrical stimulus intensities are required 95. Donohue ML, Swaminathan V, Gilbert JL, Fox CW,
to activate chronically compressed nerve roots. Smale J, Moquin RR, et al. Intraoperative
Implications for intraoperative electromyographic ­neuromonitoring: can the results of direct stimulation
pedicle screw testing. Spine. 1998;23:224–7. of titanium-alloy pedicle screws in the thoracic spine
82. Maguire J, Wallace S, Madiga R, Leppanen R, be trusted? J Clin Neurophysiol. 2012;29:502–8.
Draper V. Evaluation of intrapedicular screw posi- 96. Davis TT, Tadlock S, Bernbeck J, Fung DA,
tion using intraoperative evoked electromyography. Molinares DM. Can triggered electromyography be
Spine. 1995;20:1068–74. used to evaluate pedicle screw placement in
83. Glassman SD, Dimar JR, Puno RM, Johnson JR, hydroxyapatite-coated screws: an electrical exami-
Shields CB, Linden RD. A prospective analysis of nation. J Clin Neurophysiol. 2014;31:138–42.
intraoperative electromyographic monitoring of 97. Anderson DG, Wierzbowski LR, Schwartz DM,
pedicle screw placement with computed tomo- Hilibrand AS, Vaccaro AR, Albert TJ. Pedicle screws
graphic scan confirmation. Spine. 1995;20:1375–9. with high electrical resistance: a potential source of error
84. Isley M, Pearlman R, Wadsworth J. Recent advances with stimulus-evoked EMG. Spine. 2002;27:1577–81.
in intraoperative neuromonitoring of spinal cord 98. Houten JK, Alexandre LC, Nasser R, Wollowick
function: pedicle screw stimulation techniques. AL. Nerve injury during the transpsoas approach
Neurodiagn J. 1997;37:93–126. for lumbar fusion. J Neurosurg Spine.
85. Lenke LG, Padberg AM, Russo MH, Bridwell KH, 2011;15:280–4.
Gelb DE. Triggered electromyographic threshold for 99. Jahangiri FR, Sherman JH, Holmberg A, Louis R,
accuracy of pedicle screw placement. An animal model Elias J, Vega-Bermudez F. Protecting the genitofem-
and clinical correlation. Spine. 1995;20:1585–91. oral nerve during direct/extreme lateral interbody
86. Isley MR, Zhang X-F, Balzer JR, Leppanen fusion (DLIF/XLIF) procedures. Am
RE. Current trends in pedicle screw stimulation J Electroneurodiagnostic Technol. 2010;50:321–35.
techniques: lumbosacral, thoracic, and cervical lev- 100. Uribe JS, Isaacs RE, Youssef JA, Khajavi K, Balzer
els. Neurodiagn J. 2012;52:100–75. JR, Kanter AS, et al. Can triggered electromyogra-
87. Bose B, Wierzbowski LR, Sestokas phy monitoring throughout retraction predict post-
AK. Neurophysiologic monitoring of spinal nerve operative symptomatic neuropraxia after XLIF?
root function during instrumented posterior lumbar Results from a prospective multicenter trial. Eur
spine surgery. Spine. 2002;27:1444–50. Spine J. 2015;24 Suppl 3:378–85.
124 J.R. Toleikis

101. Sloan TB. Muscle relaxant use during intraoperative 104. Táboríková H, Sax DS. Motoneurone pool and the
neurophysiologic monitoring. J Clin Monit Comput. H-reflex. J Neurol Neurosurg Psychiatry.
2013;27:35–46. 1968;31:354–61.
102. Holland NR. Intraoperative electromyography dur- 105. Kimura J. Principles of nerve conduction studies. In:
ing thoracolumbar spinal surgery. Spine. Kimura J, editor. Electrodiagnosis in diseases of
1998;23:1915–22. nerve and muscle: principles and practice.
103. Leppanen RE. Intraoperative applications of the Philadelphia: FA Davis; 1983. p. 353–98.
H-reflex and F-response: a tutorial. J Clin Monit 106. Slimp JC. Electrophysiologic intraoperative moni-
Comput. 2006;20:267–304. toring for spine procedures. Phys Med Rehabil Clin
N Am. 2004;15:85–105.
 The Use of Reflex Responses
for IOM 8
Ronald Leppanen

nal cord systems activating the motor units are

Introduction not monitored [1]. Our ability to interact with the
environment involves not only intact sensation
There are three goals for intraoperative neuro- and strength but also complex coordinated motor
physiologic monitoring. The first is to reduce the behavior. SSEPs and MEPs monitor those sys-
risk of neurologic complications by detecting tems that are responsible for sensation and
insult to neuronal structures. The second is to strength. Reflexes can be used to monitor com-
provide guidance that may affect a surgeon’s plex integrated motor behavior. When used in
approach or actions, such as mapping the loca- conjunction with SSEPs, tcMEPs, free-run and
tion of sensory and motor tracts within the spinal electrically stimulated electromyography reflexes
cord. The third is to perform studies detailed provide a multisystems approach to monitoring.
enough to help understand normal and patho- The advantage of the use of intraoperative
physiologic function. Intraoperative reflex tech- reflexes to detect neuronal compromise is that the
niques are used to help accomplish these three recordings are single sweep. They are real time
goals. They are used to monitor the function of with no delay after the onset of compromise that
peripheral nerve, plexus, nerve root, and segmen- is present when using averaged SEPs. They pro-
tal and suprasegmental function. These reflex vide the surgeon with immediate feedback and
techniques will be reviewed in this chapter. can be acquired continuously throughout surgery
At the spinal cord level, orthodromic ascend- with little or no noticeable patient movement.
ing somatosensory-evoked potentials (SSEPs) Reflexes may be monitored in patients in whom
monitor spinal cord sensory function. Motor-­ SSEPs and tcMEPs cannot be recorded because
evoked potentials (MEPs) recorded with tran- of a pre-existing neurologic deficit [1–4].
scranial electrical motor stimulation (tcMEPs)
monitors spinal cord motor function but only
4–5 % of the motor units innervating a muscle are Anatomy and Neurophysiology
activated. Therefore, 95–96 % of the motor spi-
Reflex processing can be considered relatively
simple. An example is the monosynaptic or oli-
R. Leppanen, Ph.D. (*) gosynaptic H-reflex that involves simple spinal
Clinical Neurophysiologist, Knoxville Neurology cord processing. It can also be considered
Clinic, 2200 Sutherland Avenue,
Knoxville, TN 37919, USA complex, with polysynaptic reflexes involving
e-mail: [email protected] processing at multiple spinal cord levels.

© Springer International Publishing AG 2017 125

A. Koht et al. (eds.), Monitoring the Nervous System for Anesthesiologists
and Other Health Care Professionals, DOI 10.1007/978-3-319-46542-5_8
126 R. Leppanen

Monosynaptic reflex muscle recordings are of provide information about the degree of coupling
short latency, short duration, simple configura- between CPGs. Acute or chronic damage to the
tion, and high amplitude. These parameters are peripheral afferent input, the descending supra-
stable and vary little from one stimulus to the segmental systems, the propriospinal systems, or
next. Polysynaptic recordings are of longer the segmental interneurons results in the uncou-
latency, longer duration, complex configura- pling of these components. This changes the
tion, and low amplitude. Polysynaptic record- level of excitability of the segmental interneu-
ings are not stable and vary from one stimulus rons, which results in a change in the segmental
to the next [5]. reflex gain. The change in the reflex gain can be
At the spinal cord level, one approach to detected by changes in reflex processing recorded
understanding complex motor behavior is to con- from muscles. Changes in reflex processing can
sider spinal cord integrating function to be con- be used as a monitoring technique to detect acute
trolled by a system of tightly electrically coupled and chronic compromise of the spinal cord supra-
central pattern generators (CPGs). The integrated segmentally and segmentally, and of the nerve
activity of these spinal cord CPGs is responsible roots and peripheral nerve (Fig. 8.1).
for controlling the stepping mechanism of gait
and the coordination of upper and lower extrem-
ity function [6–10]. Spinal Cord Pathophysiology
Spinal cord CPGs may be thought of as having
four components: the segmental interneurons, the Acute spinal cord transection causes spinal shock
descending suprasegmental systems, the proprio- that is characterized by complete paralysis, hypo-
spinal systems, and the peripheral afferent input. reflexia, loss of sensation, and muscle hypotonia
The point of control is the level of excitation of caudal to the lesion. Spinal cord injury (SCI) dis-
interneurons, which is determined by the inte- rupts or disinhibits the suprasegmental influence
grated summated synaptic excitatory and inhibi- over segmental interneurons mediating presyn-
tory effect of the other components on the aptic inhibition. The hyporeflexia associated with
interneurons and motor neurons. The level of spinal shock may be due to an increase in the
excitation of interneurons determines the level of efficacy of presynaptic inhibition [12]. Over
reflex gain. Sensory afferent and antidromic time, presynaptic activity decreases, resulting in
motor response signals following peripheral enhanced spinal reflexes [13]. Observations in
nerve stimulation provide the time-locked syn- cats indicate that rostral acute SCI causes post-
chronization of the system. Summated activity synaptic caudal lumbar motor neuron changes. In
from descending spinal cord systems, especially cats, rostral acute SCI causes hyperpolarization
the corticospinal, rubrospinal, vestibulospinal, of caudal lumbar segment motor neurons [14–
and reticulospinal systems, contribute to con- 17]. Immediately following rostral spinal cord
trolling the gain set by the interneurons. transection, the monosynaptic reflexes from the
Vestibulospinal and reticulospinal tracts control medial and lateral gastrocnemius, soleus, poste-
proximal function, and rubrospinal and cortico- rior biceps, and semitendinous muscles are
spinal tracts control distal lower extremity func- reduced in amplitude or completely absent.
tion [11]. The gain is also controlled by short, During the 6 h following transection there is
intermediate, and long propriospinal systems that some recovery of reflex activity [18]. When using
control processing at multiple spinal cord levels cold as a model for acute reversible spinal cord
ipsilaterally and contralaterally. Interaction transection, hyperpolarization of caudal motor
between the cervical and lumbosacral networks neurons occurs within 30 s and monosynaptic
is mediated by propriospinal neurons [8]. The reflex amplitudes are gradually decreased. These
output from the system is through the motor neu- changes persist during the application of cold.
rons, which is measured by reflex recordings Rewarming restores reflex amplitudes to original
from muscle. Intraoperative reflex recordings values in 30 s, and the resting motor neuron
8 The Use of Reflex Responses for IOM 127

Fig. 8.1 Spinal cord central pattern Reflex and F-responses monitor integrated spinal cord activity
generator (CPG) components. The point Complex Motor Behavior
of control is the level of excitability of
segmental interneurons, which is Monitors - uncoupling of spinal cord CPGs
determined by the integrated activity of
the other components. Disturbance of
any component results in CPG
Descending
uncoupling which is detected by reflexes Peripheral afferent
Corticospinal
and F-responses from the output through segmental input
Rubrospinal
anterior horn cells
Vestibulospinal
Recticulospinal Long, intermediate
Segmental and short propriospinal
interneuronal systems
activity

Anterior horn cell

output

Disturbance of any component - uncoupling

of CPGs - changes in reflex gain - detected by Muscle

membrane potential back to the original value in Spinal Nerve Root Pathophysiology
1 min [14–16].
The hyperpolarization of motor neurons fol- Spinal nerve roots are more susceptible to injury
lowing spinal cord transection or cooling is than are peripheral nerves [24]. Spinal nerve
thought to be secondary to decreased supraseg- roots are susceptible to injury by two mecha-
mental facilitation of motor neurons, which usu- nisms. The first is that proximally the dorsal and
ally keeps them in a slightly hypopolarized state ventral roots split into rootlets and minirootlets
[14–17]. Fusimotor drive is also depressed early [25]. The area where this split occurs is the
after acute spinal cord injury, for gamma motor central-­
peripheral transitional region, which is
neurons are also hyperpolarized [19]. H-reflexes the point where the nerve root is more susceptible
and F-responses reflect the level of excitation of to mechanical injury. The axons at this point are
a large percentage of the motor neuron pool in the enclosed by a thin root sheath, cerebrospinal fluid
spinal cord gray matter [20, 21]. These record- and meninges, and lack the protective covering
ings may be very helpful in detecting intraopera- of epineurium and perineurium that is present in
tive spinal cord ischemic insults, since the level peripheral nerve [26]. The second mechanism of
of gray matter excitability is depressed more with injury is that there is an area of hypovascularity
ischemia than is posterior column function. In at the junction of the proximal and middle one-
animals the dorsal horn potential that is gener- third of the dorsal and ventral roots. This is the
ated by postsynaptic gray matter activity disap- point of anastomosis between the central vasa
pears within 3–5 min after cessation of spinal corona and peripheral segmental vessels. At this
cord perfusion. Posterior column potentials per- point, the nerve roots are more susceptible to
sist for 12–15 min [22]. mechanical injury [25].
Changes in spinal cord electrophysiologic sig- The relationship between compression and
nal processing can be used to help understand the traction forces on nerve conduction is well
electrophysiologic mechanisms occurring during described in peripheral nerves and nerve roots
acute SCI [23]. Changes in serial, parallel, and [24, 27, 28]. The conduction abnormalities are
oscillatory processing, hyperpolarization, inhibi- coupled to the perfusion of the nerve root [24, 28].
tion, and disinhibition may be observed. A 3-mm retraction distance (pressure around
128 R. Leppanen

70 g/cm2) caused the intraneural blood flow to types of mechanical force, but also to electrical
decrease to 20 % of initial value [28]. If nerve root stimulation.
dysfunction is recognized in the reversible phase,
permanent injury can be avoided. Compression
and traction may produce a mechanical effect on Late Responses
the nerve tissue in addition to compromising the
blood supply to the nerve. Physiologic block is Following electrical stimulation of a mixed sen-
the first sign of nerve dysfunction and can be sory–motor peripheral nerve, the compound
reversed in seconds. motor action potential (CMAP) or M-wave is
When more axons in the nerve are blocked, recorded from the peripherally innervated mus-
the latency of the averaged nerve action poten- cle. The M-wave is the result of orthodromic
tials may increase and the amplitude may motor conduction from the point of stimulation
decrease. A more intense or prolonged compres- to the muscle. In addition to the short latency
sion or traction may produce myelin deformation M-wave, three late responses can be recorded:
at the edges of compression and prolongation of H-reflex, F-response, and A-wave (axon reflex).
the nerve action potential latencies [28]. Further The mechanisms responsible for generating
mechanical loading may produce a conduction F-responses and A-waves are different from
block from segmental demyelination with reduc- H-reflexes [32]. When recording intraoperative
tions in action potential amplitudes. With injury, H-reflexes, the F-response and A-wave should
reversible functional changes precede more not be confused with H-reflexes.
severe morphologic nerve changes such as neura- The F-response is not a reflex. After supra-
praxia (segmental demyelination) and axonotme- maximal stimulation of a mixed peripheral
sis (Wallerian degeneration). Although both these nerve, antidromic motor nerve impulses are con-
morphologic changes may be reversible in weeks ducted proximally to the ventral horn where they
to months, perineural and intraneural scarring activate from 1 to 5 % of the motor neurons. This
may result and propagate further nerve damage is followed by orthodromic conduction through
[26, 27, 29, 30]. motor fibers, and the F-response is recorded
The mechanosensitivity of normal and abnor- from muscle. The F-response is of low ampli-
mal dorsal root ganglia and axons has been tude, and the latency, configuration, and duration
defined in human and animal models [24]. In a vary from one stimulus to the next. This variabil-
cat model, rapidly applied mechanical forces ity occurs because each time the motor neuron
induce short-duration (200 ms) impulses in nor- pool is activated, a different population of 1–5 %
mal nerve root. Static mechanical forces do not of the motor neuron pool is activated that
induce impulses in normal nerve root. Rapidly have different conduction characteristics [33].
and statically applied mechanical forces induce Both afferent and efferent components of the
long periods (15–30 s) of repetitive impulses in F-response follow the same motor neurons. The
irritated or in regenerating nerves in continuity. subpopulation of the motor neuron pool acti-
Minimal acute compression of normal dorsal root vated by the F-response and H-reflex are not the
ganglion induces prolonged (5–25 min) repeti- same [34]. F-response persistence is a measure
tive firing of nerve [31]. Mechanically induced of the excitability of the motor neuron pool.
trains of nerve action potentials can desynchro- Persistence is the number of recorded F-responses
nize afferent 1a nerve action potentials during divided by the number of stimuli. Clinically, a
H-reflex recordings and decrease the H-reflex persistence less than 50 % is considered to be
amplitude. When interpreting intraoperative abnormal [35] (Fig. 8.2).
motor nerve root studies, it is important to under- The A-wave is a late motor response that is
stand the pathophysiologic mechanisms of nerve present with constant latency, configuration, and
root injury and to understand the response of nor- amplitude. Low-intensity stimulation elicits the
mal and pathologic nerve to not only different A-wave, and it is usually blocked by higher inten-
8 The Use of Reflex Responses for IOM 129

10 mV F−SNS: 1 mV 10 ms

M - waves
80 . 0mA

80 . 0mA

80 . 0mA

80 . 0mA

80 . 0mA F - responses
F - responses
80 . 0mA

80 . 0mA

80 . 0mA

80 . 0mA

80 . 0mA

10 mV F−SNS: 500 uV 10 ms
80 . 0mA

80 . 0mA

80 . 0mA

80 . 0mA

80 . 0mA

80 . 0mA

b
a

Fig. 8.2 F-responses are generated by antidromic motor change with each stimulus. Sequentially, recorded abduc-
impulses that activate 1–5 % of the motor neuron pool. tor hallucis F-responses (a) following tibial nerve stimula-
The amplitude, latency, duration, and configuration tion at the ankle that are superimposed in (b)

sity of stimulation. A-wave latency is between the motor nerve. The collateral sprout innervates mus-
CMAP and the F-response latency or exceeds the cle. When the antidromic impulse reaches the
F-response latency. It may also appear with a point of damage, a portion of the electrical impulse
latency between the M-wave and H-reflex latency proceeds distally along the collateral sprout, and a
or the latency may exceed the H-reflex latency. small portion of the muscle is activated. Depending
A-wave amplitude is less than the H-reflex ampli- on the nerve stimulated, A-waves may be normal
tude. The A-wave should not be confused with the or abnormal (Fig. 8.3) [32].
H-reflex or F-response. The A-wave is generated
by peripheral nerve changes rather than changes in
the central nervous system signal processing. The H-Reflexes: Monosynaptic,
physiology of the A-wave is that there is periph- Oligosynaptic
eral neural damage with the presence of a collat-
eral sprout from a proximal point of damaged See Chap. 7 for a discussion of H-reflex testing.
130 R. Leppanen

Fig. 8.3 A-wave amplitude, duration, latency, and configuration do not change from one stimulus to the next, but
F-responses do. These are ten abductor hallucis recordings following tibial nerve stimulation at the ankle

Neurophysiologic Basis of H-Reflexes reflects the refinement of motor neuron pool acti-
vation with central nervous system maturation. In
The H-reflex is a CMAP recorded from muscle adults, they are also frequently found in the quad-
after electrical afferent activation of a monosyn- riceps and plantar foot muscles [36–38].
aptic reflex. The afferent pathway involves elec- The H-reflex was first described by Hoffman
trical activation of the large 1a nerve fibers in 1918 [39] and characterized more in the 1950s
originating from muscle. After entering the dorsal [40]. The reflex is most easily recorded from the
horn of the spinal cord, the 1a fibers synapse with gastrocnemius muscle after stimulation of the
the motor neurons. The efferent pathway involves tibial nerve in the popliteal fossa. The fast-­
orthodromic motor conduction through motor conducting low-threshold 1a fibers are activated
fibers in the same homologous spinal segment as with long-duration (1 ms) low-intensity stimula-
the afferent pathway [32]. In normal newborns, tion. The stimulation rate is 0.5 Hz. The intensity
H-reflexes may be recorded from many widely of stimulation is slowly increased. Low-intensity
distributed muscles. After 2 years of age, they are stimulation activates the 1a fibers before the
primarily present in the gastrocnemius, soleus, motor fibers, so at low intensity of stimulation,
and flexor carpi radialis muscles. The more the H-reflex appears before the M-wave. The
restricted distribution of H-reflexes in adults H-reflex CMAP is usually of a biphasic or tripha-
8 The Use of Reflex Responses for IOM 131

5 mV 10 ms

M - waves
H - reflexes
M - wave, H - reflex and F - response
amplitude changes with increased
9 . 4mA
stimulation intensity
11 . 0mA

14 . 9mA

25 . 1mA

24 . 3mA
5 mV 10 ms
25 . 9mA

28 . 2mA M - waves

30 . 6mA

35 . 3mA

35 . 3mA

F - responses
H - reflexes and F - responses

Fig. 8.4 The H-reflex is the first component elicited from the gastrocnemius muscle with increasing stimulus intensity
of the tibial nerve in the popliteal fossa. The second is the M-wave and the third is the F-response

sic configuration. The 1a fibers are activated first The H-reflex amplitude usually peaks at or just
either because they have a lower threshold than before the M wave becomes present. Further
motor fibers to long-duration stimulation or increases in stimulation intensity result in a steady
because they are located anatomically more increase in the M-wave amplitude. When the M
superficial than the motor fibers in the popliteal wave no longer increases in amplitude, the H-reflex
fossa [41]. is usually replaced by the F-response (Fig. 8.4).
As the intensity of stimulation is increased, a When the stimulus intensity is held constant from
greater percentage of the motor neuron pool is one stimulus to the next, the H-reflex is of short
activated and the H-reflex amplitude increases latency, short duration, simple configuration, and
[32]. In an awake human, when recording the gas- constant amplitude [32]. To determine if a CMAP
trocnemius H-reflex, the percentage of the motor is a H-reflex, the amplitude should exceed the
neuron pool activated averaged 50 % (range, 24.0– M-wave amplitude, and the configuration and
100 %) [20]. The motor neurons recruited with an latency should be the same from one stimulus to
increasing intensity of stimulation obey the the next [32]. Changes in the effect of the central
Henneman size principle. Low-­force motor neu- facilitation and inhibition on the spinal interneu-
rons are recruited with low intensity of stimula- rons and motor neurons may change the pattern of
tion, and high-force motor neurons are recruited how the H-reflex is activated. With an increased
with higher intensity of stimulation [42]. presynaptic inhibition and hyperpolarization of
132 R. Leppanen

T
R
C
Vastus Medialis
Trig Trig
10 ms 1 10 ms
/
200 uV 5 200 uV
Amp 1 Amp 5

Tibialis anterior

Trig Trig
2
10 ms 10 ms
/
200 uV 6 200 uV
Amp 2 Amp 6
M - wave
H - reflex Gastrocnemius

#
Trig Trig
3
10 ms 10 ms
/
2 mV 7 200 uV
Amp 3 Amp 7

Abductor Hallucis

Trig Trig
10 ms 4 10 ms
/
200 uV 8 200 uV
Amp 4 Amp 8
Left Right

Fig. 8.5 Homonymous monosynaptic left intraoperative gastrocnemius H-reflex recorded following electrical stimula-
tion of the left tibial nerve in the popliteal fossa

motor neurons, the M-wave may be present before In the lower extremity of man, there are two
the H-reflex, and the H-reflex amplitude may be types of prewired monosynaptic 1a H-reflex con-
smaller than the M-wave amplitude. With a nections: homonymous (homosynaptic) and het-
decreased presynaptic inhibition, the H-reflex may eronymous (heterosynaptic). They are both part
be robust, and it may not be possible to suppress of the functional synergistic spinal cord CPGs.
and replace it with the F-response. Both types may be recorded in the operating
The H-reflex has been thought of as a mono- room. Homonymous monosynaptic H-reflexes
synaptic reflex. The central conduction time are H-reflexes that are recorded from muscles
between dorsal and ventral roots reveals only that are innervated by the same nerve root as the
enough time for one synapse, that is, between 0.5 1a-activated sensory fibers. The gastrocnemius
and 1.0 ms [32]. There is also evidence to indicate H-reflex is an example of a homonymous mono-
that the H-reflex is an oligosynaptic reflex. Low- synaptic H-reflex (Fig. 8.5). 1a sensory action
threshold motor neurons may be activated by a potentials may also make monosynaptic connec-
single (monosynaptic) synapse through the fastest tions with motor neurons at spinal cord levels
1a afferents. Higher threshold motor neurons may other than the 1a sensory segmental level. As a
be activated through several (oligosynaptic) syn- result of this activation, H-reflexes may be
apses by the fastest 1a afferents and through sin- recorded from muscles having segmental
gle synapses by slower 1a afferents [32]. innervation other than the 1a segmental afferent
8 The Use of Reflex Responses for IOM 133

Fig. 8.6 Following Heterosynaptic la Facilitation

intraoperative low-intensity
electrical stimulation in the H - Reflexes
right popliteal fossa, decreased
T
spinal cord presynaptic R
inhibition may allow not only C
the homonymous H-reflex to H-reflex
be recorded from the Trig
1 10 ms
gastrocnemius muscle but also Vastus medialis /
heteronymous H-reflexes may 5 100 uV
Amp 5
be recorded from the vastus
medialis, tibialis anterior, and
abductor halluces muscles H-reflex

Trig
2 10 ms
Tibialis anterior /
6 100 uV
Amp 6

H-reflex
Trig
3 10 ms
Gastrocnemius /
500 uV
7
Amp 7

H-reflex
Trig
4 10 ms
Abductor hallucis /
8 100 uV
Amp 8

Right

activation. These H-reflexes are called heterony- tibialis anterior and intrinsic hand muscles, may
mous monosynaptic H-reflexes (Fig. 8.6). indicate a disordered central motor system state.
In humans, heteronymous connections nor- These changes occur because of uncoupling of the
mally exist between ankle and knee muscles. different components of CPGs [44–47].
Functionally heteronymous connections provide
coupling between muscles operating at different
joints. These transjoint monosynaptic connections Gastrocnemius H-Reflex
are important for maintaining equilibrium during
bipedal stance and during gait. Intraoperatively,  astrocnemius H-Reflex Normal
G
heteronymous H-reflexes are inhibited by presyn- Parameters
aptic inhibition. When presynaptic inhibition is In the lower extremity, the H-reflex can be
decreased, heteronymous H-reflexes may become recorded from the gastrocnemius and soleus
present [43]. muscle after electrical stimulation of the poste-
The presence of H-reflexes in muscles where rior tibial nerve in the popliteal fossa [48]. This
they are not usually recorded can be an indication reflex is mediated by segmental S1 afferent and
of a lesion of the suprasegmental central nervous efferent activity [45]. Intraoperative normal
system that results in a decreased effect of presyn- parameters have not been established. Normal
aptic inhibition on motor neurons. The abnormal parameters established for clinical studies may
distribution of H-reflexes in adults, such as in the serve as a guide for intraoperative studies.
134 R. Leppanen

Clinically, the gastrocnemius H-reflex latency For subdermal electroencephalographic

varies with age and leg length and has a mean (EEG) recording, needle electrodes are inserted
latency of 28.9 ± 2.7 ms in an awake human. A in the medial head of the gastrocnemius muscle.
regression equation may be used to calculate the The H-reflex may also be recorded in the calf
expected latency for each individual: H-reflex (in from the soleus muscle. The technique for record-
ms) = 9.14 + 0.46 (leg length in centimeters) + 0.1 ing the H-reflex from the soleus muscle is the
(age in years). A normogram based on this equa- same as that for recording from the gastrocne-
tion is available as a reference [49]. Clinically, mius muscle, only that the recording electrodes
the normal side-to-side amplitude difference are placed over the mid-dorsal line of the leg with
between 21 and 67 years of age may reach 60 % the active electrode 4 cm above the point where
[50]. The upper limit of normal clinical side-to-­ the two heads of the gastrocnemius muscle join
side latency difference is 1.5 ms [37]. When mea- the Achilles tendon. The reference electrode is
suring latency, the most concise departure from placed 3 cm distal to the active electrode [48].
the baseline occurs when the active recording Monopolar electromyographic (EMG) needle
electrode is over the motor point. It is therefore electrodes and longer uncoated stainless steel
necessary to know the location of the motor point needle electrodes may also be used.
of the gastrocnemius muscle [51]. The H:M ratio Fine Teflon-coated silver wires with the wire
is a measure of H-reflex motor neuron pool acti- exposed at the end and that are inserted with a
vation or excitation. It is calculated by dividing spinal tap needle may also be used for recordings
the maximum H-reflex amplitude by the maxi- when the subcutaneous tissue is thick. The active
mum M-wave amplitude. It is normally less than electrode is inserted at the motor point of the gas-
0.7 [37]. Intraoperatively, onset latencies may be trocnemius muscle and the reference electrode is
greater because of the decreased limb tempera- inserted over tendon or bone. The needles are
ture. In the operating room, H-reflex and M-wave secured with tape. A range of different high- and
amplitudes that are recorded in the clinical set- low-frequency filters are used. A high-frequency
ting may be reduced by the use of neuromuscular filter of 10 KHz and a low-frequency filter of
junction (NMJ) blocking agents. Intraoperatively, 20 Hz are most often used [32]. The time base is
the gastrocnemius H-reflex parameters that have 100 ms. Recordings are single sweep.
been monitored are H-reflex amplitude, latency, In addition to recording H-reflexes from the
and the H:M ratio. Right–left amplitude and gastrocnemius muscle, recordings may also be
latency differences are also used. made bilaterally and simultaneously from the
vastus medialis, tibialis anterior, and abductor
 astrocnemius H-Reflex Stimulation
G hallucis muscles. Recording from these muscles
and Recording Techniques will allow for the detection of heteronymous
Stimulation is with needle or surface electrodes. H-reflexes. This also allows for the monitoring of
The cathode is placed proximally in the popliteal proximal vestibulospinal and reticulospinal con-
fossa between the tendons of the medial and lateral trolled motor neurons and distal rubrospinal and
hamstring muscles. The anode is placed 2–4 cm corticospinal controlled motor neurons [11]. The
distal to the cathode. The stimulation rate is 0.5 Hz gastrocnemius H-reflex may be used to monitor
and the stimulus duration is 1.0 ms. The stimulus peripheral tibial nerve, proximal sciatic nerve,
intensity is adjusted so that the H-reflex amplitude sensory and motor S1 nerve root, and S1 segmen-
is maximal. The most effective stimulus intensity is tal spinal cord function. These reflexes can also
chosen such that any increase or decrease in stimu- be used to monitor the function of a variety of
lus intensity results in a decrease in the H-reflex suprasegmental descending spinal cord systems
amplitude. Baseline recordings are made with the that control the S1 segmental interneurons [52].
patient anesthetized before the start of the surgical When present, heterosynaptic H-reflexes may be
procedure. Any variability in latency and ampli- used to monitor other nerve roots. The ability to
tude should be noted in the baseline recordings. record lower extremity H-reflexes may be
8 The Use of Reflex Responses for IOM 135

affected by pre-existing abnormalities, such as a and the H:M ratio. Right–left latency and ampli-
generalized polyneuropathy, plexopathy, or tude differences are also used.
radiculopathy. H-reflexes may be absent, laten-
cies may be prolonged, amplitudes may be  lexor Carpi Radialis H-Reflex
F
decreased, and the CMAP configuration may Stimulation and Recording Techniques
change. Amplitudes may also be decreased with Stimulation is with needle electrodes spaced
the presence of a myopathy [1, 2]. 2 cm apart unilaterally over the distal medial
upper arm or over the anterior medial elbow at
0.5 Hz and 1.0 ms duration. The cathode is proxi-
Flexor Carpi Radialis H-Reflex mal. The stimulus intensity is adjusted so that the
H-reflex amplitude is maximal. The most effec-
 lexor Carpi Radialis H-Reflex
F tive stimulus intensity is chosen such that any
Background and Normal Parameters increase or decrease in stimulus intensity results
In the upper extremity, the flexor carpi radialis in a decrease in the H-reflex amplitude. Baseline
H-reflex can be recorded after electrical stimula- recordings are made with the patient anesthetized
tion of the median nerve over the distal medial before the start of the surgical procedure. In the
upper arm or over the anterior medial elbow. This operating room, H-reflex amplitudes that are
reflex is mediated by segmental C6/C7 afferent recorded in the clinical setting may be reduced by
and efferent activity (Fig. 8.7). Intraoperative nor- the use of NMJ-blocking agents. Any variability
mal parameters have not been established. Normal in latency and amplitude should be noted in the
parameters established for clinical studies may baseline recordings.
serve as a guide for intraoperative studies. For recording subdermal EEG, needle elec-
Clinically, the flexor carpi radialis H-reflex latency trodes are inserted in the flexor carpi radialis
varies with the arm length. When measuring muscles. Monopolar EMG needle electrodes and
latency, the most concise departure from the base- longer uncoated stainless steel needle electrodes
line occurs when the active recording electrode is may also be used. Fine Teflon-coated silver wires
over the motor point. It is therefore necessary to with the wire exposed at the end that are inserted
know the location of the motor point of the flexor with a spinal tap needle may also be used for
carpi radialis muscle [51]. In awake humans, the recordings when the subcutaneous tissue is thick.
mean latency is 17.07 ± 1.77 ms. The interlatency The active electrode is inserted at the motor point
time is calculated by subtracting the M wave from and the reference electrode is inserted distally
the H-reflex latency. The interlatency time mean over tendon or bone. The needles are secured
latency is 14.5 ± 1.8 ms. The maximum side-to- with tape. A range of different high- and low-­
side H-reflex latency difference is 0.002 ± 0.42 ms. frequency filters are used. A high-frequency filter
The maximum side-to-­ side interlatency time of 10 KHz and a low frequency filter of 20 Hz are
latency is 0.11 ± 0.44 ms. A regression equation is most often used. Flexor carpi radialis H-reflexes
used to calculate the expected H-reflex latency: may be used to monitor median peripheral nerve,
H-reflex (in ms) = 0.29 ± 0.195 × arm-length in cen- brachial plexus, and segmental sensory and
timeters. The equation for the interlatency time is motor spinal nerve root and spinal cord function.
−2.08 + 0.1878 × arm-length in centimeters. A nor- These reflexes can also be used to monitor the
mogram based on these equations is available as a function of a variety of suprasegmental descend-
reference. The arm length is measured from the tip ing spinal cord systems that control the C6/C7
of the third finger to the C6 spinous process with segmental interneurons. The ability to record
the arm pronated and the shoulder abducted to 90° flexor carpi radialis H-reflexes may be affected
[53, 54]. Intraoperatively, onset latencies may be by pre-­existing abnormalities, such as a general-
greater because of the decreased limb temperature. ized polyneuropathy, plexopathy, or radiculopa-
Flexor carpi radialis H-reflex parameters that have thy. H-reflexes may be absent, latencies may be
been monitored are H-reflex amplitude, latency, prolonged, amplitudes may be decreased, and the
136 R. Leppanen

Fig. 8.7 Baseline intraoperative right flexor carpi radialis flexor carpi radialis muscles. Baseline free-run EMG
M-wave and H-reflex recordings. Free-run EMG activity activity is present in the left deltoid muscle secondary to
is also recorded bilaterally from the trapezius, deltoid, and left C5 and/or C6 nerve root irritation

CMAP configuration may change. Amplitudes technique used must allow for the recording of all
may also be decreased with the presence of a these signals with maximum sensitivity (see
myopathy process [1, 2]. Chap. 19 for a discussion of general anesthesia
monitoring) [2, 55–60].
The technique used varies depending on the
Anesthetic Technique surgical procedure, patient’s age, medical history,
and the presence of a pre-existing neurologic
To record reflexes intraoperatively, it is critical deficit. Frequently, total intravenous anesthesia
that the anesthetic technique does not signifi- (TIVA) with constant infusion of propofol and
cantly inhibit the activity of suprasegmental spi- fentanyl and other opioids are used. Bolus injec-
nal cord function, spinal interneurons, and tions of these agents should be avoided to prevent
segmental motor neurons. Also, adequate neuro- suppression of recordings and if nitrous oxide is
muscular junction (NMJ) transmission must be used it should not exceed 50 vol.% [61]. Ketamine
present. Reflexes and F-responses are usually may serve an adjunctive role to reduce propofol
monitored during the recording of SSEPs, utilization. High dose of ketamine may suppress
tcMEPs, EEG, and free-run EMG. The anesthetic tcMEP amplitudes [58]. Ketamine is an excitatory
8 The Use of Reflex Responses for IOM 137

agent and increases SSEP, tcMEP, and H-reflex dose of 2 mg/kg decreased the H-reflex amplitude
amplitudes [62]. and H:M ratio. The following 10-min infusion did
Dexmedetomidine can be used to supplement not further decrease these values. The 6-μg/mL
propofol during TIVA. Suggested therapeutic injection did not change the H-reflex amplitude
levels (0.5–0.7 μg/kg/h) suppress MEP ampli- and H:M ratio. The 9-μg/mL injection decreased
tudes but not SSEP amplitudes [63, 64]. One the amplitude and H:M ratio. They recommended
study found that 0.6 μg/kg/h did not suppress that the propofol induction dose be 1.0–2.5 mg/kg
SSEP, tcMEPs, and visual evoked potentials [65]. and this should be followed by an infusion from
Subtherapeutic doses of less than 0.5 μg/kg/h 100 to 200 μg/kg/min. They concluded that the
combined with propofol do not suppress tcMEPs immobility during propofol anesthesia is not
[66, 67]. caused by a depression of spinal motor neuron cir-
The author studied the effect of dexmedetomi- cuit excitability. Propofol does not decrease axon
dine on intraoperative SSEPs, tcMEPs, EEG, conduction peripherally nor transmission at the
H-reflexes, and F-responses in 17 patients during neuromuscular junction [69].
correction of scoliosis. After completion of dis- Soleus H-reflexes were used to determine the
traction and fixation, a loading dose of 1 μg/kg of level of motor neuron excitability intraopera-
dexmedetomidine was added to the fentanyl and tively. In ten normal human volunteers, 1.0–1.5 %
propofol technique and this was followed by an enflurane was found to decrease H-reflex ampli-
infusion of 0.5 μg/kg/h to the end of surgery. No tudes from 35 to 100 % of baseline values [70].
muscle relaxants were used. The addition of dex- The effect of isoflurane alone and isoflurane
medetomidine had no effect on SSEPs or EEG with N20 on the soleus H-reflex was studied
but the effect on motor signals was variable. under general anesthesia in 25 patients. No NJM-­
Lower extremity tcMEP amplitudes decreased blocking agents were used. Twenty-three of these
from 0 to 100 % (mean, 86 ± 16.6 %), F-response patients had consistently measurable stable
amplitudes decreased from 0 to 100 % (mean, H-reflexes with baseline amplitudes varying
58 ± 12.5 %), and H-reflex amplitudes decreased from 3.43 to 11.97 mV. The addition of 0.68 %
from 0 to 100 % (mean, 71 ± 13.9 %). Because of isoflurane decreased the amplitude to 48 % of the
the variable and at times pronounced effects on baseline. Increasing the isoflurane to 1.37 %
motor signals, it was recommended that dexme- decreased the amplitude to 33.8 % of baseline.
detomidine not be used to monitor tcMEPs, The combination of 0.81 % isoflurane with 30 %
H-reflexes, or F-responses [68]. N20 decreased the amplitude to 66.2 % of base-
NMJ-blocking agents should be avoided other line. The combination of 0.37 % isoflurane with
than using short-acting agents for intubation. 70 % N20 decreased the amplitude to 30.4 % of
NMJ function needs to be carefully monitored. the baseline. They reported that increasing isoflu-
The most common technique is the train-of-four rane concentration results in decreased H-reflex
(TOF) response. Another technique is the T1 % amplitude. Increasing the N20 concentration
response. T1 % = first unblocked response/con- results in decreased H-reflex amplitude. They
trol response × 100. NMJ monitoring should be concluded that the H-reflexes may be recorded
conducted on muscles that may be affected by the with a combination of isoflurane and N20. The
surgical procedure. H-reflex is maximally stable within the range of
The effects of propofol on the soleus H-reflex anesthetic concentrations used to achieve surgi-
were studied in 33 patients. No NMJ-blocking cal immobility [55].
agents were used. H-reflexes were recorded before The effect of isoflurane and N20 on spinal
administration of anesthetics and after an initial motor neuron excitability was studied in eight
dose of 2 mg/kg over 1 min followed by a continu- adult patients by monitoring soleus H-reflexes and
ous infusion at a rate of 167 μg/kg/min for 10 min. abductor hallucis F-responses. No NMJ-­blocking
Measurements were also made after infusing pro- agents were used. H-reflex amplitude was
pofol to a blood level of 6 and 9 μg/mL. The initial decreased to 48.4 ± 18.6 % of the baseline with 0.6
138 R. Leppanen

minimum alveolar concentration (MAC) isoflu- sevoflurane (0.8 vol.%). Soleus H-reflexes were
rane and to 33.8 ± 19.1 % with 1.2 MAC isoflu- used to study the depressive effects of these
rane. MAC is defined as the alveolar concentration agents individually. The amount of depression
of an anesthetic agent that prevents movement in was found to be dependent on the size of the
50 % of patients in response to a painful stimulus. H-reflex and is different at different stimulus
F-response amplitude and persistence decreased to intensities, indicating a varying effect of propofol
52.2 ± 22.8 % and 44.4 ± 26.0 % of the baseline at and sevoflurane on motor neurons of different
0.6 MAC isoflurane and to 33.8 ± 26.0 % and size. H-reflex depression is more for both agents
21.7 ± 22.8 % at 1.2 MAC. With 1.0 MAC isoflu- at a lower intensity of stimulation than at a higher
rane, the H-reflex amplitude was decreased by intensity of stimulation, indicating that smaller
32.5 ± 19.2 %, 33.3 ± 20.8 %, and 30.4 ± 23.5 % of motor neurons are depressed more than larger
baseline levels at 30, 50 and 70 % nitrous oxide, motor neurons. In contrast, excitability of motor
respectively [71]. neurons by supraspinal effects affects the larger
In 12 adult patients, a comparison of the before smaller motor neurons [74]. From a prac-
effects of isoflurane on tcMEPs and F-responses tical standpoint, if H-reflexes are being recorded
were studied. Anesthesia was maintained with with a low stimulus intensity and injury occurs
60 % N20, 100 μg/kg/min of propofol and supple- that effects the larger motor neurons, this injury
mentary fentanyl, 0.5–1.0 μg/kg. Recordings could be missed because only the function of the
were made before and after adding 0.5 % isoflu- smaller motor neurons would be monitored. The
rane. Baseline tcMEP amplitudes (median, smaller motor neuron population would already
205 μV, 25th–75th percentiles, 120–338 μV), be suppressed by the anesthesia, further decreas-
F-response amplitudes (median, 100 μV, 25th– ing the sensitivity for detecting injury. A higher
75th percentiles, 64.2–137.5 μV) and F-response intensity of stimulation would activate small and
persistence (59 ± 29 %) were decreased to 0.0 μV large motor neurons, allowing the monitoring of
(0–15 μV), 49 μV (12.4–99.6 μV), and 30 ± 31 %, a greater percentage of the motor neuron pool.
respectively, by 0.5 % isoflurane. The tcMEPs
were suppressed more than the F-responses. No
NMJ-blocking agents were used [72].  linical Correlation of H-Reflexes
C
Soleus H-reflexes and abductor hallucis and F-Responses
F-responses were monitored to determine the effect
of hyperventilation and hypoventilation on motor Studies of the normal and abnormal electrophysi-
neuron excitability during isoflurane anesthesia. ology of the spinal cord CPG components in
H-reflex and F-responses were recorded before and humans and animals have resulted in the
after changing the end-tidal CO2 (ETCO2) concen- understanding of the mechanisms involved in
­
tration. Anesthesia was maintained with 0.8 % iso- acute complete and partial SCI in humans. How
flurane and no muscle relaxants were used. An H-reflex and F-response changes correlate with
ETCO2 of 25 mmHg decreased the preanesthetic the patient’s postoperative status is derived from
H-reflex amplitude from 6.8 ± 2.7 mV to the publications discussed as follows.
4.0 ± 2.0 mV and to 2.0 ± 2.2 mV at an ETCO2 of Soleus H-reflexes and abductor hallucis
45 mmHg. F-response persistence decreased from F-responses were recorded in 32 patients during
the preanesthetic value of 100 % to 77 % ± 24 % at spinal cord surgery. In six patients, an abrupt fall
an ETCO2 of 25 mmHg and to 61 ± 19 % at an in H-reflex amplitude beyond 3 SD from the
ETCO2 of 45 mmHg. They concluded that hyper- baseline or significant drop in F-response persis-
ventilation and hypoventilation effects motor neu- tence coincided to perturbation or injury to the
ron excitability and may affect the probability of spinal cord. Transient suppression occurred in
patient movement during surgery [73]. four patients with less than a 50 % drop in
Volunteers were sedated with a constant level H-reflex amplitude or abductor hallucis
of propofol (2 mg/L) and to a constant level of F-response persistence. These patients did not
8 The Use of Reflex Responses for IOM 139

develop a new postoperative neurologic deficit. H-reflexes may indicate intact spinal cord
Suppression exceeded 90 % of the baseline val- function with changes in SSEPs. Soleus
ues and persisted through surgery in two patients. H-reflexes and tibial SSEPs were monitored in a
Both patients had profound postoperative neuro- patient during T7–T12 laminectomy for spinal
logic deficits. The author concluded that rostral stenosis. During laminectomy, the left SSEP
SCI suppresses H-reflexes and F-responses and became absent and the right amplitude was sig-
the degree of suppression reflects the severity of nificantly transiently reduced. No H-reflex
injury. The mechanism responsible for these changes occurred. Postoperatively, no lower
changes is thought to be hyperpolarization of extremity motor deficits were present and no new
caudal motor neurons, which occurs within sec- sensory deficits [76].
onds of injury [75]. Spinal cord function was monitored in 278
H-reflexes were recorded in the lower extrem- pediatric spine with gastrocnemius H-reflexes
ities in 31 patients during spinal surgery. A sig- and SSEPs. Combined H-reflex and SSEP moni-
nificant change in amplitude was considered toring improved the reliability for detecting spi-
significant if it exceeded 3 SD of the mean post- nal cord compromise compared with either
anesthetic baseline. In six patients, a significant procedure alone. H-reflexes changed more than
decrease in H-reflex amplitude occurred. The SSEPs. The changes reflected changes in spinal
onset of H-reflex suppression coincided with a cord gray matter function related to acidosis and
potentially injurious event in each case. In one changes in hematocrit and blood pressure [77].
case involving a cervical myelotomy, a large syr- In a clinical setting, soleus H-reflexes and
inx collapsed during decompression, producing abductor halluces F-responses were recorded in
immediate fluctuation in the H-reflex amplitude. 14 patients following SCI that resulted in either
The amplitude recovered to the baseline level partial injury without spinal shock or injury with
9 min later. Postoperatively, no neurologic deficit spinal shock. Deep tendon reflexes following tap
was noted. Another case involved mechanical of the Achilles tendon and patellar tendons were
reduction of a T-8 spinal fracture. Increased vari- also evaluated. Patients were evaluated within
ability in H-reflexes developed with manipula- 24 h of injury and on day 10, 20, and 30 after
tion of the spine before reduction. The first injury. F-responses were absent in patients with
attempt to reduce the fracture produced a tran- spinal shock, reduced in persistence in patients
sient fall in H-reflex amplitude. The second with acute injury without spinal shock, and nor-
attempt at reduction produced a pronounced mal in persistence in patients with chronic injury.
reduction in amplitude to less than 10 % of base- F-response changes persisted up to 2 weeks fol-
line. H-reflexes remained suppressed until the lowing SCI. H-reflexes were absent or markedly
end of surgery and postoperatively the patient suppressed in patients with spinal shock within
had severe motor and sensory deficits in both legs 24 h of injury but recovered to normal amplitude
that were not present preoperatively. A third case within several days of the injury. Deep tendon
involved a cervical myelotomy for decompres- reflexes were proportionally more depressed in
sion of a posttraumatic syrinx. spinal shock than H-reflexes. This demonstrated
During surgery, cervical cord hemorrhage dissociation between electrically and mechani-
occurred and was followed by reduction H-reflex cally induced reflexes during spinal shock. The
amplitude. The amplitude steadily declined and observation that the stretch reflex is more
the H-reflexes disappeared and remained absent. depressed than the H-reflex is consistent with
Postoperatively, the patient had profound weak- depressed fusimotor drive with SCI [19].
ness in both lower extremities. The changes in A 63-year-old woman with a large T8–9 her-
these patients demonstrated that H-reflex changes niated disk that was causing spinal stenosis with
occur immediately at the time of spinal injury. cord compression was treated with a bilateral
H-reflex changes may be reversible and reflect T8–9 laminectomy with left far lateral costover-
the severity of SCI [4]. tebral exposure for disk removal. She presented
140 R. Leppanen

clinically with right T8 de