Helge Refsum Ilmar A.

Sulg
Knut Rasmussen (Eds.)

HEART ~~ BRAIN
BRAIN ~ HEART

With 159 Figures, Some in Color

Springer-Verlag Berlin Heidelberg GmbH

Prof. Dr. med. HELGE REFSUM
University of Tromsl/l
Department of Physiology
Institute of Medical Biology
Tromsl/l, Norway
Prof. Dr. med. ILMAR A. SULG
University of Tromsl/l
Department of Neurology
Institute of Clinical Medicine
TromSI/l, Norway
Prof. Dr. med. KNUT RASMUSSEN
University of Tromsl/l
Section of Cardiology
Department of Internal Medicine
Tromsl/l, Norway

ISBN 978-3-642-83458-5 ISBN 978-3-642-83456-1 (eBook)

DOI 10.1007/978-3-642-83456-1

Library of Congress Cataloging-in-Publication Data. Heart & brain, brain & heart / Helge Refsum,
IImar A. Sulg, Knut Rasmussen (eds .). p. cm. Based on a symposium held at the University of
Tromsl'l, Norway, June 24-27,1987. Includes bibliographies and index. ISBN 0-387-19186-0 (U.S.):
1. Heart - Diseases - Complications and sequelae - Congresses. 2. Brain - Diseases - Complica-
tions and sequelae - Congresses. 3. Heart - Physiology - Congresses. 4. Brain - Physiology - Con-
gresses. 5. Heart - Innervation - Congresses. 6. Brain - Blood-vessels - Congresses. I. Refsum,
Helge. II. Sulg, IImar A., 1919- . III. Rasmussen, Knut. IV. Title: Heart & brain, brain & heart.
[DNLM: 1. Autonomic Nervous System Diseases - congresses. 2. Cardiovascular Diseases - con-
gresses. 3. Central Nervous System Diseases - congresses. WG 100 H4345 1987] RC682.H35
1989 616.1'2 - dc19 DNLMIDLC for Library of Congress 88-39684 CIP

This work is subject to copyright. All rights are reserved, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broad-
casting, reproduction on microfilms or in other ways, and storage in data banks. Duplication of this
publication or parts thereof is only permitted under the provisions of the German Copyright Law of
September 9,1965, in its version of June 24, 1985, and a copyright fee must always be paid. Violations
fall under the prosecution act of the German Copyright Law.

© Springer-Verlag Berlin Heidelberg 1989

Originally published by Springer-Verlag Berlin Heidelberg New York in 1989
Softcover reprint of the hardcover 1st edition 1989

The use of registered names, trademarks, 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.
Product Liability: The publisher can give no guarantee for information about drug dosage and applica-
tion thereof contained in this book. In every individual case the respective user must check its ac-
curacy by consulting other pharmaceutical literature .
Typesetting and printing: Petersche Druckerei GmbH & Co. Offset KG, Rothenburg ob der Tauber
Bookbinding: Konrad Triltsch, Graphischer Betrieb, D-8700 Wiirzburg
2121/3130-543210 - Printed on acid-free paper
Preface

The aim of this book is to focus on the important relationships between the
heart and the brain in health and disease. The brain and nervous system
may cause or influence heart disease, e.g., by causing arrhythmias or modi-
fying the response to ischemia. Disorders of the heart and circulation may
cause brain damage, e.g., by releasing emboli resulting in cerebral infarc-
tion. Furthermore, both the brain and the heart are frequently targets of
the same disease process. The heart and brain have electrophysiologically
active cells, which may respond in similar ways to diseases and various
interventions. Finally, many drugs affect both the brain and the heart, and
drugs used for heart diseases usually have side effects on the brain and vice
versa.
With today's increasing subspecialization in medicine, we feel the time
has come to present a book that integrates basic and clinical aspects of
cardiology, neurology, cardiovascular surgery, and neurosurgery. We hope
this cross fertilization will broaden horizons and advance both understand-
ing and practice.
This book is based on a symposium held at the University of TromS!/l,
Norway, 24-27 June 1987, organized by Ilmar A. Sulg, Knut Rasmussen,
Svein Ivar Mellgren, Dag S!/lrlie, and Helge Refsum of the Departments of
Clinical Neurophysiology, Medicine, Neurology, Surgery, and Physiology,
respectively. Weare grateful to the distinguished group of contributors for
not only outlining their pioneering studies, but also describing their recent
work and indicating important possibilities for the future. The excellent
secretarial work of Elisabeth Richardsen is greatly appreciated.
We hope the book will prove to be of use to a wide variety of clinicians
as well as basic scientists and students.

Troms!/l HELGE REFSUM ILMAR A. SULG KNuT RASMUSSEN

Contents

Part I: Physiology and Pharmacology

Chapter 1
Spasm of Cerebral and Coronary Vessels: Effects of Calcium
Antagonists
K.-E.ANDERSSON, L.BRANDT, and B.LJUNGGREN . . . . . . . 3

Chapter 2
Opioid Peptides and the Cardiovascular System with Especial
Reference to Low Perfusion States
J. R. PARRATI. With 4 Figures . . . . . . . . . . . 20

Chapter 3
Cardiac and Cerebral Effects of Local Anesthetics
R. HOTVEDT and H. REFSUM. With 10 Figures .......... 37

Chapter 4
Neurotoxins as Tools in Studying Cardiac Excitation-Contraction
Coupling
U. RAVENS and E. WE'ITWER. With 8 Figures . . . . . . . . . . . 51

Chapter 5
Adenosine and ATP Interactions with Autonomic Neural Control
of the Heart
A. PELLEG. With 1 Figure . . . . . . . . . . . . . . . . . . . . .. 62

Part II: Autonomic Nervous System and Arrhythmias

Chapter 6
Sympathetic Influences on Arrhythmogenesis in the
Ischemic Heart
K. A. YAMADA, G. P. HEATHERS, S. M. POGWIZD, and P. B. CORR . . .. 79

Chapter 7
Sympathetic Nervous System and Malignant Arrhythmias:
Evidence for Further Links
S.G. PRIORI and P. J. SCHWARTZ. With 6 Figures . . . . . . . 98
Contents VII

Chapter 8
Modulation of Cardiac Arrhythmias by the Autonomic Nervous
System
R. F. GILMOUR JR., J. J. SALATA, and D. P. ZIPES. With 8 Figures . 108
Chapter 9
Supraventricular Tachycardia and the Autonomic Nervous System
D. L. Ross. With 8 Figures . . . . . . . . . . . . . . . . . . . . . . 120
Chapter 10
Heart Rate Changes and ECG Rhythm Disturbances in the Cluster
Headache Syndrome
D. RUSSELL. With 10 Figures . . . . . . . . . . . . . . . . . . . . . 131
Chapter 11
Blood Pressure Assessment in a Broad Chronobiologic Perspective
F. HALBERG, E. BAKKEN, G. CORNELISSEN, J. HALBERG, E. HALBERG,
and P. DELMORE. With 8 Figures . . . . . . . . . . . . . . . . . . . 142

Part III: Syncope and Sudden Death

Chapter 12
Mechanisms of Syncope and of Sudden Death Due to Ventricular
Tachyarrhythmias
G. BREITHARDT, M. BORGGREFE, A. PODCZECK, and A. MARTINEZ-RuBIO
With 8 Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
Chapter 13
Possible Role of the Fear Paralysis Reflex in Sudden Cardiac Death
B.KAADA. With 3 Figures . . . . . . . . . . . . . . . . . . . . . . 185
Chapter 14
Some Clinical Neurological Aspects of Syncope
O.JOAKIMSEN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
Chapter 15
Differential Diagnosis in Syncope and Epilepsy:
Clinical Neurophysiological and Cardiological Aspects
I.A.SuLG. With 1 Figure . . . . . . . . . . . . . . . . . . . . . . . 202

Part IV: Thromboembolism and Ischemia

Chapter 16
Thromboembolic Complications in Atrial Fibrillation
J. GODTFREDSEN and P. PETERSEN . . . . . . . . . . . . . . . . . . . 225

Chapter 17
Echocardiography and Embolic Sources in the Heart
K.-A.JOHANNESSEN. With 3 Figures . . . . . . . . . . . . . . . . . . 230
VIII Contents

Chapter 18
31pNuclear Magnetic Resonance Spectroscopy of Cerebral and
Cardiac Ischemia
M.M.CoHEN, S.J.Kopp, J.W. PEITEGREW, and T. GLONEK . . . 239
Chapter 19
Oxygen Radicals in Heart and Brain Tissue Injury
K.YTREHus and O.D.MJ0s. With 8 Figures .. 244
Chapter 20
Prevention of Ischemic Brain Damage Following Cardiac Arrest
L.MoGENSEN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255

Part V: Sleep and Sleep Apnea

Chapter 21
Influence of Sleep, Activity and Circadian Rhythm on Heart Rate,
QT Interval and Cardiac Arrhythmias
C. GUILLEMINAULT and A. M. GILLIS . . . . . . . . . . . . . . . . . . 263
Chapter 22
Pulmonary Hemodynamics in Obstructive Sleep Apnea Syndromes
J. KRIEGER, E. WEITZENBLUM, B. REITZER, and D. KURTz. With 1 Figure 272
Chapter 23
Sleep Apnea Syndrome as an Occupational Disease
P. MONsTAD, I. A. SULG, A. K. ROM, T. NISSEN, and S. I. MELLGREN
With 2 Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280

Part VI: Disease in the Cardiovascular and Nervous System

Chapter 24
Concomitant Manifestations of Disease in the Cardiovascular
and Nervous System: An Overview
K. RASMUSSEN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
Chapter 25
Cardiac Involvement in Kearns-Sayre Syndrome
B. SCHWARTZKOPFF, G. BREITHARDT, M. BORGGREFE, B. LOSSE,
K.-Y. TOYKA, and H. FRENZEL. With 6 Figures . . . . . . . . 293
Chapter 26
Some Neurological and Hereditary Aspects of Progressive External
Ophthalmoplegia and Mitochondrial Encephalomyopathy
S. I. MELLGREN, T. TORBERGSEN, E.B.MATHIESEN, N.J.BRAUTAsET,
and S. LINDAL. With 3 Figures . . . . . . . . . . . . . . . . .. 311
Chapter 27
Neurological and Cardiological Findings in Systemic Lupus
Erythematosus
R.OMDAL, S. I. MELLGREN, and G.HuSBY . . . . . . . . . . . . . . . 318
Contents IX

Part VII: Hypo- and Hyperbaric Environment

Chapter 28
High Altitude Physiology and Pathophysiology: Medical
Observations During the Norwegian Mount Everest Expedition
K. T. SWKKE. With 3 Figures . . . . . . . . . . . . . . . . . 327
Chapter 29
Heart and Brain Under Hyperbaric Conditions in Man
A. O. BRuBAKK. With 8 Figures . . . . . . . . . . . . . . . . . . . . 343
Chapter 30
Effects of Positive End-Expiratory Pressure Ventilation
on Intracranial Pressure and Cerebral Blood Flow
O.HEVR0Y, N.-E.KL0W, 0. NYGAARD, and J.H.TRUMPY
With 3 Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357

Chapter 31
Positive End-Expiratory Pressure and Cardiac Function:
The Role of Extraventricular Constraint
O.A.SMISETH, I. KINGMA, N.W. ScoTI-DoUGLAs, E.RSMITH,
and J. V. TYBERG. With 6 Figures . . . . . . . . . . . . . . . . . . . 364

Part VIn: Cerebral Function and Cardiac Surgery

Chapter 32
Neuromonitoring in High Risk Surgery: Physiological Tolerance
Limits for Central Nervous System
I. A. SULG. With 10 Figures . . . . . . . . . . . . . . . . . . . 375

Chapter 33
Monitoring Brain Function During Cardiovascular Surgery:
Hypoperfusion vs Microembolism as the Major Cause
of Neurological Damage During Cardiopulmonary Bypass
E. RJoHN, L.S.PRiCHEP, RJ.CHABOT, and W.O.ISOM. With 9 Figures 405
Chapter 34
Cerebral Blood Flow During Cardiopulmonary Bypass
L.HENRIKSEN. With 7 Figures . . . . . . . . . . . . . . . . . . . . . 422
Chapter 35
Cerebral Hemodynamics During Nonpulsatile Cardiopulmonary
Bypass
T. LUNDAR. With 7 Figures . . . . . . . . . . . . . . . . . . . . . . 432

Chapter 36
Cerebral Outcome After Open Heart Surgery: A Long-term
Multidimensional Follow-up of Valvular Replacement Patients
K. SOTANIEMI. With 5 Figures . . . . . . . . . . . . . . . . . . . . . 440
x Contents

Chapter 37
Cerebral Protection During Open Heart Surgery:
Clinical, Psychometric, Enzymological, and Radiological Data
T. ABERG. With 3 Figures 452

Subject Index . . . . . . 459

List of Contributors

ABERG, T.
Department of Thoracic and Cardiovascular Surgery, University Hospital,
75124 Uppsala, Sweden
ANDERSSON, K.-E.
Department of Clinical Pharmacology, University Hospital, 22185 Lund,
Sweden
BAKKEN, E.
Medtronic Inc., Minneapolis, Minnesota 55455, USA
BORGGREFE, M.
Abteilung fur Kardiologie, Pneumologie und Angiologie, Medizinische
Klinik, Universitat Dusseldorf, MoorenstraBe 5, 4000 Dusseldorf, FRG
BRANDT, L.
Department of Neurosurgery, University Hospital, 22185 Lund, Sweden
BRAUTASET, N.J.
Department of Neurology, Central Hospital, 8000 Bodj/l, Norway
BREITHARDT, G.
Universitat Munster, Medizinische Universitatsklinik, 4400 Munster, FRG
BRUBAKK, A.O.
Department of Biomedical Engineering, Medical Faculty, Regional Hospi-
tal, University of Trondheim, 7000 Trondheim, Norway
CHABOT, R. J.
Department of Psychiatry, New York University Medical Center, New York,
New York 10016, USA
COHEN,M.M.
Department of Neurological Sciences, Rush-Presbyterian-St. Luke's Medi-
cal Center, 600 South Paulina, Chicago, Illinois 60615, USA
CORNELISSEN, G.
Chronobiology Laboratories, University of Minnesota, 420 Washington
Ave. S.E., Minneapolis, Minnesota 55455, USA
CaRR, P.B.
Cardiovascular Division, Washington University School of Medicine, 660
South Euclid Avenue, St.Louis, Missouri 63110, USA
XII List of Contributors

DELMORE, P.
Medtronic Inc., Minneapolis, Minnesota 55455, USA
FRENZEL, H.
Institut fur Pathologie, Medizinische Klinik, Universitat Dusseldorf, 4000
Dusseldorf, FRG
GILLIS, A. M.
Division of Cardiology, Stanford University School of Medicine, Stanford,
California 94305, USA
GILMOUR, R.E, JR.
Krannert Institute of Cardiology, Indiana University School of Medicine,
1100 West Michigan Street, Indianapolis, Indiana 46223, USA
GLONEK, T.
Nuclear Magnetic Resonance Laboratory, Chicago College of Osteopathic
Medicine, Chicago, Illinois 60615, USA
GODTFREDSEN, J.
Department of Cardiology, University of Copenhagen, Herlev University
Hospital, 2730 Herlev, Denmark
GUILLEMINAULT, C.
Sleep Disorders Center, Stanford University School of Medicine, Stanford,
California 94305, USA
HALBERG, E.
Chronobiology Laboratories, University of Minnesota, 5-187 Lyon Labo-
ratories, 420 Washington Ave. S.E., Minneapolis, Minnesota 55455, USA
HALBERG, E
Chronobiology Laboratories, University of Minnesota, 5-187 Lyon Labo-
ratories, 420 Washington Ave. S.E., Minneapolis, Minnesota 55455, USA
HALBERG, J.
Hennepin County Medical Center, Minneapolis, Minnesota 55455, USA
HEATHERS, G. P.
Cardiovascular Division, Washington University School of Medicine, 660
South Euclid Avenue, St.Louis, Missouri 63110, USA
HENRIKSEN, L.
Department of Neurosurgery, Rigshospitalet, Blegdamsvej 9, 2100 Copen-
hagen, Denmar~
HEVR0Y, O.
Department of Physiology, Institute of Medical Biology, University of
TromSj1j, 9000 TromSj1j, Norway
HOTVEDT, R.
Department of Anesthesiology, University Hospital, 9012 Tromsj1j, Norway
List of Contributors XIII

HUSBY, G.
Department of Rheumatology, University Hospital, 9012 Troms!/!, Norway
!sOM, W.O.
Department of Cardiothoracic Surgery, Cornell University Medical Col-
lege, New York , New York 10021, USA
O.
JOAKIMSEN,
Department of Neurology, University Hospital, 9012 Troms!/!, Norway
JOHANNESSEN, K.-A.
Department of Medicine, Haukeland Hospital, University of Bergen, 5000
Bergen, Norway
JOHN, E.R.
Department of Psychiatry, New York University Medical Center, and the
Nathan S. Kline Institute for Psychiatric Research, Orangeburg, New York
10962, USA
KAADA, B.
Laboratory of Clinical Neurophysiology, Rogaland Central Hospital, 4000
Stavanger, Norway
KINGMA, I.
Departments of Medicine and Medical Physiology, University of Calgary,
Calgary, Alberta TIN 4N1, Canada
KL0W, N.-E.
Department of Physiology, Institute of Medical Biology, University of
Troms!/!, 9001 Troms!/!, Norway
KOPp, S.J.
Department of Physiology, Chicago College of Osteopathic Medicine,
Chicago, Illinois 60615, USA
KRIEGER, J.
Service d'Explorations Fonctionnelles du Systeme Nerveux, CHU 1 Place
de I'Hopital, 67091 Strasbourg Cedex, France
KURTZ, D.
Service d'Explorations Fonctionnelles du Systeme Nerveux, CHU 1 Place
de I'Hopital, 67091 Strasbourg Cedex, France
LINDAL, S.
Department of Pathology, University Hospital, 9012 Troms!/!, Norway
LJUNGGREN,B.
Department of Neurosurgery, University Hospital, 22185 Lund, Sweden
LOSSE, B.
Abteilung fUr Kardiologie, Pneumologie und Angiologie, Medizinische
Klinik, Universitat Dusseldorf, MoorenstraBe 5, 4000 Dusseldorf, FRG
XIV List of Contributors

LUNDAR, T.
Department of Neurosurgery, Rikshospitalet, The National Hospital, Uni-
versity of Oslo, 0027 Oslo 1, Norway
MARTINEZ-RUBIO, A.
Abteilung fur Kardiologie, Pneumologie und Angiologie, Medizinische
Klinik, Universitat Dusseldorf, MoorenstraBe 5,4000 Dusseldorf, FRG
MATHIESEN, E. B.
Department of Neurology, University Hospital, 9012 Troms!/!, Norway
MELLGREN, S. I.
Department of Neurology, University Hospital, 9012 Troms!/!, Norway
M.r0s,0.D.
Department of Physiology, Institute of Medical Biology, University of
Troms!/!, 9001 Troms!/!, Norway
MOGENSEN, L.
Division of Cardiology, Department of Internal Medicine, Karolinska
Hospital, 10401 Stockholm, Sweden
MONSTAD, P.
Department of Neurology, University Hospital, 9012 Troms!/!, Norway
NISSEN, T.
Department of Neurology, University Hospital, 9012 Troms!/!, Norway
NYGAARD,0.
Department of Neurosurgery, University Hospital, 9012 Troms!/!, Norway
OMDAL, R.
Department of Internal Medicine, Rogaland Central Hospital, 4000 Sta-
vanger, Norway
PARRATI, J.R.
Department of Physiology and Pharmacology, University of Strathclyde,
Glasgow G11XW, United Kingdom
PELLEG, A.
The Lankenau Medical Research Center, Lancaster A venue west of City
Line, Philadelphia, Pennsylvania 19151, USA
PETERSEN, P.
Department of Cardiology, University of Copenhagen, Herlev University
Hospital, 2730 Hedev, Denmark
PETTEGREW, J. W.
Neurophysics Laboratory, Western Psychiatric Institute and Clinics, Pitts-
burgh, Pennsylvania 15213, USA
PODCZECK, A.
Abteilung fur Kardiologie, Pneumologie und Angiologie, Medizinische
Klinik, Universitat Dusseldorf, MoorenstraBe 5, 4000 Dusseldorf, FRG
List of Contributors xv
POGWIZD, S. M.
Cardiovascular Division, Washington University School of Medicine, 660
South Euclid Avenue, St. Louis, Missouri 63110, USA
PRICHEP, L. S.
Department of Psychiatry, New York University Medical Center, and the
Nathan S. Kline Institute for Psychiatric Research, Orangeburg, New York
10962, USA
PRIORI, S. G.
Istituto Clinica Medica Generale e Terapia Medica, Universita degli Studi
di Milano, Via F. Sforza 35, 20122 Milan, Italy

RASMUSSEN, K.
Section of Cardiology, Department of Internal Medicine, University Hos-
pital, 9012 Troms!/!, Norway .

RAVENS, U.
Institut fUr Pharmakologie, Universitat Essen, HufelandstraBe 55, 4300
Essen 1, FRG

REFSUM, H.
Department of Physiology, Institute of Medical Biology, University of
Troms!/!, 9001 Troms!/!, Norway
REITZER, B.
Service d'Explorations Fonctionnelles du Systeme Nerveux, CHU 1 Place
de l'Hopital, 67091 Strasbourg Cedex, France
ROM, A.K.
Occupational Health Service, Stakkevollan, 9000 Troms!/!, Norway
Ross, D.L.
Cardiology Unit, Department of Medicine, Westmead Hospital, Sydney
NSW 2145, Australia
RUSSELL, D.
Department of Neurology, Rikshospitalet, The National Hospital, Univer-
sity of Oslo, 0027 Oslo 1, Norway
SALATA, J.J.
Krannert Institute of Cardiology, Indiana University School of Medicine,
1100 West Michigan Street, Indianapolis, Indiana 46223, USA

SCHWARTZ, P. J.
Istituto Clinica Medica Generale e Terapia Medica, Universita degli Studi
di Milano, Via F. Sforza 35, 20122 Milan, Italy

SCHWARTZKOPFF, B.
Abteilung fur Kardiologie, Pneumologie und Angiologie, Medizinische
Klinik, Universitat Dusseldorf, MoorenstraBe 5, 4000 Dusseldorf, FRG
XVI List of Contributors

SCOTI-DoUGLAS, N. W.
Departments of Medicine and Medical Physiology, University of Calgary,
Calgary, Alberta T2N 4N1, Canada
SMISETII,O.A.
Department of Medicine, Rikshospitalet, The National Hospital, Univer-
sity of Oslo, 0027 Oslo 1, Norway
SMITII, E. R.
Departments of Medicine and Medical Physiology, University of Calgary,
Calgary, Alberta T2N 4N1, Canada
SOTANIEMI, K.
Department of Neurology, University of Oulu, 90220 OuIu, Finland
STOKKE, K.T.
Department of Clinical Chemistry, Rikshospitalet, The National Hospital,
University of Oslo, 0027 Oslo 1, Norway
SULG,I.A.
Departments of Neurology and Clinical Neurophysiology, University Hos-
pital, 9012 TromSjlj, Norway
TORBERGSEN, T.
Department of Neurology, University Hospital, 9012 TromSjlj, Norway
TOYKA, K.-V.
Abteilung fUr Neurologie, Medizinische Klinik, Universitat Dusseldorf,
MoorenstraBe 5, 4000 Dusseldorf, FRG
TRUMPY,J.H.
Department of Neurosurgery, University Hospital, 9012 Tromsjlj, Norway
TYBERG, J. V.
Departments of Medicine and Medical Physiology, University of Calgary,
Calgary, Alberta T2N 4N1, Canada
WEITZENBLUM, E.
PavilIon Laennec, 67091 Strasbourg Cedex, France
WETIWER, E.
Institut fUr Pharmakologie, Universitat Essen, HufelandstraBe 55, 4300
Essen 1, FRG
YAMADA, K. A.
Cardiovascular Division, Box 8086, Washington University School of Medi-
cine, 660 South Euclid Avenue, St.Louis, Missouri 63110, USA
YTREHUS, K.
Department of Physiology, Institute of Medical Biology, University of
TromSjlj, 9001 TromSjlj, Norway
ZIPES, D.P.
Krannert Institute of Cardiology, Indiana University School of Medicine,
1100 West Michigan Street, Indianapolis, Indiana 46223, USA
Part I
Physiology and Pharmacology
Chapter 1

Spasm of Cerebral and Coronary Vessels:

Effects of Calcium Antagonists
K.-E. ANDERSSON, L. BRANDT, and B. LJUNGGREN

Introduction

Calcium antagonists [130], blocking calcium entry into cells, have a well-known re-
laxant effect on vascular smooth muscle. However, this does not mean that they pro-
duce a uniform vasodilation in all vascular beds [5, 27, 28]. It is generally considered
that calcium antagonists are more potent in coronary and cerebral vessels than in
peripheral vessels [27]. The reasons for such a preferential action have not been
elucidated. A selective action on cerebral and coronary vessels may contribute to a
beneficial effect of these drugs in several cerebral and cardiac disorders, and particu-
larly in the syndromes of cerebral and coronary vasospasm where the use of calcium
antagonists has been rewarding.
Neither in cerebral nor in coronary vessels have the factors controlling contrac-
tion been established. Assuming that calcium antagonists act mainly at the cell mem-
brane level by preventing the influx of extracellular calcium, they may give informa-
tion on what role the extracellular calcium concentration has in maintaining basal
tone and in the contractile activation produced by different agents. Whether or not
their effects can also be used as indicators of what factors are involved in cerebral
and coronary vasospasm can only be speculated upon.
Contractile activation of vascular smooth muscle may be conveniently studied in
vitro using isolated vascular segments. Despite the obvious limitations of this ap-
proach [122, 135] it may give information relevant for the understanding of a clinical
condition. Considering the large species variation in reactivity of both cerebral and
coronary arteries to different agents, experiments performed on human vessels may
be particularly relevant. In the present review, emphasis has been given to experi-
ments performed on isolated human cerebral and coronary arteries to elucidate
mechanisms behind the clinical syndromes of cerebral and coronary vasospasm. Al-
though arterial spasm has been implicated in both syndromes, and calcium antago-
nists seem to provide an effective prophylactic treatment, this does not mean that the
mechanisms evoking the contractions are similar.

Cerebrovascular Effects of Calcium Antagonists

Contractile Activation of Brain Vessels

Bevan [12] pointed out that the characteristics of the vascular smooth muscle of
larger pial arteries seem to differ both qualitatively and quantitatively from those of
4 K.-E. Andersson et al.

other arteries of similar size. This includes, e.g., electrophysiological properties, re-
ceptor characteristics, and receptor-response coupling mechanisms [12]. The marked
species variation in the responses to vasoactive agents, human vessels often differing
from those of animals, was stressed by Toda [120].

Myogenic Tone. One important characteristic of the cerebral circulation is its ability
to autoregulate, i.e., respond to an increase in perfusion pressure with vasoconstric-
tion and to a decrease in perfusion pressure with vasodilation. Evidence has been
presented for the predominance of either myogenic [23] or metabolic [133] mecha-
nisms in cerebral flow regulation. Harder [47] showed that elevation of transmural
pressure in cat middle cerebral arteries resulted in muscle depolarization and that
this was a myogenic response. Spontaneous action potentials were observed and in-
creased in frequency when the pressure was increased from 100 to 140mmHg. This
spike activity was inhibited by lowering extracellular calcium and by verapamil. Re-
duction of extracellular calcium produced a pressure-induced decrease in vascular di-
ameter. Osol and Halpern [93] and Halpern and Osol [44] suggested, based on ex-
periments on rats, that purely myogenic responses can account for a portion of the
adjustment in arterial caliber required for cerebral flow regulation and that they are
supplemented by metabolic mechanisms operative in the intact brain.
Myogenic tone, i.e., tone occurring in vitro in the absence of applied agonists,
has been demonstrated in isolated cerebral blood vessels from man as well as from
several animal species. Bevan and Hwa [13] reported this tone to be stretch depen-
dent and also dependent on the concentration of calcium in the extracellular me-
dium. It was relatively resistant to calcium antagonists compared with agonist and
K+-induced responses [12, 13]. Nakayama et at. [88,89] found that quick stretches
applied to helical strips of rabbit cerebral artery produced a pronounced, delayed
tension development. This response was more resistant to the removal of extracel-
lular calcium and to calcium antagonists than contractions evoked by K+ or electrical
stimulation and it was suggested that it was associated with calcium release from the
inner surface of the plasma membrane [89].
If myogenic mechanisms are of decisive importance for autoregulation of cere-
bral blood flow, calcium antagonists would not be expected to influence autoregula-
tory responses. Published results seem conflicting. Thus, Harris et at. [49] reported in-
tact vasodilator autoregulation in baboons receiving nimodipine, whereas McCalden
et at. [79] reported an impairment. Sahlin et at. [100] found that nimodipine im-
paired the autoregulatory response to hypertension, whereas Harris et at. [49] did
not. Whether or not calcium antagonists cause a clinically significant impairment of
cerebral autoregulation remains to be established.

Stimulus-Contraction Coupling. The electrical events accompanying mechanical re-

sponses in cerebral vessels seem to vary considerably, not only between different
species and modes of excitation, but also between different parts of the cerebral cir-
culation [46]. However, both electromechanical [45, 73] and pharmacomechanical
[30,48,62] coupling have been demonstrated to occur in brain vessels.
McCalden and Bevan [78] investigated the sources of activator calcium in rabbit
basilar artery. They recorded the responses to K+, noradrenaline (NA) , and 5-
hydroxytryptamine (5-HT) before and after incubation in calcium-free Krebs' solu-
Spasm of Cerebral and Coronary Vessels 5

tion and also studied the effects of the calcium antagonist D 600. They found that
both calcium-free solution and D 600 effectively reduced the responses to all three
agonists and concluded that K+ caused contraction by using a single calcium pool,
probably of extracellular origin. NA and 5-HT also primarily utilized extracellular
calcium. Skarby et al. [109] studied the effects of calcium-free medium and nifedipine
on K+ - and NA-induced contraction in rat and cat middle cerebral arteries. Their
findings suggested that in cat middle cerebral artery, both K+- and NA-induced
contractions are almost exclusively dependent on the presence of calcium in the extra-
cellular medium and that activation occurs through pathways sensitive to calcium an-
tagonists. This is in contrast to findings in isolated human pial vessels [15] where treat-
ment in a calcium-free medium for 30 min and exposure to nifedipine or nimodipine
markedly reduced K+ -, but not NA- or 5-HT-induced contractions. It was suggested
that this was due to the amines using intracellularly stored calcium for their contraction.
Sasaki et al. [102] investigated the dependence of cerebral arterial contractions
on intracellularly stored calcium. They compared the responses induced by a stable
thromboxane A2 (TxA2) analogue, prostaglandin F2a (PGF2a ), and 5-HT in canine
and monkey basilar arteries which had been exposed to a calcium-free medium. This
exposure inhibited completely contractions induced by K+. 5-HT, PGF2a , and the
TxA2 analogue used caused contractions amounting to a maximum of 23% (TxA2
analogue) of controls in normal calcium-containing medium. This was interpreted to
reflect release of intracellular calcium. Uski and Andersson [127] found that in cal-
cium-free medium PGF2a was able to induce a biphasic contraction in feline basilar
arteries, probably by releasing cellularly bound calcium from two different stores.
The second contractile phase could be blocked by increasing the EGTA concentra-
tion in the extracellular medium and by nifedipine. Brandt et al. [15] found that in
isolated human pial arteries even high concentrations of nimodipine and nifedipine,
which abolished K+ -induced contraction, reduced PGF2a-induced contractions by
only about 60% .
The effectiveness of calcium antagonists as inhibitors of contraction in isolated
human cerebral arteries has been demonstrated by several investigators. Thus, irre-
spective of what agent was used for producing contraction, e.g., K+, NA, 5-HT,
PGF2a , blood, or posthemorrhagic cerebrospinal fluid, calcium antagonists to vari-
ous degrees blocked the contraction [15, 16,22,87,101].

Comparison with Peripheral Vessels. Shimizu et al. [106] showed that verapamil,
nifedipine, and diltiazem more effectively relaxed cerebral vessels (dog basilar and
middle cerebral artery) than peripheral vessels (coronary and mesenteric vessels)
contracted by PGF2a , and several other studies both in vitro [2, 12, 15, 51, 56, 87,
123,124,140] and in vivo [50, 53, 65, 79] suggest that contractions in cerebrovascular
smooth muscle can be more effectively prevented or relaxed by calcium antagonists
than contractions in peripheral vascular smooth muscle. This may be related to a
greater dependence on extracellular calcium of cerebral than of peripheral vessels,
but this has not been definitely established. Thus, Brandt et al. [15] found that the
relaxant potency of nimodipine was similar in K+ -contracted isolated human pial and
mesenteric arteries. Available information suggests that the requirement of extra-
cellular vs intracellular calcium for contraction of brain arteries is dependent on what
agent is used for contractile activation.
6 K.-E. Andersson et al.

Cerebral Vasospasm

According to Wilkins [138] intracranial arterial spasm (also known as cerebral vaso-
spasm) may be defined as: (a) an arteriographically evident narrowing of the lumen
of one or more of the major intracranial arteries at the base of the brain due to
contraction of the smooth muscle within the arterial wall or morphological changes
in the arterial wall and along its endothelial surface that occur in response to vessel
injury; (b) the delayed onset of a neurological deficit after subarachnoid hemorrhage
thought to be due to ischemia or infarction of a portion of the brain; or (c) the com-
bination of these two features (symptomatic vasospasm).
Arterial narrowing may be detected in 30% -70% of arteriograms performed 4-
12 days after aneurysmal subarachnoid hemorrhage (SAH). However, not more
than 20%-30% of such patients suffer delayed ischemic deterioration (DID) [52, 63].
The latter patients almost invariably have regional or generalized vasospasm on their
arteriograms. Ljunggren et al. [71] found that in 137 patients who were operated
within 72 h after the bleed 13% developed symptoms of DID with permanent deficits.
The onset of angiographic vasospasm is gradual, rarely pronounced before the
4th day after the initial hemorrhage, and reaches a peak around the 7th day [63].
This time course, and the fact that once established vasospasm is not reversed by
available vasodilators or by inhibitors of specific vasoactive substances, make it ques-
tionable whether it is only a vasoconstrictive phenomenon. There is no doubt that
pathological changes occur in the vessel wall in vasospasm after SAH [63, 135]. As
these may be secondary to vasoconstriction [61], attempts to prevent vasoconstric-
tion as soon as possible after SAH would be a logical approach to the problem.
The agents responsible for cerebral vasospasm are unknown. Fisher et al. [26]
and Mizukami et al. [84] found a clear correlation between the site of the major sub-
arachnoid blood clots and the occurrence of severe vasospasm. It appears that sub-
arachnoid blood contains or causes the release of the agent(s) causing the vasospasm
and many mediator candidates have been suggested. These include adrenaline, NA,
5-HT, angiotensin, prostanoids, hemoglobin, K+, and others [52, 63, 113, 136].
However, none has been shown to be more important than the others, and no an-
tagonist of a single mediator candidate has been demonstrated to be therapeutically
effective [138].
It has been suggested that there is an increased sensitivity of the vessels to vaso-
constrictor stimuli after experimental SAH [114]. Whether this is the case in man is
not known. Wijdicks et al. [137] showed the presence of a substance reacting with
digoxin antibodies in SAH patients not receiving digoxin. The presence of the sub-
stance was significantly related to the amount of blood and to the presence of blood
in the frontal interhemispheric fissure. They suggested that in SAH patients a digoxin-
like factor with natriuretic properties is released, probably as a result of hypothalamic
damage. If such a factor inhibits Na+-K+-adenosine triphosphatase (ATPase) like
digoxin, it might contribute to an increased reactivity of cerebral vessels. Having
natriuretic properties, it may also contribute to the hyponatriemia frequently noted
when vasospasm develops [52].
If an increased smooth muscle activity contributes to symptomatic vasospasm,
the experimental findings that calcium antagonists effectively prevent contractions
induced by almost all contractile agents make it logical to use these drugs in patients
Spasm of Cerebral and Coronary Vessels 7

at risk of developing cerebral vasospasm. Even if calcium antagonists seem to have

little effect when the vasospasm has developed [43], evidence is accumulating that it
is possible to reduce symptomatic vasospasm after SAH by these drugs [1, 8, 21, 67,
68, 70, 72, 89, 90, 94]. Allen et al. [1] found in a multicenter, placebo-controlled trial
that orally administered nimodipine reduced the occurrence of severe neurological
deficits in patients with aneurysmal SAH and angiographically confirmed vaso-
spasm. In 140 patients subjected to acute stage clipping for a ruptured supratentorial
aneurysm within 72h after hemorrhage, and additional postoperative intravenous
nimodipine, Ljunggren et al. [72] reported an overall incidence of DID with perma-
nent neurological deficit of 3%.

Postischemic Cerebrovascular Spasm

Brain tissue damage after a period of ischemia is a complex phenomenon involving

several pathogenic factors [107]. Hossman and Sato [55] showed that the restoration
of the blood supply after more than 10 min of global ischemia produced functional
neuronal recovery, but this was again suppressed in the later phase of the postischemic
period. Several observations indicate that an increased vascular resistance to flow
may be a decisive factor behind this reperfusion-induced tissue damage. Localized
vascular changes have been suggested to be responsible for the fact that when blood
flow to the brain is restored after prolonged ischemia, some areas are not reperfused
and ultimately die [4]. Such vascular damage may involve the endothelium leading to
loss of an endothelial-derived relaxant factor (EDRF [31, 129]). It has been demon-
strated that the pial artery response to, e.g., NA, was potentiated by endothelium re-
moval [104]. As neither intravascular clotting nor platelet aggregates appear to cause
the vascular obstruction, the contractile state of the cerebral resistance vessels may
be responsible for the postischemic reperfusion damage. Brandt et al. [15] produced
focal ischemia in cats by clamping the middle cerebral artery. After an initial vaso-
dilatation, a marked, long-standing vasoconstriction occurred. This vasoconstriction
was immediately reversed by topical application of nifedipine. In a later study, Teas-
dale et al. [115] were able to demonstrate a direct relationship between the vasocon-
striction and an increased extravascular K+ concentration. Kazda and Mayer [66]
showed that 7 min of global ischemia in cats produced an increase of the extracellular
K+ concentration at the surface of the temporoparietal cortex up to values of more
than 50 mmolil. They suggested that a high extracellular K+ concentration produces
sustained vascular contraction by depolarizing the vessels. A role of extracellular K+
for the development of cerebrovascular spasm was also suggested by Brandt et al.
[15] who found that the threshold for K+ -induced contraction in human pial arteries
was below 10mmolil.
K+ is considered to induce vascular contraction by stimulating calcium influx
through potential sensitive calcium channels. This vascular constriction may be one
reason for the ischemic damage. Another might be that K+ -induced calcium influx
can be increased also in the cerebral neurons and that this together with the ischemic
insult will aggravate the ischemic damage. Thayer et al. [116] investigated the effects
of dihydropyridine drugs on the voltage-sensitive influx of Ca2+ into central nervous
system neurons grown in primary culture. They found that the depolarization in-
8 K.-E.Andersson et al.

duced by increasing the extracellular calcium concentration was associated with a

calcium influx which was enhanced by the calcium channel agonist Bay K 8644 and
inhibited by the antagonist nitrendipine. The effects of the drug were different in dif-
ferent brain regions, but suggested a widespread distribution of functional dihydro-
pyridine-sensitive calcium channels in the central nervous system.
If K+ -induced calcium influx through voltage-operated calcium channels plays an
important role in ischemic brain damage, it should be expected that calcium antago-
nists which effectively block voltage-operated calcium channels would be therapeuti-
cally effective. Steen et al. [110] subjected dogs to 10 min of complete ischemia by
temporary ligation of the aorta. They found that nimodipine improved neurological
recovery and nearly doubled cerebral blood flow in the delayed postischemic hypo-
perfusion period without significant effect on metabolism. Four of five treated dogs
were normal whereas six of seven controls were either severely damaged or dead.
Kazda and Mayer [66] found that pretreatment with nimodipine in cats subjected to
global ischemia did not reduce the ischemic increase in extracellular K+ but com-
pletely prevented the postischemic reduction in cerebral blood flow. Also when
nimodipine treatment was delayed up to 60 min after reperfusion, cerebral blood
flow increased and outcome was improved [82]. Flunarizine given under similar con-
ditions did not, however, improve either cerebral blood flow or neurological out-
come [91]. In rats subjected to a 20-min period of high-grade forebrain ischemia,
Vibulsresth et al. [132] found no beneficial effect of nimodipine when given 3 min
after restoration of circulation to the brain. Similar results were obtained by Gotoh
et al. [42] who administered nimodipine 5 min after occlusion of the middle cerebral
artery. Nimodipine modified neither the pattern of cerebral blood flow distribution
after the occlusion nor the extent of ischemic brain damage as determined by histo-
logical examination.
Even if pretreatment with calcium antagonists may have a beneficial effect on re-
perfusion associated brain damage [39] the results thus seem conflicting as to
whether or not the drugs are effective when given after the restoration of blood flow.
In man, the possibility of such a beneficial effect is suggested by the open study of
Roine et al. [98]. Of 19 patients resuscitated after ventricular fibrillation and receiv-
ing intravenous nimodipine, 14 survived and 12 could be discharged home. In a con-
trol group (historical controls), only 5 of 19 patients survived and could be dis-
charged home.

Coronary Vascular Effects of Calcium Antagonists

Contractlle Activation of Coronary Vessels

The coronary vasculature can be divided into the small resistance vessels, the large
epicardial coronary arteries, and the coronary collateral vessels. The responses of
these types of vessels to constrictor agents and to dilators may be different [103]. For
all types of vessels, however, it may be concluded that their basal tone and state of
contractile activation are regulated by complex systems [9, 25, 92].
Spasm of Cerebral and Coronary Vessels 9

Myogenic Tone. The high basal tone of the coronary vasculature, and its ability
to autoregulate, seem to be mainly myogenic in origin, even if a contribution of a-
adrenoceptor-mediated contraction cannot be excluded [92]. In isolated human epi-
cardial arteries, basal tone was found to be dependent on extracellular calcium [34,
131]. The role of extracellular calcium for contractile activation of coronary smooth
muscle has not been elucidated in detail. Feigl [25], reviewing the literature, con-
cluded that calcium per se produced constriction of coronary vessels, but that there
is no evidence that extracellular calcium is an important physiological mediator in
the local control of the coronary circulation. This does not exclude that extracellular
calcium may be important both for maintaining basal tone and for contractile activa-
tion of coronary vessels produced by different agents. Such vessels also demonstrate
spontaneous rhythmic periods of contraction and relaxation [34, 40, 64, 99, 131],
which were abolished by verapamil, D 600, diltiazem [34, 131], and nifedipine [40],
but was not affected by agents blocking a- or /3-adrenoceptors, muscarinic receptors,
or histamine receptors [34]. Vedernikov [131] suggested that the occurrence of spon-
taneous contractions depends on the conditions of pacemaker activation and the
propagation of excitation along the smooth muscle cells. The activity may be
triggered by calcium influx through voltage-sensitive calcium channels [35], which
explains its sensitivity to calcium antagonists.

Stimulus-Contraction Coupling. The control of the coronary circulation both at rest

and during exercise [111] appears to be exerted by a combination of neural and meta-
bolic events and to involve a number of mediators both of contraction and of relaxa-
tion [see, e.g. 25, 92, 111].
Studies performed mainly on isolated canine arteries have indicated that large
epicardial vessels have a predominance of postjunctional a-adrenoceptors (al and a2)
mediating contraction whereas small coronary arteries are equipped almost exclu-
sively with /3-adrenoceptors (PI) mediating relaxation [25, 86]. Studies on isolated
human epicardial coronary arteries suggest the occurrence of both a- and /3-adreno-
ceptors [6, 35, 118, 131]. The a-adrenoceptors seem to be of both al- and a2-type
[121]. The response to a-adrenoceptor activation was variable and weaker than that
of other agonists studied [35, 59, 118], and contraction was inversely related to the
distance from the coronary artery orifice [118]. /3-Adrenoceptor stimulation (iso-
prenaline, but not salbutamol) relaxed isolated human coronary arteries [35].
Van Breemen and Siegel [128] investigated the sources of calcium used for NA-
induced activation of canine coronary conduit arteries. They found that blockade of
calcium influx by removal of extracellular calcium, or addition of 10 mM LaCh, pre-
vented the NA-induced contraction. In addition, NA failed to stimulate calcium
efflux. This contrasted with findings in the rabbit aorta. Van Breemen and Siegel
[128] suggested that NA activates isolated canine coronary arteries by stimulation of
a-adrenoceptors leading to an opening of calcium channels in the membrane and cal-
cium influx, but not to a significant release of intracellular calcium. They also sug-
gested this to be the reason why coronary arteries were particularly sensitive to cal-
cium antagonists. However, their findings may not be valid for human coronary ar-
teries, nor do they exclude that agonists other than NA can contract coronary vessels
by calcium-ant agonist-insensitive pathways or by causing release of intracellularly
stored calcium. Based on their experiments, Ginsburg et al. [36] concluded that
10 K.-E.Andersson et al.

human epicardial coronary vessels used both extra- and intracellularly stored calcium
for agonist, e.g., NA-induced, activation, and Vedernikow [131] arrived at a similar
conclusion.
Human coronary arteries constrict in response to muscarinic receptor stimulation
[32,35,59, 118, 131]. This effect is independent of the endothelium and appears to
be mediated by muscarinic receptors on the smooth muscle cells [32]. Pig coronary
arteries contracted by ACh were effectively relaxed by calcium antagonists [29].
Vedernikov [131] showed in isolated human coronary arteries that even in the pres-
ence of calcium antagonists ACh was able to induce contraction. This was taken
as evidence for ACh being able to release calcium from membrane or intracellular
stores.
Histamine contracts isolated human coronary arteries by stimulation of HI recep-
tors in the smooth muscle cells [37, 41, 118, 131]. The response is considered to be
partially masked by vasodilatation mediated via H 2-receptors in smooth muscle and
by a relaxing factor released by activation of endothelial Hrreceptors [122]. His-
tamine was less potent than most other agonists but had a high intrinsic activity [35].
Coronary arteries contain high amounts of histamine and 5-HT, and vessels from
cardiac patients were found to be hyperreactive to these agents [60].
5-HT is as powerful a constrictor of isolated human coronary arteries as his-
tamine or ACh [59]. Miiller-Schweinitzer [86] suggested that the well-known con-
tractant effect of ergonovine was mediated via 5-HT-receptors, at least in dog coro-
nary vessels, but this was not felt to be the case in human coronary arteries [35]. His-
tamine- and 5-HT-contracted pig coronary arteries were effectively relaxed by cal-
cium antagonists [29].
Prostaglandins produce either contraction or relaxation of human coronary ar-
teries. PGE2, PGH2, and PGF2a all contract the vessels as does the endoperoxide
U 44069 [35]. Toda [119] reported carbocyclic thromboxane A2 (cTxA2) to be the
most potent vasoconstrictor among substances ever tested in isolated human coro-
nary arteries, including NA, ACh, and histamine. Ginsburg [35] found prostacyclin
(PGI2) to cause relaxation at low but contraction at high concentrations. In isolated
human coronary arteries, maximally precontracted with cTxA2 or partially precon-
tracted with PGF2a , Toda [119] demonstrated that PGh induced concentration-
related relaxations.
Canine coronary arteries were potently contracted by the TxA2 mimetic U 44619,
which was able to induce contraction independent of the extracellular calcium con-
centration [7]. Verapamil was almost ineffective against U 44619-produced contrac-
tions. On the other hand, in isolated human coronary arteries, precontracted with
cTxA2, verapamil caused a slowly developing relaxant response [119]. Nifedipine,
and verapamil in high concentrations, also suppressed rhythmic contractile activity
induced by cTxA2 [119, 134]. Human coronary arteries partially precontracted with
PGF2a were effectively relaxed by diltiazem [121].
In vivo calcium antagonists have been shown to dilate both large coronary vessels
and resistance vessels [103]. Holtz et al. [54] showed that vasodilator drugs may have
two effects on large epicardial coronary arteries, one flow dependent, which is
caused by release of a relaxant factor from the endothelium, and one direct caused
by direct smooth muscle relaxation. Holtz et al. [54] found that nifedipine, verapamil,
and diltiazem dilated the vessels even if the increase in flow caused by these agents
Spasm of Cerebral and Coronary Vessels 11

was prevented. They therefore seem to be direct dilators of epicardial coronary ar-
teries. There may be differences between the drugs, nifedipine and verapamil being
more effective than diltiazem [103].
Nifedipine, verapamil, and diltiazem all dilate coronary resistance vessels, but re-
sults on their effects on collateral vessels have been contradictory [see 103]. Interest-
ingly, nifedipine and diltiazem were found to decrease ischemia-induced vasodilata-
tion of the coronary resistance vessels [10, 11]. This is in contrast to the effects of
some calcium antagonists in the brain, where lidoflazine and flunarizine enhanced
reactive hyperemia elicited by an anoxic challenge, whereas verapamil had no effect
[95].

Endothelial-Derived Relaxant Factor (EDRF). ACh and several other agents may act
as vasodilators by releasing EDRF. As mentioned above, ACh and other muscarinic
agonists contract isolated human coronary arteries [32, 35, 59, 118]. ACh, when
given by intracoronary injection in man, also causes vasoconstriction [142]. Forster-
mann et al. [32] confirmed the inability of human coronary arteries to relax in re-
sponse to ACh, but stressed that this does not exclude that EDRF-mediated vasodi-
lation induced by other agents may occur in these vessels. The vessels investigated by
Forstermann et al. [32] were reported to be free of atherosclerosis. Diverging results
were reported by Bossaller et al. [14] who found that isolated human arteries without
atherosclerosis "responded to ACh with concentration-related relaxations. This was
not found in atherosclerotic vessels. Atherosclerotic arteries were still relaxed by
substance P, histamine, and the Ca2+ ionophore A 23187. It was suggested that
atherosclerotic arteries had a muscarinic receptor defect and not an inability of the
endothelium to release EDRF or the smooth muscle to respond to it.
Altura and Altura [3] using isolated canine coronary arteries found that removal
of extracellular Mg2+ inhibited the ability of these vessels to relax when challenged
with ACh and suggested that Mg2+ is an important cofactor for ACh-induced endo-
thelium-dependent relaxation. Mg2+ deficiency may induce coronary vasoconstric-
tion which is endothelium independent [69] and which can be effectively blocked by
calcium antagonists [57, 125]. The formation and release of ED RF appear to be de-
pendent also on calcium in the extracellular fluid. Singer and Peach [108] found that
both verapamil and nifedipine partially inhibited relaxation believed to be mediated
via EDRF. However, Jayakody et al. [58] found that nicardipine and diltiazem had
no significant effect on synthesis/release and the relaxant response to EDRF in the
rabbit aorta. Whether or not calcium antagonists interfere with synthesis, release, or
effects of EDRF is presently unclear. Obviously, their own vasodilatory action is in-
dependent of EDRF; it may be speculated whether or not they may affect the relaxant
response to other vasodilators.

Coronary Artery Spasm

It is now widely accepted that coronary artery spasm, defined as a transient constric-
tion of a large coronary artery resulting in myocardial ischemia [141], may play an
important role in the pathogenesis of many clinical manifestations of ischemic heart
disease including angina pectoris and myocardial infarction [19, 75, 76, 77, 117]. It
12 K.-E. Andersson et al.

has been accepted as the "proven" cause of variant angina, i.e., angina occurring at
rest and associated with ST-segment elevation in the electrocardiogram [81]. How-
ever, the factors responsible for coronary artery spasm are still unknown although
several mechanisms have been discussed [33, 105].
Yasue [141] stressed the marked variation in the incidence of anginal attacks in
variant angina patients, being more frequent from midnight to early morning when
the patients are at rest. In the early morning, even mild exercise could induce anginal
attacks, whereas in the afternoon exercise evoked few attacks. Theroux and Waters
[117] speculated that the cyclic pattern may be due to a central neurohormonal ef-
fect, possibly originating from the hypothalamus. If the hypothetical hypothalamic
factor has the properties of an endogenous digoxin-like substance, i.e., inhibits Na+-
K+ -ATPase, an increased sensitivity of the coronary smooth muscle to contractile
stimuli would result. Such a sensitization of the coronary vasculature is known to
occur after administration of cardiac glycosides, is dependent on extracellular cal-
cium, and can be prevented by calcium antagonists [29]. It cannot be excluded that
variant angina is the coronary manifestation of a generalized vasospastic disorder.
Such a hypothesis is supported by the significantly higher prevalence of migraine
headaches and Raynaud's phenomenon in patients with variant angina than in a con-
trol population [83].
The variation in the incidence of anginal attacks may, however, also be explained
by an increased parasympathomimetic activity during night, leading to decreased
metabolic demands, which are known to constrict large coronary arteries [74]. As
mentioned previously, ACh contracts isolated human coronary arteries, and when
injected intracoronary, ACh induces coronary spasm and anginal attacks in patients
with variant angina [142]. Other factors may contribute. At rest, particularly at night
when the metabolic activity is low, the production of hydrogen ions decreases. An in-
crease in pH increases vascular tone in isolated coronary arteries [29], and coronary
vasospasm and anginal attacks can be induced by hyperventilation and Tris-buffer in-
fusion [143]. A third possibility is variations in a-adrenoceptor-mediated coronary
tone [141]. Support for the involvement of a-adrenoceptors in variant angina was
given by the results of Tzivoni et al. [126] who found prazosin effective in six pa-
tients. On the other hand, Winniford et al. [139] found prazosin ineffective in a long-
term placebo-controlled study, and Chierchia et al. [17] concluded from their data
that activation of a-adrenoceptors does not play an important role in the genesis of
coronary spasm in variant angina.
Thus, several physiological stimuli seem to be able to evoke coronary vasospasm
in variant angina patients. This may reflect a local hyperreactivity to vasoconstrictors
in the large coronary arteries. Such a hyperreactivity seems to be nonspecific as
spasm can be induced by, e.g., ergonovine, ACh and methacholine [24, 142, 144],
histamine [138] as well as by cold pressor test [96] and hyperventilation [143]. Several
observations suggest that the increased sensitivity is localized to one coronary artery,
usually at the site of an atherosclerotic lesion [76]. This favors speculations that a loss
of EDRF may be involved [129].
In patients with variant angina, nifedipine, verapamil, and diltiazem effectively
reduce or eliminate angina caused by coronary spasm [see, e.g., 112]. In addition,
these drugs prevent vasospasm induced by ergonovine, exercise [see, e.g., 117], and
alkalosis [143]. This is in contrast to experiences with antagonists of several putative
Spasm of Cerebral and Coronary Vessels 13

mediators of coronary vasospasm. Thus, no beneficial effects were documented for

TxA2 antagonists [18, 97], the 5-HT2-receptor blocker ketanserin [20], or low-dose
aspirin [18].

Conclusions

Cerebral and coronary arteries depend on extracellular calcium both for maintaining
tone and for contractile activation by several agonists. Human arteries from these re-
gions seem to be able to use both extra- and intracellular calcium in the activation
process. It appears, however, that calcium influx triggering contraction to an impor-
tant extent occurs through pathways sensitive to calcium antagonists.
The factors causing constriction of cerebral arteries in the syndromes of symptom-
atic and postischemic cerebral vasospasm are still unknown. The beneficial effects of
prophylactic treatment with calcium antagonists suggest, but in no way prove, that
vasoconstriction is an important pathophysiological factor and that this vascular re-
sponse, at least initially, is mediated by calcium influx which can be blocked by cal-
cium antagonists.
The factors leading to coronary vasospasm also remain unknown, but are prob-
ably different from those involved in cerebral vasospasm. Local hyperreactivity to
vasoconstrictor stimuli seems to make it possible for several mediators to trigger the
contractile process. Irrespective of what mediator is involved, activation seems to be
mediated by calcium influx through calcium-antagonist-sensitive pathways.
The therapeutic efficacy of calcium antagonists in variant angina is well docu-
mented. Even if their beneficial effect can well be explained by a direct effect on
coronary artery smooth muscle, this does not exclude that other actions may also be
important.

References

1. Allen GS, Ahn HS, Preziosi TJ, Battye R, Boone SC, Chou SN, Kelly DL, Weir BK, Crabbe
RA, Lavik PJ, Rosenbloom SB, Dorsey FC, Ingram CR, Mellits DE, Bertsch LA, Boisvert
DPJ, Hundley MB, Johnson RK, Strom JA, Transou CR (1983) Cerebral arterial spasm - a
controlled trial of nimodipine in patients with subarachnoid hemorrhage. N Engl J Med 308:
619-624
2. Allen GS, Banghart SB (1979) Cerebral arterial spasm 9. In vitro effects of nifedipine on
serotonin-, phenylephrine-, and potassium-induced contractions of canine basilar and femoral
arteries. Neurosurgery 4:37-42
3. Altura BT, Altura BM (1987) Endothelium-dependent relaxation in coronary arteries requires
magnesium ions. Br J Pharmacol 91 : 449-451
4. Ames A, Wright RL, Kowada M, Thurston SM, Majno G (1968) Cerebral ischemia II. The no-
reflow phenomenon. Am J PathoI52:437-453
5. Andersson KE (1986) Pharmacodynamic profiles of different calcium channel blockers. Acta
Pharmacol Toxicol [Suppl 2] 58: 31-42
6. Andersson R, Holmberg S, Svedmyr N, Aberg G (1972) Adrenergic a and ,B-receptors in coro-
nary vessels in man. Acta Med Scand 191 :241-244
14 K.-E. Andersson et al.

7. Angus JA, Brazenor RM, LeDuc MA (1983) Responses of dog large coronary arteries to con-
strictor and dilator substances: implications for the cause and treatment of variant angina pec-
toris. Am J Cardiol 52: 52A-60A
8. Auer LM (1984) Acute operation and preventive nimodipine improve outcome in patients with
ruptured cerebral aneurysms. Neurosurgery 15: 57-66
9. Bache RJ, Dymek DJ (1981) Local and regional regulation of coronary vascular tone. Progr
Cardiovasc Dis 24: 191-212
10. Bache RJ, Dymek DJ (1982) Effect of diltiazem on myocardial blood flow. Circulation [Suppl1]
65:119-126
11. Bache RJ, Tockman BA (1982) Effect of nitroglycerin and nifedipine on subendocardial perfu-
sion in the presence of a flow-limiting coronary stenosis in the awoke dog. Circ Res 50: 678-687
12. Bevan JA (1982) Selective action of diltiazem on cerebral vascular smooth muscle in the rabbit:
antagonism of extrinsic but not intrinsic maintained tone. Am J Cardiol49: 519-524
13. Bevan JA, Hwa JJ (1985) Myogenic tone and cerebral vascular autoregulation: The role of a
stretch-dependent mechanism. Ann Biomed Eng 13:281-286
14. Bossaller C, Habib GB, Yamamoto H, Williams C, Wells S, Henry PD (1987) Impaired mus-
carinic endothelium-dpendent relaxation and cyclic guanosine 5' monophosphate formation in
atherosclerotic human coronary artery and rabbit aorta. J Clin Invest 79: 170-174
15. Brandt L, Andersson KE, Edvinsson L, Ljunggren B (1981) Effects of extracellular calcium
and calcium antagonists on the contractile responses of isolated human pial and mesenteric ar-
teries. J Cereb Blood Aow Metab 1: 339-347
16. Brandt L, Ljunggren B, Andersson KE, MacKenzie ET, Tamura A, Teasdale G (1982) Effects
on feline cortical pial microvasculature of topical application of a calcium antagonist (nifedipine)
under normal conditions and in focal ischemia. J Cereb Blood Flow Metab 3: 44-50
17. Chierchia S, Davies G, Berkenboom G, Crea F, Crean P, Maseri A (1984) a-Adrenergic recep-
tors and coronary spasm: an elusive link. Circulation 69:8-14
18. Chierchia S, de Caterina R, Crea F, Patrono C, Maseri A (1982) Failure of thromboxane A2
blockade to prevent attacks of vasospastic angina. Circulation 66: 702-705
19. Conti CR (1985) Large vessel coronary vasospasm: diagnosis, natural history and treatment.
Am J Cardiol 55 : 41B-49B
20. De Caterina R, Carpeggiani C, L'Abbate A (1984) A double blind, placebo-controlled study
of ketanserin in patients with Prinzmetal's angina. Evidence against a role for serotonin in the
genesis of coronary vasospasm. Circulation 69: 889-894
21. Disney L, Weir B (1987) Nimodipine treatment in poor grade patients. Results of a multi-
center, double-blind, placebo controlled study. In: Kassell N (ed) Proceedings of cerebral vaso-
spasm - a research conference. Williams and Wilkins, Baltimore (in press)
22. Edvinsson L, Brandt L, Andersson KE, Bengtsson B (1979) Effect of a calcium antagonist on
experimental constriction of human brain vessels: possible efficacy in cerebrovascular spasm.
Surg Neurolll :327-330
23. Ekstrom-Jodal B (1970) Effect of increased venous pressure on cerebral blood flow in dogs.
Acta Physiol Scand [Suppl] 350: 51-61
24. Endo M, Hirosawa K, Kaneko N, Hase K, Inone Y, Konno S (1976) Prinzmetal's variant
angina. Coronary arteriogram and left ventriculogram during angina induced by methacholine.
N Engl J Med 294: 252-255
25. Feigl EO (1983) Coronary physiology. Physiol Rev 63: 1-205
26. Fisher CM, Roberson GH, Ojeman RG (1977) Cerebral vasospasm with ruptured saccular
aneurysm - the clinical manifestations. Neurosurgery 1: 245-248
27. Aaim SF (1982) Comparative pharmacology of calcium blockers based on studies of vascular
smooth muscle. In: Flaim SF, Zelis R (eds) Calcium blockers. Mechanism of action and clinical
applications. Urban and Schwarzenberg, Baltimore, pp 155-178
28. Flaim SF, Kanda K (1982) Comparative pharmacology of calcium blockers based on studies on
cardiac output distribution. In: Aaim SF, Zelis R (eds) Calcium blockers. Mechanisms of action
and clinical applications. Urban and Schwarzenberg, Baltimore, pp 155-178
29. Aeckenstein A (1983) Calcium antagonism, a basic principle of drug-induced smooth muscle
relaxation. In: Aeckenstein A (ed) Calcium antagonism in heart and smooth muscle. Experi-
mental facts and therapeutic prospects. Wiley, London, pp 209-285
30. Fujiwara S, Ito Y, Itoh T, Kuriyama H, Suzuki H (1982) Diltiazem-induced vasodilatation of
smooth muscle cells of the canine basilar artery. Br J Pharmacol 75: 455-467
Spasm of Cerebral and Coronary Vessels 15

31. Furchgott RF (1984) The role of endothelium in the responses of vascular smooth muscle to
drugs. Annu Rev Pharmacol Toxicol 24: 175-197
32. Forstermann U, Mugge A, Frolich JC (1986) Endothelium-dependent relaxation of human epi-
cardial coronary arteries: frequent lack of effect of acetylcholine. Eur J Pharmacol128: 277-281
33. Ganz P, Alexander RW (1985) New insights into the cellular mechanism of vasospasm. Am J
Cardiol 56: 11E-15E
34. Ginsburg R (1983) The isolated human epicardial coronary artery. Am J Cardiol 52: 61A-66A
35. Ginsburg RC (1984) Myogenic tone of the isolated human epicardial artery: regulatory con-
trols. Acta Med Scand [Suppl) 694 : 29-37
36. Ginsburg R, Bristwo MR, Davis K, Dibiase A, Billingham ME (1984) Quantitative pharma-
cologic responses of normal and atherosclerotic isolated human epicardial coronary arteries.
Circulation 69: 430-440
37. Ginsburg R, Bristow MR, Stinson EB, Harrison DC (1980) Histamine receptors in the human
heart. Life Sci 26: 2245-2249
38. Ginsburg R, Bristow MR, Kantrowitz N, Bairn DS, Harrison DC (1981) Histamine provoca-
tion of clinical coronary artery spasm: implications concerning pathogenesis of variant angina
pectoris. Am Heart J 102: 819-822
39. Gisvold SE, Steen PA (1985) Drug therapy in brain ischaemia. Br J Anaesth 57: 96-109
40. Godfraind T, Finet M, Lima JS, Miller RC (1984) Contractile activity of human coronary
arteries and human myocardium in vitro and their sensitivity to calcium entry blockade by
nifedipine. J Pharmacol Exp Ther 230: 514-518
41. Godfraind T, Miller RC (1983) Effects of histamine and histamine antagonists mepyramine and
cimetidine on human coronary arteries in vitro. Br J Pharmacol 79: 979-984
42. Gotoh 0, Mohamed AA, McCulloch J, Graham DJ, Harper AM, Teasdale GM (1986)
Nimodipine and the haemodynamic and histopathologic consequences of middle cerebral artery
occlusion in the rat. J Cereb Blood Flow Metab 6: 321-331
43. Grotenhuis JA, Bettag W, Fiebach BJD, Dabir K (1984) Intracarotid slow bolus injection of
nimodipine during angiography for treatment of cerebral vasospasm after SAH. J Neurosurg
61:231-240
44. Halpern W, Osol G (1985) Influence of transmural pressure on myogenic responses of isolated
cerebral arteries of the rat. Ann Biomed Eng 13 :287-293
45. Harder DR (1980) Comparison of electrical properties of middle cerebral and mesenteric artery
in cat. Am J Physiol 23: C23-C26
46. Harder DR (1983) Heterogeneity of membrane properties in vascular muscle cells from various
vascular beds. Fed Proc 42: 253-256
47. Harder DR (1984) Pressure-dependent membrane depolarization in cat middle cerebral artery.
Circ Res 55: 197-202
48. Harder DR, Abel PW, Hermsmeyer K (1981) Membrane electrical mechanisms of basilar
artery constriction and pial artery dilation by norepinephrine. Circ Res 49: 1237-1242
49. Harris RJ, Bronston NM, Symon L, Bayhan M, Watson A (1982) The effects of a calcium-
antagonist, nimodipine, upon physiological responses of the cerebral vasculature and its pos-
sible influence upon focal cerebral ischaemia. Stroke 13: 759-766
50. Haws CW, Gourley JK, Heistad DD (1983) Effects of nimodipine on cerebral blood flow. J
Pharmacol Exp Ther 225: 24-28
51. Hayashi S, Toda N (1977) Inhibition by Cd2+, verapamil, and papaverine of Ca2+ -induced
contractions in isolated cerebral and peripheral arteries of the dog. Br J Pharmacol 60: 35-43
52. Heros RC, Zervas NT, Varsos V (1983) Cerebral vasospasm after subarachnoid hemorrhage:
an update. Ann NeuroI14:599-608
53. Hof RP (1983) Calcium antagonists and the peripheral circulation: differences and similarities
between PY 108-068, nicardipine, verapamil and diltiazem. Br J Pharmacol 78: 375-394
54. Holtz J, Giesler M, Bassenge E (1983) Two dilatory mechanisms of antianginal drugs on epi-
cardial coronary arteries in vivo: indirect, flow-dependent, endothelium-mediated dilatation
and direct smooth muscle relaxation. Z Kardiol [Suppl 3) 72: 98-106
55. Hossman KA, Sato K (1970) Recovery of neuronal function after prolonged cerebral ischemia.
Science 168: 375-376
56. Hogestatt ED, Andersson KE, Edvinsson L (1982) Effects of nifedipine on potassium-induced
contraction and noradrenaline release in cerebral and extracranial arteries from rabbit. Acta
Physiol Scand 114: 283-296
16 K.-E. Andersson et al.

57. Iseri LT, French JH (1984) Magnesium: Nature's physiologic calcium blocker. Am Heart J 108:
188-193
58. Jayakody KL, Kappagoda cr, Senaratne MPJ, Sreeharan N (1987) Absence of effect of cal-
cium antagonists on endothelium-dependent relaxation in 'rabbit aorta. Br J Pharmacol 91:
155-164
59. Kalsner S (1985) Coronary artery reactivity in human vessels: some questions and some answers.
Fed Proc 44: 321-325
60. Kalsner S, Richards R (1984) Coronary arteries of cardiac patients are hyperreactive and con-
tain stores of amines: a mechanism for coronary spasm. Science 223: 1435-1437
61. Kapp JP, Neill WR, Neill CL, Hodges LR, Smith RR (1982) The three phases of vasospasm.
Surg Neurol 18: 40-45
62. Karashima T, Kuriyama H (1981) Electrical properties of smooth muscle cell membrane and
neuromuscular transmission in the guinea-pig basilar artery. Br J Pharmacol 74: 495-504
63. Kassel NF, Sasaki T, Colohan ART, Nazar G (1985) Cerebral vasospasm following aneurysmal
subarachnoid hemorrhage. Stroke 16: 562-572
64. Kawasaki K, Seki K, Hosoda S (1981) Spontaneous rhythmic contractions in isolated human
coronary arteries. Experientia 37: 1291-1292
65. Kazda S, Garthoff B, Krause HP, Schlossman (1982) Cerebrovascular effects of the calcium
antagonistic dihydropyridine derivative nimodipine in animal experiments. Arzneimittelfor-
schung (Drug Res) 32:331-338
66. Kazda S, Mayer D (1985) Postischemic impaired reperfusion and tissue damage: consequences
of a calcium dependent vasospasm. In: Godfraind T, Vanhoutte PM, Govoni S, Paoletti R
(eds) Calcium entry blockers and tissue protection. Raven, New York, pp 129-149
67. Kazner E, Sprung CH, Adelt D, Ammerer HP, Karnick R, Baumann H, Boker DK, Groten-
huis JA, Jaksche H, Istaitih AR, Klein HJ, Langelaar G, Rusegger L, Sachfenheimer W,
Schackert G, Schramm J (1985) Clinical experience with nimodipine in the prophylaxis of
neurological deficits after subarachnoid hemorrhage. Neurochirurgia 28: 110-113
68. Koos WT, Pemeczky A, Auer LM, Boker DK, Gaab M, Jaksche H, Kostron H, Meinig G,
Muizelaar JP, van der Werff AJM, Seibert HK, Ulrich F, Sprung CH (1985) Nimodipine treat-
ment of ischemic neurologic deficits due to cerebral vasospasm after subarachnoid hemorrhage.
Neurochirurgia 28 : 114-117
69. Ku DD, Ann HS (1987) Magnesium deficiency produces endothelium-dependent vasorelaxa-
tion in canine coronary arteries. J Pharmacol Exp Ther 241 :961-966
70. Ljunggren B, Brandt L, Sliveland H, Nilsson PE, Cronquist S, Andersson KE, Vinge E (1984)
Outcome in 60 consecutive patients treated with early aneurysm operation and intravenous
nimodipine. J Neurosurg 61: 864-873
71. Ljunggren B, Sliveland H, Brandt L (1983) Causes of unfavorable outcome after early aneurysm
operation. Neurosurgery 13: 629-633
72. Ljunggren B, Sliveland H, Brandt L, Romner B, Andersson KE (1987) Nimodipine in aneurys-
mal subarachnoid hemorrhage. In: Kassell N (ed) Proceedings of cerebral vasospasm - a re-
search conference. Williams and Wilkins, Baltimore (in press)
73. Lusamvuku NAT, Sercombe R, Aubineau P, Seylaz J (1979) Correlated electrical and mechan-
ical response of isolated rabbit pial arteries to some vasoactive drugs. Stroke 10: 727-732
74. Macho P, Hintze TH, Vatner SF (1981) Regulation of large coronary arteries by increases in
myocardial metabolic demands in conscious dogs. Circ Res 49: 594-599
75. Maseri A (1983) The changing face of angina pectoris: practical implications. Lancet 1: 746-749
76. Maseri A (1987) Role of coronary artery spasm in symptomatic and silent myocardial ischemia.
JAm Coll Cardiol9:249-262
77. Maseri A, Chierchia S (1982) Coronary artery spasm - demonstration, definition, diagnosis
and consequences. Prog Cardiovasc Dis 25 : 169-192
78. McCalden TA, Bevan JA (1981) Sources of activator calcium in rabbit basilar artery. Am J
Physiol241: HI29-H133
79. McCalden TA, Nath RG, Thiele K (1984) The effects of a calcium antagonist (nimodipine) on
basal cerebral blood flow and reactivity to various agonists. Stroke 15: 527-530
80. Mee EW, Dorrance DE, Low D, Neil-Dwyer G (1986) Cerebral blood flow and neurological
outcome: a controlled study of nimodipine in patients with subarachnoid hemorrhage. J Neurol
Neurosurg Psychiatry 49: 469
Spasm of Cerebral and Coronary Vessels 17

81. Meller H, Pichard A, Dack S (1976) Coronary arterial spasm in Prinzmetals angina: a proven
hypothesis. Am J Cardiol 37: 938-940
82. Milde LN, Milde JH, Michenfelder JD (1986) Delayed treatment with nimodipine improves
cerebral blood flow after complete cerebral ischemia in the dog. J Cereb Blood Flow Metab 6:
332-337
83. Miller D, Waters DD, Warmica W, Szlachcic J, Kreeft J, Theroux P (1981) Is variant angina
the coronary manifestation of a generalized vasospastic disorder. N Engl J Med 304: 763-766
84. Mizukami M, Takemae T, Tazawa T, Kawase T, Matsuzaki T (1980) Value of computer
tomography in the prediction of cerebral vasospasm after aneurysmal rupture. Neurosurgery
7:583-586
85. Moreland RS, Bohr DF (1983) Adrenergic control of coronary arteries. Fed Proc 43: 2857-2861
86. Miiller-Schweinitzer EM (1980) The mechanism of ergometrine-induced coronary arterial
spasm: In vitro studies on canine arteries. J Cardiovasc Pharmacol 2: 645-655
87. Miiller-Schweinitzer E, Neumann P (1983) In vitro effects of calcium antagonists PN 200-110,
nifedipine, and nimodipine on human and canine cerebral arteries. J Cereb Blood Flow Metab
3:354-361
88. Nakayama K (1982) Calcium-dependent contractile activation of artery produced by quick
stretch. Am J Physiol 242: H760-H768
89. Nakayama K, Suzuki S, Sugi H (1986) Physiological and ultrastructural studies on the mecha-
nisms of stretch-induced contractile activation in rabbit cerebral artery smooth muscle. Jpn J
Physiol36: 745-760
90. Neil-Dwyer GA (1985) A controlled study of nimodipine in subarachnoid hemorrhage patients.
Proceeding of 13th World congress of neurology, September 1-6, 1985, Hamburg
91. Newberg LA, Steen PA, Milde JH, Michenfelder JD (1984) Failure of flunarizine to improve
cerebral blood flow or neurological recovery in a canine model of complete cerebral ischemia.
Stroke 15: 666-671
92. Olsson RA, Bugni WJ (1986) Coronary circulation. In: Fozzard HA, Haber E, Jennings RB,
Katz AM, Morgan HE (eds) The heart and cardiovascular system. Raven, New York, pp 987-
1037
93. Osol G, Halpern W (1985) Myogenic properties of cerebral blood vessels from normotensive
and hypertensive rats. Am J PhysioI249:H914-H921
94. Philippon J, Grob R, Dagreou F, Guggiari M, Rivierez M, Viars P (1986) Prevention of vaso-
spasm in subarachnoid hemorrhage. A controlled study with nimodipine. Acta Neurochir 82:
110-114
95. Phillis JW, Delong RE, Towner JK (1985) The effects of lidoflazine and flunarizine on cerebral
reactive hyperemia. Eur J Pharmacol 112: 329
96. Raizner AE, Chahine RA, Ishimori T, Verani MS, Zacca N, Jamal N, Miller RR, Luchi RJ
(1980) Provocation of coronary artery spasm by the cold pressor test. Hemodynamic, arterio-
graphic and quantitative angiographic observations. Circulation 62: 925-932
97. Robertson RM, Robertson D, Roberts U (1981) Thromboxane A2 in vasotonic angina pec-
toris. Evidence from direct measurements and inhibitor trials. N Engl J Med 304: 998-1003
98. Roine RO, Kaste M, Kinnunen A, Nikki P (1987) Safety and efficacy of nimodipine in resus-
citation of patients outside hospital. Br Med J 294: 20
99. Ross G, Stinson E, Schroeder J, Ginsburg R (1980) Spontaneous phasic activity of isolated
human coronary arteries. Cardiovasc Res 14: 613-618
100. Sahlin C, Brismar J, Delgado T, Owman C, Salford LG, Svendgaard NA (1987) Cerebrovascu-
lar and metabolic changes during the delayed vasospasm following experimental subarachnoid
hemorrhage in baboons, and treatment with a calcium antagonist. Brain Res 403: 313-332
101. Salaices M, Maria J, Rico ML, Gonzalez C (1983) Effects of verapamil and manganese on the
vasoconstrictor responses to noradrenaline, serotonin and potassium in human and goat cere-
bral arteries. Biochem PharmacoI32:2711-2714
102. Sasaki T, Kassel NF, Zuccarello M (1986) Dependence of cerebral arterial contractions on
intracellularly stored Ca + +. Stroke 17: 95-97
103. Schwartz JS, Bache RJ (1987) Pharmacologic vasodilators in the coronary circulation. Circula-
tion [Suppl 1] 75: 162-167
104. Sercombe R, Verrechia C, Oudort N, Dimitriadon V, Seylaz J (1985) Pial artery responses to
norepinephrine potentiated by endothelium removal. J Cereb Blood Flow Metab 5: 312-317
18 K.-E. Andersson et al.

105. Shepherd JT, Vanhoutte PM (1986) Mechanisms responsible for coronary vasospasm. JAm
Coli Cardiol 8: 50A-54A
106. Shimizu K, Ohta T, Toda N (1980) Evidence for greater susceptibility of isolated dog cerebral
arteries to Ca antagonists than peripheral arteries. Stroke 11 : 261-265
107. Siesjo BK (1984) Cerebral circulation and metabolism. J Neurosurg 60:883-908
108. Singer HA, Peach MJ (1982) Calcium- and endothelial-mediated vascular smooth muscle
relaxation in rabbit aorta. Hypertension [SuppI2] 4: 19-25
109. Skiirpy T, Hogestatt ED, Andersson KE (1984) Influence of extracellular calcium and nifedipine
on al- and aZ-adrenoceptor mediated contractile responses in isolated rat and cat cerebral and
mesenteric arteries. Acta Physiol Scand 123: 445-456
110. Steen PA, Newberg LA, Milde JH, Michenfelder JD (1983) Nimodipine improves cerebral
blood flow and neurologic recovery after complete cerebral ischemia in the dogs. J Cereb Blood
Flow Metab 3: 38-43
111. Stone HL (1983) Control of the coronary circulation during exercise. Ann Rev Physiol 45:
213-227
112. Stone PH (1987) Calcium antagonist for Prinzmetal's variant angina, unstable angina and silent
myocardial ischemia: Therapeutic tool and probe for identification of pathophysiologic mecha-
nisms. Am J Cardiol 59: 101B-115B
113. Sundt TM Jr, Szurszewski J, Sharbrough FW (1977) Physiological considerations important for
the management of vasospasm. Surg Neurol 7: 259-267
114. Svendgaard NA, Edvinsson L, Owman C, Sahlin C (1977) Increased sensitivity of the basilar
artery to norepinephrine and 5-hydroxytryptamine following experimental subarachnoid hemor-
rhage. Surg Neurol8: 191-195
115. Teasdale G, Legrain Y, Mackenzie E, Graham DF (1983) Potassium release and vascular
events in focal cerebral ischemia. J Cereb Blood Flow Metab [Suppll] 3: 395-396
116. Thayer SA, Murphy SN, Miller RJ (1986) Widespread distribution of dihydropyridine sensitive
calcium channels in the central nervous system. Mol Pharmacol 30: 505-509
117. Theroux P, Waters DD (1983) Calcium channel blocking agents in the therapy of vasospastic
angina. In: Stone PH, Antman EM (eds) Calcium channel blocking agents in the treatment of
cardiovascular disorders. Futura, Mount Kisco, New York, pp 269-319
118. Toda N (1983) Isolated human coronary arteries in response to vasoconstrictor substances. Am
J Physiol 245 : H937 - H941
119. Toda N (1984) Responses of human, monkey and dog coronary arteries in vitro to carbocyclic
thromboxane A2 and vasodilators. Br J Pharmacol 83: 399-408
120. Toda N (1985) Reactivity in human cerebral artery: species variation. Fed Proc 44: 326-330
121. Toda N (1986) a-Adrenoceptor subtypes and diltiazem actions in isolated human coronary
arteries. Am J Physiol 250: H718-H724
122. Toda N (1987) Is histamine a human coronary vasospastic substance? Trends Pharmacol Sci 8:
289-290
123. Towart R (1981) The selective inhibition of serotonin-induced contractions of rabbit cerebral
vascular smooth muscle by calcium antagonistic dihydropyridines. An investigation of the
mechanism of action of nimodipine. Circ Res 48: 650-657
124. Towart R, Wehinger E, Meyer H, Kazda S (1982) The effects ofnimodipine, its optical isomers
and metabolites on isolated vascular smooth muscle. Arzneimittelforschung (Drug Res) 32:
338-346
125. Turlapaty PDMV, Weiner R, Altura BM (1981) Interactions of magnesium and verapamil on
tone and contractility of vascular smooth muscle. Eur J Pharmacol 74: 263-272
126. Tzivoni D, Kerm A, Benhorin J, Gottlieb S, Atlas D, Stem S (1983) Prazosin therapy for
refractory variant angina. Am Heart J 105: 262-266
127. Uski T, Andersson KE (1984) Effects of prostanoids on isolated feline cerebral arteries. II.
Roles of extra- and intracellular calcium for the prostaglandin F 2a-induced contraction. Acta
Physiol Scand 120: 197-205
128. van Breemen C, Siegel B (1980) The mechanism of a-adrenergic activation of the dog coronary
artery. Circ Res 46 : 426-429
129. Vanhoutte PM (1986) Could the absence or malfunction of vascular endothelium precipitate
the occurrence of vasospasm? J Mol Cell Cardiol 18: 679-689
130. Vanhoutte PM (1987) The expert committee of the world health organization on classification
of calcium antagonists: the view point of the raporteur. Am J Cardiol59: 3A-8A
Spasm of Cerebral and Coronary Vessels 19

131. Vedernikov yP (1986) Mechanisms of coronary spasm of isolated human epicardial coronary
segments excised 3 to 5 hours after sudden death. J Am Coli Cardiol 8: 42A-49A
132. Vibulsresth S, Dietrich WD, Busto R, Ginsberg MD (1987) Failure of nimodipine to prevent
ischemic neuronal damage in rats. Stroke 18: 210-216
133. Wei EP, Kontos HA (1982) Responses of cerebral arterioles to increased venous pressure. Am
J Physiol243: H442-H447
134. Weinheimer G, Golenhofen K, Mandrek K (1983) Comparison of the inhibitory effects of
nifedipine, nitroglycerin, and nitroprusside sodium on different types of activation in canine
coronary arteries, with comparative studies in human coronary arteries. J Cardiovasc Pharma-
col 5 : 989-997
135. Wellum GR, Peterson JW, Zervas NT (1985) The relevance of in vitro smooth muscle experi-
ments to cerebral vasospasm. Stroke 16: 573-581
136. White RP (1979) Multiplex origins of cerebral vasospasm. In: Price TR, Nelson E (eds) Cere-
brovascular diseases. Raven, New York, pp 307-319
137. Wijdicks EFM, Vermeulen M, van Brummelen P, denBoer NC, van Gijn J (1987) Digoxin-like
immunoreactive substance in patients with aneurysmal subarachnoid haemorrhage. Br Med J
294: 729-732
138. Wilkins RH (1986) Attempts at prevention or treatment of intracranial arterial spasm: an up-
date. Neurosurgery 18: 808-825
139. Winniford MD, Filipchuk N, Hillis LD (1983) Alpha-adrenergic blockade for variant angina: a
long-term, double-blind, randomized trial. Circulation 67: 1185-1188
140. Yamamoto M, Ohta T, Toda N (1983) Mechanisms of relaxant action of nicardipine, a new
Ca++-antagonist, on isolated dog cerebral and mesenteric arteries. Stroke 14:270-275
141. Yasue H (1984) Coronary artery spasm and calcium ions. In: Opie LH (ed) Calcium antagonists
and cardiovascular disease. Raven, New York, pp 117-128
142. Yasue H, Horio Y, Nakamura N, Fuji H, Imoto H, Sonoda R, Kugiyama K, Obata K,
Morikami Y, Kimura T (1986) Induction of coronary artery spasm by acetylcholine in patients
with variant angina: possible role of the parasympathic nervous system in the pathogenesis of
coronary artery spasm. Circulation 74: 955-963
143. Yasue H, Nagao M, Omote S, Takizawa A, Miwa K, Tanaku S (1978) Coronary arterial spasm
and Prinzmetal's variant form of angina induced by hyperventilation and Tris buffer infusion.
Circulation 58 : 56-62
144. Yasue H, Touyama M, Shimamoto M, Kato H, Tanaka S, Akiyama F (1974) Role of autonomic
nervous system in the pathogenesis of Prinzmetal's variant form of angina pectoris. Circulation
50:534-539
Chapter 2
Opioid Peptides and the Cardiovascular System
with Especial Reference to Low Perfusion States
J. R. P ARRATT

The purpose of this brief review is two-fold. First, in line with the rationale behind
the Tromso Symposia, it is designed to introduce the general reader to the role of
opioid peptides and their receptors in the cardiovascular system. Second, it concen-
trates in more depth on two aspects of cardiovascular stress and low tissue perfusion,
shock (and trauma) and myocardial ischaemia, where endogenous opioid peptides
may be especially important and where drugs which interact with their receptors
have therapeutic potential.

Opioid Peptides

What They Are

Three main families of peptides are involved in the endogenous opioid system and all
of these influence the cardiovascular system.

Enkephalins. These pentapeptides (leu-enkephalin, Tyr-Gly-Gly-Phe-Leu and met-

enkephalin, Tyr-Gly-Gly-Phe-Met) are derived from pre-proenkephalin A, are
localised, together with their receptors, in neurones throughout the brain and spinal
cord but especially in the hypothalamus, in the brain stem nuclei, the limbic fore-
brain and in the fibres of the sympathetic lateral columns, where they are associated
with adrenaline and noradrenaline. Outside the central nervous system enkephalins
are found in sympathetic ganglia, the myenteric plexus and in the chromaffin cells of
the adrenal medulla where they are co-stored with, and released with, catechol-
amines. This is probably the main source of enkephalins in the plasma although
release from the pituitary may also contribute [8]. In the heart and blood vessels
enkephalins are also mainly associated with neurones; Table 1 summarises the dis-
tribution of both opioid peptides (mainly enkephalins and dynorphins) and their re-
ceptors in the cardiovascular system.

Dynorphins. Are large peptides also derived from pre-proenkephalin. Dynorphin

has 17 amino acids, the sequence of the first five of which is the same as for leu-
enkephalin; a- and jJ-neoendorphin have ten and nine amino acids respectively, and
again the first five are identical with those of leu-enkephalin. Their distribution is
similar to that of leu-enkephalin and they are found in highest amounts in the hypo-
thalamus, the medulla and, in association with vasopressin, in the posterior pituitary.
Opioid Peptides and the Cardiovascular System 21

Table 1. Distribution of opioid peptides and their receptors in the cardiovascular system

Adrenal medulla Mu, kappa, delta and sigma receptors identified by ligand binding in mem-
branes (catecholamine containing cells? other cells?) [15]
Opioid peptides (especially dynorphin) present in adrenal medulla

Cardiac branches of the vagus and sympathetic nerves, cardiac chromaffin cells and superior cervical
ganglion. 'A dynorphinergic sympathetic innervation of atria and ventricles seems conceivable' [86]

Central cardiovascular regulatory centres. Predominance of mu sites in the NTS [29]

Heart Enkephalin-immunoreactive nerve fibres present (e.g., in cardiac ganglia,

SA junction)
Some fibres in the adventitia of coronary arteries (proximal arteries) and
their branches (sparse) and in paracoronary chemoreceptor structures.
Targets seem to be exclusively neural [85]
Low specific binding of 3H-bremazocine and diprenorphine [48] in car-
diac homogenates did not allow re type characterisation, specific binding
(of the order of 30%) of 3H-diprenorphine in rat hearts, unaffected by
ischaemia [77]
Immunoreactive dynorphin, (Met) and (Leu) enkephalins detected in heart
[80]. Met-enkephalin could not be extracted from ventricles [86]
7-9 peptides (derived from proenkephalin and prodynorphin) isolated from
guinea-pig hearts; 22 pmol/g wet weight and 11 pmol/g wet weight for atria
and ventricles respectively [86]
Presynaptic myocardial opioid receptors (mu?) demonstrated in rabbit
papillary muscle [73]; presynaptic opioid K-receptors (but not mu or delta)
in sinus node of the same species [82]

Dynorphin probably acts on mu and kappa receptors whilst p-neoendorphin acts

mainly on delta receptors.

Endorphins. These peptides are derived from pre-proopioidmelanocortin and have

variable amino acid sequences. Thus, p-endorphin (P-E) is a peptide with 31 amino
acid residues which has potent opioid-receptor binding and stimulatory effects; a-, r-
and 8-endorphins have similar, shorter, amino acid sequences (1-16, 1-17 and 1-27
respectively). p-Endorphin is found in the pituitary gland (anterior and intermediate
regions), in the hypothalamus and in certain neurones (especially in the medulla and
midbrain). It is possible that P-E may function as a neurohormone. It acts on mu and
delta opioid receptors but not on kappa receptors.

What They Do - Effects on Cardiovascular Function

A number of factors modify the cardiovascular effects of these peptide agonists.

These factors include the species, the presence of an anaesthetic and, in conscious
animals, the level of restraint, the dose and route (site) of administration (systemic,
central), the specificity of the agonist for particular receptor types, the status of the
animal (especially underlying trauma) and the relative dominance of other secondary
22 J. R. Parratt

Table 2. A classification of opioid receptors

Agonists Antagonists

Mu fj:-Endorphin; morphine, Naloxone; naltrexone;

BW 768 C; normorphine, {J-funaltrexamine ({J-FNA);
morphiceptin; sufestanyl; naloxone; WIN 44,441-3
meptazinol a D-Ala2 MePhe4 ,
Glf-ol-enkephalin (DAGO)
Delta Leu-enkephalin; D-Ala2 Naloxone; M 154,129;
D-Ieus enkephalin. (DADLE); ICI 174684; ICI 154129
D-Tbr, Leus-Thii enkephalin J-7747; C-7000b
(DTLET)a
Kappa Dynorphin (1-13,1-17,1-8); Mr 1452; Mr 2266 b
(benzomorphan receptor) U 50488; (ethyl)ketocyclazocine
(EKC); Mr 2034; trlfluadom
Sigma SKF10,047

a Also has some antagonist activity; b also has some agonist activity.

effects such as catecholamine and histamine release, effects on baroreceptors and on

pulmonary J receptors. In this area generalisations are difficult, and perhaps the
simplest classification is that based on the central and systemic administration of a
variety of reasonably specific opioid receptor agonists.
Much depends on the receptor type involved. Opioid receptors consist of a recog-
nition site, to which the drug binds, together with a factor that translates the binding
into a particular biochemical response. Three, possibly four, receptor subtypes have
been described. These are mu, delta, kappa and possibly sigma. Table 2 outlines the
presently known subtypes of opioid receptors and those agonists and antagonists that
interact with them. However, there is often an overlap between the activities of these
agonists and antagonists at different receptors (cross-reactivity) and also there is
often a poor correlation between relative binding affinities and the appropriate phar-
macological response. This is partly because peripheral tissues, which are often used
to evaluate these responses in bioassay studies, do not contain homogeneous opioid
receptor populations.
As with other receptors, this classification is based on both pharmacological,
functional responses and ligand-binding techniques to identify stereospecific opioid-
binding sites. The kinetic parameters of the interaction of a labelled opioid and the
binding site can be determined directly, whilst the affinity of unlabelled opioids may
be estimated indirectly from their ability to displace the binding of labelled ligands.
This subject has been extensively reviewed [38, 58, 61, 84].
Tables 3 and 4 summarise the cardiovascular effects of opioid receptor agonists
when given directly into the central nervous system or when administered systemi-
cally. Usually, administration of small amounts of mu or delta agonists in conscious
animals results in increases in arterial pressure and in heart rate, whereas large doses
of either agonist, especially in anaesthetised animals, result in systemic hypotension,
bradycardia and eventually cardiovascular collapse. There are, however, many dif-
ficulties in interpreting these results. A good example involves the precise location
Opioid Peptides and the Cardiovascular System 23

Table 3. Cardiovascular effects following administration of opioid receptor agonists into the central
nervous system"

Receptor Agonist action

Mu Hypotension and bradycardia (sometimes preceded by a pres-

(e.g., morphine, morphiceptin, sor response) when injected into NTS or into fourth ventricle
BW 768C, DAGO or dermor- (large doses) [34,37]
phine) Pressor responses (with tachycardia) after injections into the
third ventricle or after application of small doses to the NA or
NTS [26, 28] and to the hypothalamus paraventricular (PV)
and medial preoptic (MP) nuclei [14, 22, 64, 65]

Delta (i) Anaesthetised animals: hypotension and bradycardia (topi-

(e.g., DADLE) cal application to ventral surface of medulla or brain-stem or
into third ventricle) [37]
(ii) Conscious or lightly anaesthetised animals: hypertension,
tachycardia and potentiation of the carotid occlusion reflex
when injected into the NTS, lateral ventricles or by intra-
cisternal or intracerebroventricular administration. It seems
that pressor responses (blocked by naloxone or phentolamine)
occur following administration into ventrolateral vasodepres-
sor areas and depressor responses (with bradycardia) when in-
jected into the pressor area of the rostral ventrolateral medulla
(again reversed by naloxone) [34, 37]

Kappa Microinjections into PV and MP nuclei or into the NA and

(e.g., dynorphin (A species) NTS decrease pressure (sometimes preceded by an increase)
and benzomorphans, e.g., and heart rate [23, 27]. Marked hypotension in vasopressin
bremazocine) deficient (rats)

" These depend on species, doses, route and site of administration, presence or absence of anaes-
thetics (in general, anaesthesia blocks pressor responses of injected opiates and enhances their de-
pressive effects) and, in conscious animals, the degree of restraint. Results of earlier studies have
been previously reviewed [34].

of the central sites containing various opioid receptors. Thus, in the hypothalamus,
pressor and depressor sites for morphine and D-Ala2 , D-Leu5-enkephalin (DADLE)
have been found in neighbouring nuclei less than 1 mm apart [22]. Further, com-
pletely opposite effects can be obtained when the enkephalin analogue D-Ala2-Met3
enkephalin is injected into the pressor area of the rostral-ventrolateral medulla
(hypotension and bradycardia) compared with injections into the ventrolateral vaso-
depressor areas (hypertension and tachycardia) [87].
The effects of the systemic administration of opioid peptides are again dependent
on dose, species and the extent of previous surgery. These effects are summarised in
Table 4. In general, stimulation of mu receptors (e.g., by morphine or morphiceptin)
leads to bradycardia and to peripheral vasodilatation. It is still not clear in which vas-
cular beds blood flow is increased. The studies with morphine, for example, are com-
plicated by effects other than those of mu receptor stimulation, such as histamine
and catecholamine release and, in higher doses, by effects on other opioid receptors
(e.g. delta receptors). The same applies to cardiac effects ofmu receptor stimulation
in rat isolated atria; morphine has no significant electrophysiological effects (electri-
24 J. R. Parratt

Table 4. Cardiovascular effects following the systemic administration of fairly "selective" opioid
receptor agonists

Receptor Agonist action

Mu No decrease in sympathetic transmitter release (?) [44), vasodilation and bradycardia,

e.g. [3, 6, 45)
Increased cerebral blood flow, reduced duodenal blood flow [46)
Little effect on coronary blood flow [46)
Decrease in action potential duration, increased K+ conductance [59), depressed Na+
influx
Delta Inhibition of noradrenaline release and of sympathetic vasoconstriction [44)
Reduced blood pressure and heart rate [11, 45, 71); attenuated carotid occlusion re-
sponse [45)
Inhibition of calcium flux in cardiac muscle (atria) and attenuation of positive chrono-
tropic responses [30)
Kappa Inhibition of noradrenaline release and of sympathetic vasoconstriction [44, 47)
Inhibition of catecholamine release from adrenal medullary chromaffin cells [52)
Reduced blood pressure and heart rate [79)
Relaxation of isolated (cerebral) arteries [2)
Sigma Contraction of isolated cerebral arteries [2)

ht
~ 10
<1 0
~

10 min

•
0.03
•
0.1
•
0.3
• •3 •
10
•
30 pmol/l llWlenk
Fig. I. Inhibition by [leu5) enkephalin of electrically evoked vasoconstriction. The perfused ileocolic
artery of the rabbit was stimulated every minute with 5 pulses at 10 Hz. Leu5-enkephalin was added
as indicated. L1mmHg is the electrically evoked increase in perfusion pressure [from 47)

cal threshold for excitation, effective refractory period, sinus node recovery time),
does not influence spontaneous frequency and only slightly decreases contractile
force [30]. Similar negative results have also been obtained in rabbits [82].
The cardiovascular effects of intravenous morphine are reduced by naloxone,
naloxazone and f3-flunaltrexamine (f3-FNA). Holaday [34] has concluded that brady-
cardia induced in conscious rats by morphine is mainly mediated by mu receptors
whereas both delta and mu receptors are involved in the peripheral vasodilatation.
Biphasic blood pressure responses are seen following the intravenous administration
of met- and leu-enkephalins [72, 75, 76].
The effects of delta receptor stimulation are less difficult to evaluate (Table 4).
There is good evidence for inhibition of noradrenaline release and therefore of sym-
pathetically mediated vasoconstriction. In the rabbit isolated ear and ileocolic artery
preparations, leu-enkephalin (a predominantly delta agonist; Table 2) inhibits the ef-
fects of field stimulation (Fig. 1) and of noradrenaline release, an effect prevented
by the selective delta antagonist leI 154129 [43, 44, 47]. In vivo, met-enkephalin
(a rather less specific delta agonist) decreases cardiac output, induces hypotension
Opioid Pep tides and the Cardiovascular System 25

A 10 min

20
mmHg
I
Stim.
3 10 30 100 300 ~ Dyn'1_U

B 10 min

m:Hg I
Stim .
.....,
-
100
3 10
100 100
,AIM NatOlI.
~ Dyn.1-13

Fig. 2A, B. Field stimulation-induced vasoconstriction in the perfused rabbit ear artery (constant
flow preparation; increase in perfusion pressure indicates vasoconstriction). A Effect of increasing
concentrations of dynorphinl_13 (Dyn.l_13). B Effect of naloxone (Nalox.) and its interaction with
dynorphinl_13. Stimulation (Stirn.) with 5 pulses (200mA, O.3ms) every minute at 5Hz. Dynorphin
reduced field stimulation-induced vasoconstriction, an effect abolished by naloxone. Note that in
this particular preparation naloxone alone increases perfusion pressure on field stimulation. This
might indicate endogenous opioid peptide release by electrial stimulation in this preparation [from 44]

Table 5. Cardiovascular effects of opiate antagonists, especially naloxone and naltrexone

No significant effects on heart rate and blood pressure [3, 6, 45] (except under conditions described
later, when there is evidence of substantial release of endogenous opioid peptides, e .g., shock)
No change in cerebral [46, 53], coronary [31], renal, skeletal muscle blood flow. Some evidence for
reduced spinal cord blood flow and cerebral (cortical) blood flow during cerebral ischaemia
Prevent fall in blood pressure that occurs during sleep in man [71]
Opiate antagonists markedly attenuate haemodynamic responses to morphine e.g., bradycardia (by
selective mu antagonists like ,B-FNA) and hypotension (by combined mu and delta antagonists) [34]

and bradycardia, and reduces in parallel the rate of sympathetic nerve discharge [68].
The main blood flow increase is in skeletal muscle beds [17].
Kappa receptor stimulation (e.g., by dynorphin 1-13) has similar effects to delta
receptor stimulation, i.e. inhibition of noradrenaline release and of sympathetically-
mediated vasoconstriction (Table 4; Fig. 2) [47].
Although opioid receptors are present in peripheral organs, the physiological
role of these receptors in modulating sympathetic tone remains unclear. Thus,
studies in the pithed rat, have failed to demonstrate cardiovascular responses to vari-
ous enkephalins selective for mu , delta or kappa receptors [16, 24] and, except under
conditions favouring endogenous opioid peptide release, opiate antagonists such as
naloxone have little effect on blood pressure or tissue blood flow (Table 5) and do
26 J. R. Parratt

not consistently increase neurogenic vasoconstriction or 3H noradrenaline release in

field stimulated isolated vessel preparations [44, 47] (but cf. Fig. 2).
Little has been done to systematically characterise the opioid receptors present in
smooth muscle vessels in different vascular beds. The most complete studies in iso-
lated blood vessels are those of Altura's group who have used mainly basilar and
middle cerebral arteries of dogs [2]. Although these arteries contract in response to
certain sigma receptor agonists (phencyclidine and its analogues) no evidence could
be found for mu or delta receptors mediating relaxation. Typical kappa receptor
ligands such as (ethyl)ketacyclazocine (EKC) and bremazocine (Table 2) did, how-
ever, relax precontracted cerebral arteries although other, so-called kappa agonists
(such as Mr 2034 and U 50488H; Table 2) contracted rather than relaxed these same
cerebral arteries [2].

Endogenous Opioid Peptides and Their Receptors in Shock and Trauma

One might expect that the release of endogenous opioid peptides in response to stress
would exert beneficial effects; for example, pain relief following traumatic injury. It
might seem surprising therefore that a concept has arisen suggesting that such a re-
lease of opioid peptides might have detrimental effects. This concept arose from the
observation that there are similarities between symptoms of overdosage with narcotic
drugs and those of circulatory shock and therefore that such opioid peptides, re-
leased during stress accompanying shock, might contribute to tissue hypoperfusion
and cardiovascular collapse.
This led Holaday and Faden in 1978 [36] to examine the effects of naloxone in ex-
perimental shock. The original studies used conscious rats (with indwelling arterial
and venous catheters) in which naloxone was administered shortly after an injection
of purified endotoxin derived from negative gram organisms (E. coli). Naloxone

Fig.3. The effects of naloxone (1.0mg/100g at 2h and then O.1mg/100g hourly thereafter) on
plasma-glucose levels in unanaesthetised rats with chronically implanted vascular catheters given
E. coli endotoxin. (e) Endotoxin (E) alone; (_) E + naloxone. The histograms give the number of
survivors at two hourly intervals (E alone (hatched columns); E + naloxone (open columns)) [from 1]
Opioid Peptides and the Cardiovascular System 27

Table 6. Inhibition by meptazinol (0.1 mg/100g at 2h, then O.OS mg/1oog hourly thereafter, i.v.) of
the delayed hypoglycaemia induced by E. coli endotoxin response in unanaesthetised rats with
indwelling vascular catheters (values are means ± s.e.m.; glucose levels in mmoll- 1)

Pre-E 2 3 S 7hafterE
Endotoxin (E) 6.S ±0.26 6.81 ±0.18 S.30 ± 0.26 3.S6±0.22 1.86 ±0.21
E + meptazinol 6.6 ± 0.20 6.67 ±0.28 7.61 ±0.26" 7.03 ±0.26 b 6.27 ±0.22b
Meptazinol alone 6.4 ± 0.26 6.60 ± 0.22 8.08 ± 0.26 6.81 ±0.20 6.72 ±0.22

"P<O.OS; b P<O.02 compared to endotoxin alone. Data from [1] and unpublished.

Table 7. The effect of meptazinol and naloxone on endotoxin-induced lacticacidaemia in unanaes-

thetised rats with indwelling vascular catheters. Meptazinol (0.1 mg/100 g, then O.OS mg/1oo g each
hour, i.v.) or naloxone (l.Omg/lOOg, then O.lmg/lOOg each hour i.v.) were given initially 2.Sh into
shock (values are means ± s.e.m.; lactate values in mmolll)

Pre-E 2 3 S 7hafterE

Endotoxin (E) 0.24 ± 0.03 0.S6±0.03 0.66 ± 0.04 0.84±0.07 0.82 ±O.OS
E + meptazinol 0.2S ±0.06 0.43 ±0.04 0.44 ± O.OS" O.4S ±0.04 b 0.48 ± O.OSb
E + naloxone 0.26 ± 0.01 O.Sl ±0.03 0.48 ± 0.04" 0.46 ±O.OSb 0.36±0.06b

Lactic acid levels did not change in either control (saline treated) animals or animals given meptazinol
or naloxone alone (values in the range 0.2S ± 0.04 mmolll).
"P<O.OS; b P<O.Ol compared to endotoxin alone animals.
Data from [1] and unpublished.

reversed the profound endotoxin-induced systemic hypotension; this reversal was

maintained throughout the 2h experimental period. Survival, however, was not af-
fected. Because this attenuation of depressed cardiovascular function occurred with
the (-) isomer but not with (+) naloxone it was suggested that this protection was
due to an action on specific opioid receptors.
These initial studies have, in general, been confirmed and extended to other
species and to other forms of shock (haemorrhagic, spinal, traumatic). These studies
have been extensively reviewed [32, 35]. Two examples will suffice.
In conscious rats given E. coli endotoxin Adeleye, Furman and I found in 1982
that both naloxone and the partial agonist meptazinol prevented the hypoglycaemia,
attenuated the metabolic acidosis and improved survival [1]. Some ofthese results are
illustrated in Fig. 3 and in Tables 6 and 7. These results suggest that opioid peptide
release is involved not only in the cardiovascular depression (shock) that results from
endotoxin administration, but also in some of its metabolic consequences. It is of
interest, however, that not all of the many effects of endotoxin are prevented by
these drugs; haemoconcentration and hyperthermia were not modified at all. The
second example is a quite different experimental model of septicaemia. In mice given
Borrelia duttoni organisms, naloxone not only attenuated the shock-induced reduc-
tions in circulatory white blood cells and platelets but also the reduction in body tem-
perature that accompanies infection in this species [89].
28 J. R. Parratt

It is not easy to summarise adequately the large number of studies with naloxone
in shock because of differences in experimental protocols, in dose, in species and in
the methods used to induce shock. Certainly there is good evidence for increased
levels of opioid peptides in shock [18, 21, 33, 83] and large doses of these peptides
administered systemically induce cardiovascular collapse. The evidence obtained
using the stereoisomers of naloxone indicates that it presumably acts by displacing
these endogenous opioid peptides from specific receptors, or preventing their access
to these receptors. There is evidence for attenuation by naloxone of the depressed
cardiovascular effects of endotoxin (reduced myocardial contractility and output,
systemic hypotension [9, 67], hypoglycaemia (Table 6), acidosis (Table 7) [1, 66],
haemoconcentration [67], and release of lysosomal enzymes and of myocardial de-
pressant factor [10] that result from endotoxaemia). The evidence for an improve-
ment in survival is less firm. Thus in the Reynolds study [67] survival at 24h was im-
proved by naloxone (from 21% in the control group to 83% in those dogs given
naloxone) but those animals still alive after 24 h were extremely sick and may well
have died shortly afterwards. As with the current controversy over infarct size limita-
tion, there is an important difference between actually improving survival from
shock and simply delaying death by a few hours. In our own conscious rat studies, all
those rats given only endotoxin died within 8 h whereas 66% of the naloxone-treated
rats were still alive after 20h [1]. However, further unpublished studies (Furman,
McKechnie and Parratt 1984) have shown that ultimate survival was not significantly
improved. The most likely explanation is that, as has been recently discussed [62],
many factors are involved in determining lethality in shock. Opioid peptides may in-
deed be involved (the evidence now seems almost conclusive) but an approach in
which just one of many possible factors is interfered with is almost certainly too
simplistic. In shock (as in myocardial infarction) the simultaneous use of several dif-
ferent approaches is likely to yield the most benefit.
Mechanisms of this naloxone protection are poorly understood but an interaction
with the sympatho-medullary system seems essential. The current hypothesis is that
endogenous opioid systems act at sites within the brain to inhibit sympatho-medul-
lary discharge. Naloxone would antagonise these central actions to improve cardio-
vascular function. Certainly the pressor effects of naloxone in shock, and its benefi-
cial effects on myocardial contractility, are mediated through increased circulating
levels of catecholamines and are prevented by adrenal demedullation or by blockade
of a- and ,B-adrenoceptors [74]. Such a mechanism might well explain the relative
failure of opiate antagonists to improve long-term survival in shock. Catecholamine
release is essential for short-term survival [25] but, if excessive, is probably detri-
mental in the later stages because of the transudation of fluid from the plasma into
the tissue spaces. This results from a noradrenaline-maintained post-capillary vaso-
constriction which is exaggerated under conditions of metabolic acidosis [50].
Although there was some initial optimism about the use of naloxone in clinical
shock most recent studies have failed to demonstrate reliable improvement in cardio-
vascular function [4, 69] or in survival [13]. An obvious major disadvantage is rever-
sal of opiate analgesia but a number of adverse effects have also been reported in-
cluding an analeptic action, pulmonary oedema [69] and ventricular fibrillation.
Recent studies have clarified the subtype of opioid receptor involved in the pro-
tective effect of naloxone and of other opiate antagonists in shock. This is important
Opioid Peptides and the Cardiovascular System 29

because, since shock and pain usually occur together, a general opiate antagonist
such as naloxone might intensify pain by ant agonising the effects of endogenous
opioid peptides on mu receptors. Further, analgesics such as morphine would then
not relieve pain in this situation. If, however, the opioid receptors concerned with
shock reversal were different to those involved in pain suppression it should be pos-
sible to attempt to reverse shock without modifying the protective analgesic effects
of released endogenous opioid peptides.
The available evidence suggests that delta receptors are those most likely in-
volved in mediating the protective effects in shock of opiate antagonists. Thus, M
154,129 and the ICI compounds 154129 and 174864 effectively reverse the hypoten-
sion associated with shock, whereas mu antagonists like naloxazone and fi-FNA do
not [12, 34, 35, 40, 54]. In haemorrhagic shock in cats the kappa antagonist M 2266
(Table 2) was ineffective whereas both J-7747 (a mainly delta antagonist, Table 2)
and C-7000 (a delta antagonist and kappa agonist) significantly reduced the increase
in plasma myocardial depressant factor activity; those cats given J-7747 had the high-
est final arterial pressure [11]. Possibly the best alternative to naloxone in shock
treatment would be a partial agonist combining mu (possibly kappa) mediated
analgesia with the cardiovascular benefits of delta receptor antagonism. Meptazinol,
a partial agonist at mu-1 receptors (see below), improves blood pressure in shock
[60] and nalbuphine (a competitive antagonist at mu receptors and a partial agonist
at kappa and delta receptors) improves haemodynamics and survival in hypovolemic
rats [57] and in dogs subjected to haemorrhagic shock [42].

Possible Involvement of Endogenous Opioid Peptides

in the Early Events of Myocardial Ischaemia

The studies, outlined above indicating a release of opioid peptides in various forms
of shock, suggested that they might also be released and active in other forms of
acute and prolonged stress. There is indeed evidence for this [70]. In particular, we
wondered about peptide release during the stress involved in cardiac pain arising
from myocardial ischaemia. The argument at that time (1981) was that if opioid re-
ceptor antagonists are beneficial in various shock states because they antagonise
some of the detrimental cardiovascular effects of opioid peptides (e.g. tissue hypo-
perfusion), then they might also exert useful effects in the tissue (myocardial) hypo-
perfusion associated with myocardial ischaemia. We examined this in a relatively
simple experimental model in anaesthetised rats, which we had then just developed
for evaluating drug effects on the various ventricular arrhythmias that occur soon
after coronary artery occlusion [7]. In addition, because of the possible modifying ef-
fects of anaesthesia on the release and actions of endogenous opioid peptides, we ex-
tended our existing collaboration with Szekeres's group in Szeged, Hungary, to ex-
amine effects of opioid receptor antagonists on the early, life-threatening ventricular
arrhythmias that arise in conscious rats subjected to coronary artery occlusion. The
initial studies [19] demonstrated a marked increase in survival (from 27% after 16h
in the controls to 73 % in those rats pretreated with naloxone 4 mg kg -1), a protective
effect mainly resulting from a reduction in the severity of occlusion-induced ventricular
 30 J. R. Parratt

Table 8. Effect of naloxone on the survival rate and occurrence of arrhythmias in the acute phase of
experimental myocardial infarction in conscious rats

n Survival after Occurrence of arrhythmias

20 min 16h None VF VT Other'
n % n % n % n % n % n %
Control 26 8 31 7 27 0 0 23 88 24 92 17 65
Naloxone
2mg/kg 12 8 67 b 7 58 3 25 c 7 58 b 7 58 b 4 33
4mg/kg 11 9 82 c 8 73 c 4 36c 2 18d 3 27 d 5 45

, Including single ventricular ectopic beats, atrioventricular block

b P<0.05
c P<0.01
d P<O.OOl (Chi-squared test);

[From 19].

100 a b 100
Fig. 4a, b. The percentage incidence of ven-
LL.
> tricular fibrillation (VF) (a) and of survivors
"0 I!! at both 20 min (solid line) and at 16h (dotted
~
'"g 50 50 .~ line) post-ligation (b) in conscious rats pre-
'"
"0
'0
:::J
(/)
treated with either saline (n = 26, open
columns) or meptazinol, 1mg kg- 1 (n = 12,
.S: "#.
"#. diagonally hatched columns) or 2mg kg- 1
(n = 12, vertically hatched columns) .
0 0 * P<0 .01 ; ** P<0.05 [from 20]

arrhythmias, especially fibrillation (Table 8). This antiarrhythmic effect of naloxone

has since been confirmed in other models of ischaemic arrhythmias, e.g. rat isolated
perfused hearts [90], rats subject to hypoxia [88], and in anaesthetised dogs subjected
to coronary artery occlusion and reperfusion [41). There is some evidence that this
antiarrhythmic effect might involve myocardial cyclic-AMP [51] .
The analgesic meptazinol, which is a partial agonist at mu receptors, and which
selectively binds to mu-l receptors [81], has a similar antiarrhythmic effect and also
improves survival from coronary artery occlusion in conscious rats (Fig. 4) [20]. Pro-
longation of action potential duration (Table 9) could explain this antiarrhythmic ef-
fect and the pronounced protection afforded by this compound in vivo. Whether this
electrophysiological effect of meptazinol results from interaction with specific (mu)
receptors on the sarcolemma has not been examined, but more comprehensive
studies are in progress to examine the possible role of different opioid receptors in
mediating changes in ionic currents in cardiac muscle fibres.
Although these results suggest that opioid receptors are involved in this protec-
tion there may well be other pharmacological actions that could explain the anti-
arrhythmic effect of naloxone and meptazinol in myocardial ischaemia. The possible
role of opioid receptors has been evaluated by using the stereoisomers of antagonists
Opioid Peptides and the Cardiovascular System 31

Table 9. Action potential characteristics of papillary muscle from rats given either saline (control) or
meptazinol (2 mg kg -I) intravenously 15 min before excision

Group n RMP APA MRD APD so APD 90

(mV) (mV) (VIs) (ms) (ms)

Control 49 76.6 ±0.8 96.9 ± 0.6 106±4 22.0± 1.0 56.5 ± 2.5
Meptazinol
2mgkg-1 i.v. 47 75.8 ± 0.7 96.4 ±0.7 100±4 31.5 ± 1.1" 80.2 ±2.7"
% change (+43) (+42)

n is the number of observations obtained from 5 papillary muscles (stimulation frequency 1 Hz) in
each group.
"P<0.05. Values shown are mean ± s.e.m.; RMP = resting membrane potential;APA = action po-
tential amplitude; MRD = maximum rate of depolarisation; and APD = action potential duration;
[From 20].

Table 10. Effect of the stereoisomers (-)-MrI452 and (+ )-MrI453 and (-)-WIN44,441-3 and (+)-
WIN 44,441-2 on the incidence and severity of ventricular arrhythmias resulting from acute coronary
artery occlusion in anaesthetised rats

Duration (in s) and % incidence of

n VEB's VT VF

Controls 16 1011 ± 280 42± 14 (94) 21 ± 12 (56)

WIN44,441-3 (3mg kg-I) 11 176 ± 51 b 3± Ib (64) 0 (O)b
WIN44,441-2 (3 mg kg-I) 11 516 ± 169 21 ± 8 (82) 6 ± 5 (27)
Controls 15 1123 ± 177 66 ± 15 (100) 8 ± 5 (33)
Mr1452 (Imgkg-l) 9 462± 194 35 ± 19 (100) 2.3 ± 2.3 (22)
Mr1452 (4mg kg-I) 10 367 ± 152b 18± 10 (90) 0.1 ± 0.1 (10)
Mr1452 (lOmg kg-I) 7 362 ± 159 b 8± 5 (43)" 0 (0)"
Mr1453 (4mgkg- l ) 10 809 ± 183 44± 16 (90) l.3± 1 (20)
Mr1453 (lOmg kg-I) 7 670 ± 169 34± 11 (86) 0.1 ± 0.1 (14)

" P<0.05; b P<O.OOl

The table shows the mean value ± s.e.m. for ventricular ectopic count (VEB's), and the incidence
(%) and duration of ventricular tachycardia (VT) and fibrillation (VF) [From 63].

that act mainly on mu receptors (Win 44,441-3) and kappa receptors (Mr 1452) [63];
as with ,B-adrenoceptor blocking drugs most of the activity at receptor level lies in the
(-) isomer. The results (Table 10) show that the Boehringer compound (Mr 1452),
which is the (-) isomer, is about twice as active as the (+) isomer (Mr 1453). The
evidence with the Win compounds also suggests that the antiarrhythmic activity lies
mainly in the (-) isomer (Table 10). Further studies with stereoisomers are clearly
required and it would be especially valuable if the in vivo effects of stereoisomers
selective for particular receptors could be correlated with changes in the cardiac
muscle action potential.
32 J. R. Parratt

Table 11. The effects of buprenorphine on the severity of the ventricular arrhythmias that occurred
over the 0-30 min post-occlusion period in anaesthetised rats (method of Clark et al. [7])

Group n Ventricular Duration Duration %VT %VF Mor-

ectopic count ofVT(s) ofVF(s) tality

Controls 11 1089 ± 291 54.6 ± 20.3 4.7 ± 3.1 100 64 27

Buprenorphine 9 714± 191 30.5 ± 2.9 3.8±2.9 100 22 0
(0.1 mg kg-l)
Buprenorphine 10 860 ± 350 41.0 ± 20.8 1.3 ±0.8 90 20 0
(0.3mgkg-l)
Buprenorphine 10 486± 106 17.3 ± 4.5 0.1 ±0.1 90 lOa 0
(1.0mgkg- 1)

The table shows the mean values ± s.e.m. for the ventricular ectopic count, the incidence and dura-
tion of ventricular tachycardia (VT) and ventricular fibrillation (VF) and the mortality. Values for
the ventricular ectopic count and the duration of VT and VF are taken only from rats that survived
beyond 30 min.
n = number of animals; a P<0.05;
[From 78 and Sitsapesan and Parratt (unpublished)].

There are two questions that have yet to be completely answered. First, which
opioid receptor type is involved in the antiarrhythmic effects of opiate antagonists in
myocardial ischaemia. Second, the precise sites of action, within the central nervous
system or in the periphery. The limited available evidence suggests that kappa recep-
tors may be the most important receptor type involved [56] and, because naloxone
methobromide (MrZ 2593), which is a quaternary complex of naloxone and does not
enter the central nervous system, is also active in the Clark model, that peripheral
sites of action are involved [55].
The possible clinical relevance of these studies is that morphine-like analgesics
are commonly used to combat pain resulting from myocardial infarction. As Helge-
sen and Refsum [30] have recently pointed out, relatively little is known about the
cardiac effects of most analgesics, especially when the myocardium is compromised
by ischaemia. The question is therefore raised as to which analgesic might be best to
use in the early stages of a myocardial infarction. Would, for example, a partial
agonist at all opioid receptors, or a compound combining mu-l receptor agonist
properties (and therefore analgesia) with say, kappa receptor blockade be more ap-
propriate than standard analgesics? A potent analgesic with antiarrhythmic proper-
ties and with an ability to maintain arterial blood pressure, and hence coronary per-
fusion, in the early period after coronary artery occlusion might be a real advance
over existing analgesics routinely used in this situation. One possibility is buprenor-
phine [78]. Table 11 shows that this reduced the severity of early ischaemia-induced
arrhythmias; it also maintained arterial pressure. This is an important, relevant area
for future research.
Opioid Peptides and the Cardiovascular System 33

References
1. Adeleye GA, Furman BL, Parratt JR (1982) Possible involvement of opiate receptors in the re-
sponse of conscious rats to Escherichia coli endotoxin; protective effect of naloxone. J Physiol
329:31P
2. Altura BT, Altura BM, Quirion R (1984) Identification of benzomorphan - Kappa opiate
receptors in cerebral arteries which subserve relaxation. Br J Pharmacol 82: 459-466
3. Bergey JL, Beil M (1983) Antiarrhythmic evaluation of naloxone against acute coronary oc-
clusion-induced arrhythmias in pigs. Eur J Pharmacol 90: 427-431
4. Bonnet F, Bilaine J, Lhoste F, Mankikian B, Kerdelhue B, Rapin M (1985) Naloxone therapy
of human septic shock. Crit Care Med 13: 972-975
5. Bruckner UB, Lang RE, Ganten D (1984) Release of opioid peptides in canine hemorrhagic
hypotension: effects of naloxone. Res Exp Med (Berl) 184: 171-178
6. Byrd LD (1983) Cardiovascular effects of naloxone, naltrexone and morphine in the squirrel
monkey. Life Sci 32:391-398
7. Clark C, Foreman MI, Kane KA, McDonald FM, Parratt JR (1980) Coronary artery ligation in
anaesthetised rats as a method for the production of experimental dysrhythmias and for the
determination of infarct size. J Pharmacol Methods 3: 357-368
8. Clement-Jones V, Lowry PJ, Rees LH, Besser GM (1980) Metenkephalin circulates in human
plasma. Nature 283: 295-297
9. Cronenwett JL, Baver-Neff BS, Grekin RJ, Sheagren IN (1986) The role of endorphins and
vasopressin in canine endotoxin shock. J Surg Res 41: 609-619
10. Curtis MT, Lefer AM (1980) Protective actions of naloxone in hemorrhagic shock. Am J Physiol
239:416-421
11. Curtis MT, Lefer AM (1983) Actions of opiate antagonists with selective receptor interactions
in hemorrhagic shock. Circ Shock 10: 131-135
12. D'Amato RJ~ Holaday JW (1984) Multiple opiate receptors in endotoxic shock: Evidence for
delta involvement and mu-delta interactions in vivo. Proc Nat! Acad Sci USA 81: 2898-2901
13. De Maria A, Craven DE, Heffernan JJ, Mcintosh TK, Grindlinger GA, McCabe WR (1985)
Naloxone versus placebo in treatment of septic shock. Lancet 1: 1363-1365
14. Diz DI, Vitale JA, Jacobowitz DM (1984) Increases in heart rate and blood pressure produced
by injections of dermorphin into discrete hypothalamic sites. Brain Res 294: 47-57
15. Dumont M, Lemaire S (1984) Opioid receptors in bovine adrenal medulla. Can J Physiol Phar-
macol 62: 1284-1291
16. Eimerl J, Feuerstein G (1986) The effect of J.l, 8, /C and S opioid receptor agonists on heart rate
and blood pressure of the pithed rat. Neuropeptides 8: 351-358
17. Eulie PJ, Rhee HM, Laughlin MH (1987) Effects of [Met 5]enkephalin on regional blood flow
and vascular resistance in rabbits. Eur J Pharmacol137: 25-31
18. Evans SF, Medbak S, Hinds CJ, Tomlin SJ, Varley JG, Rees LH (1984) Plasma levels and bio-
chemical characterisation of circulating metenkephalin in canine endotoxin shock. Life Sci 34:
1481-1486
19. Fagbemi 0, Lepran I, Parratt JR, Szekeres L (1982) Naloxone inhibits early arrhythmias result-
ing from acute coronary ligation. Br J Pharmacol 76: 504-506
20. Fagbemi 0, Kane KA, Lepran I, Parratt JR, Szekeres L (1983) Antiarrhythmic actions of
meptazinol, a partial agonist at opiate receptors, in acute myocardial ischaemia. Br J Pharmacol
78:455-460
21. Farrell LD, Harrison TS, Demers LM (1983) Immunoreactive metenkephalin in the canine
adrenal: response to acute hypovolaemic stress. Proc Soc Exp Bioi Med 173: 515-518
22. Feuerstein G, Faden AI (1982) Differential cardiovascular effect of mu, delta, and kappa opiate
agonists at discrete hypothalamic sites in the anesthetized rat. Life Sci 31: 2197-2280
23. Feuerstein G, Faden AI (1984) Cardiovascular effects of dynorphin A-(1-8), dynorphin A-(1-13)
and dynorphin A-(1-17) microinjected into the preoptic medialis nucleus of the rat. Neuro-
peptides 5: 295-298
24. Feuerstein G, Zukowska-Grojec Z (1987) Effect of dermorphin and morphine on the sympathetic
and cardiovascular system of the pithed rat. Neuropeptides 9: 139-150
25. Furman BL, Dean HG, McKechnie K, Parratt JR (1985) Plasma catecholamines during endo-
toxin infusion in conscious unrestrained rats. Circ Shock 17: 85-94
34 J. R. Parratt

26. Hassen AH, Feuerstein G (1987) ,u-Opioid receptors in NTS elicit pressor responses via sym-
pathetic pathways. Am J Physiol 252: 156-162
27. Hassen AH, Feuerstein G, Faden AI (1984) Kappa opioid receptors modulate cardiorespiratory
function in hindbrain nuclei of rat. J Neurosci 4: 2213-2221
28. Hassen AH, Feuerstein G, Faden AI (1984) Selective cardiorespiratory effects mediated by mu-
opioid receptors in the nucleus ambiguus. Neuropharmacology 23: 407-415
29. Hassan AH, Feuerstein G, Pfeiffer A, Faden AI (1982) <5 Versus ,u receptors: cardiovascular and
respiratory effects of opiate agonists microinjected into nucleus tractus solitarius of cats. Regul
Pept 4:299-309
30. Helgesen KG, Refsum H (1987) Arrhythmogenic, antiarrhythmic and inotropic properties of
opioids. Pharmacology 35 : 121-129
31. Heydorn WH, Moores WY, Bellamy RF, O'Benar JD (1985) Naloxone: ineffective in improv-
ing cardiac performance after hypoperfusion in swine. Circ Shock 17: 35-43
32. Hinds CJ (1987) Endogenous opioid peptides in shock. In: Vincent JL, Thijs LG (eds) Septic
shock, European view. Springer, Berlin Heidelberg New York Tokyo, pp 268-275 (Update in
intensive care and emergency medicine, vol 4)
33. Hinds CJ, Evans SF, Varley JG, Tomlin S, Rees LH (1985) Neuroendocrine and cardiovascular
changes in septic shock and after cardiac surgery: effect of high-dose corticosteroid therapy. Circ
Shock 15: 61-72
34. Holaday JW (1983) Cardiovascular effects of endogenous opiate systems. Annu Rev Pharmacol
Toxicol23: 541-594
35. Holaday JW (1984) Neuropeptides in shock and traumatic injury: sites and mechanisms of
action. In: Muller EE, MacLead RM (eds) Neuroendocrine perspectives, vol 3. Elsevier,
Amsterdam, pp 161-199
36. Holaday JW, Faden AI (1978) Naloxone reversal of endotoxin hypotension suggests role of
endorphins in shock. Nature 275: 450-451
37. Holaday JW, Faden AI (1982) Selective cardiorespiratory differences between third and fourth
ventricular injections of "mu" and "delta" opiate agonists. Fed Proc 41: 1468
38. Holaday JW, Long JB, Tortella FC (1985) Evidence for IC, ,u, and <5opioid-binding site inter-
actions in vivo. Fed Proc 44: 2860-2862
39. Holaday JW, Pasternak GW, D'Amato RJ (1983) Naloxazone lacks therapeutic effects in endo-
toxic shock yet blocks the effects of naloxone. Eur J Pharmacol89: 293-296
40. Holaday JW, Ruvio BA, Robles LE, Johnson CE, D'Amato RJ (1982) M 154,129, a putative
delta antagonist, reverses endotoxic shock without altering morphine analgesia. Life Sci 31:
2209-2212
41. Huang XD, Lee A YS, Wong TM, Zhan CY, Zhao YY (1987) Naloxone inhibits arrhythmias in-
duced by coronary artery occlusion and reperfusion in anaesthetized dogs. Br J Pharmacol 87:
475-477
42. Hunt LB, Gurll NJ, Reynolds DG (1984) Dose-dependent effects of nalbuphine in canine hemor-
rhagic shock. Circ Shock 13:307-318
43. Illes P, Bettermann R (1986) Classification of presynaptic opioid receptors in the rabbit ear
artery by competitive antagonists. Eur J Pharmacol 122: 153-156
44. Illes P, Pfeiffer N, Limberger N, Klaus S (1983) Presynaptic opioid receptors in the rabbit ear
artery. Life Sci 33:307-310
45. Jaszlits L, Kosa E, Tardos L, Danitz B, Kauthny K, Szekely II (1982) Cardiovascular effects of
a potent opioid peptide (D-Met 2 , Pro5)-enkephalinamide. Arch Int Pharmacodyn Ther 260:
91-99
46. Koskinen LO, Bill A (1983) Regional cerebral, ocular and peripheral vascular effects of naloxone
and morphine in unanaesthetised rabbits. Acta Physiol Scand 119: 235-241
47. Kugelgen I von, Illes P, Wolf D, Starke K (1985) Presynaptic inhibitory opioid 8- and IC-recep-
tors in a branch of the rabbit ileocolic artery. Eur J Pharmacol 118: 97-105
48. Krumins SA, Faden AL, Feuerstein G (1985) Diprenophrine binding in rat heart: modulation
of binding sites in the brain. Biochem Biophys Res Commun 127: 120-128
49. Lechner RB, Gurll NJ, Reynolds DG (1985) Naloxone potentiates the cardiovascular effects of
catecholamines in canine hemorrhagic shock. Circ Shock 16: 347 -361
50. Ledingham IMcA, Parratt JR (1976) Pathophysiology of shock. In: Ledingham IMcA (ed)
Shock. Excerpta Medica, Amsterdam, pp 1-20
Opioid Peptides and the Cardiovascular System 35

51. Lee A YS, Wong TM (1986) Naloxone attenuates augmentation of cAMP levels and arrhythmias
following myocardial ischaemia and reperfusion in the isolated perfused rat heart. Clin Exp
Pharmacol Physiol13 : 707-710
52. Lemaire I, Tseng R, Lemaire S (1978) Systemic administration of beta-endorphin: potent hypo-
tensive effect involving a serotonergic pathway. Proc Nat! Acad Sci USA 75: 6240-6242
53. Levy R, Feustel P, Severinghaus J, Hosobuchi Y (1982) Effect of naloxone on neurologic deficit
and cortical blood flow during focal cerebral ischaemia in cats. Life Sci 31: 2205-2208
54. Long JB, Ruvio BA, Holaday JW (1984) ICI 174864, a novel "d" antagonist, reverses endotoxic
shock: pretreatment with dynorphin (1-13) a "k"agonist, blocks this therapeutic effect. Neuro-
peptides 5: 291-294
55. Mackenzie JE, Parratt JR, Sitsapesan R (1986) The antiarrhythmic properties of naloxone and
MrZ 2593 (Naloxone methobromide) in vivo and in vitro. Br J Pharmacol 89: 613 P
56. Mackenzie JE, Parratt JR, Sitsapesan R (1986) The effects of drugs interacting with opioid
receptors on ischaemic arrhythmias in anaesthetised rats. Br J Pharmacol89: 614P
57. McKenzie JE, Anselmo DM, Muldoon SM (1985) Nalbuphine's reversal of hypovolemic shock
in the anesthetized rat. Circ Shock 17: 21-33
58. Martin WR (1981) Multiple opioid receptors. Life Sci 28: 1547-1554
59. North RA, Williams JT (1983) Opiate activation of potassium conductance inhibits calcium
action potentials in rat locus coeruleus neurones. Br J Pharmacol 80: 225-228
60. Paciorek PM, Todd MR, Waterfall JF (1985) The effects of meptazinol in comparison with
pentazocine, morphine and naloxone in a rat model of anaphylactic shock. Br J Pharmacol84:
469-475
61. Paterson SJ, Robson LE, Kosterlitz HW (1983) Classification of opioid receptors. Br Med Bull
39:31-36
62. Parratt JR (1983) Neurohumoral agents and their release in shock. In: Altura BM, Lefer AM,
Schumer W (eds) Handbook of shock and trauma, vol 1. Raven, New York, pp 311-336
63. Parratt JR, Sitsapesan R (1986) Stereospecific antiarrhythmic effect of opioid receptor antago-
nists in myocardial ischaemia. Br J PharmacoI87:621-622
64. Pfeiffer A, Feuerstein G, Kopin IJ, Faden AI (1983) Cardiovascular and respiratory effect of
mu, delta and kappa-opiate agonists micro-injected into the anterior hypothalamic brain area of
awake rats. J Pharmacol Exp Ther 225:735-741
65. Pfeiffer A, Feuerstein G, Zerbe RL, Faden AI, Kopin IJ (1983) {I-Receptors mediate opioid
cardiovascular effects at anterior hypothalamic pathways. Endocrinology 113: 929-938
66. Raymond RM, Harkema JM, Stoffs WV, Emerson TE (1981) Effects of naloxone therapy on
hemodynamics and metabolism following a super lethal dosage of Escherichia coli endotoxin in
dogs. Surg Gynecol Obst 152: 159-162
67. Reynolds DG, Gurll NJ, Vargish T, Lechner RB, Faden AI, Holaday JW (1980) Blockade of
opiate receptors with naloxone improves survival and cardiac performance in canine endotoxic
shock. Circ Shock 7: 39-48
68. Rhee HM, Eulie PJ, Peterson DF (1985) Suppression of renal nerve activity by methionine
enkephalin in anesthetized rabbit. J Pharmacol Exp Ther 234: 537-544
69. Rock P, Silverman H, Plump D (1985) Efficacy and safety of naloxone in septic shock. Crit Care
Med 13:28-33
70. Rossier J, French ED, Rivier C, Ling N, Guillemin R, Bloom FE (1977) Foot shock induced
stress increases p-endorphin levels in blood but not in brain. Nature 270: 618-620
71. Rubin PC, McLean K, Reid JL (1981) Opioids and blood pressure control in man. Br J Pharma-
col 74:51P
72. Sander GE, Giles TD, Kastin A, Quiroz J, Kaneish AC, Coy DH (1981) Cardiopulmonary
pharmacology of enkephalins in the conscious dog. Peptides 2: 403-407
73. Saxon ME, Ivanitsky GR, Beloyartsev FF (1982) Myocardial opiate receptors. Gen Physiol Bio-
phys 1: 447-452
74. Schadt JC, York DH (1982) Involvement of both adrenergic and cholinergic receptors in the
cardiovascular effects of naloxone during hemorrhagic hypotension in the conscious rabbit. J
Auton Nerv Syst 6:237-251
75. Schaz K, Stock G, Simon W, Schior K, Unger H, Rockhold T, Ganten R (1980) Enkephalin ef-
fects on blood pressure, heart rate and baroreceptor reflex. Hypertension 2: 395-407
76. Simon W, Schaz K, Ganten U, Stock G, Schlor KH, Ganten D (1978) Effects of enkephalins on
arterial blood pressure are reduced by propranolol. Clin Sci Mol Med 55: 237-241
36 J. R. Parratt: Opioid Peptides and the Cardiovascular System

77. Sitsapesan R, Parratt JR (1988) Opioid binding in rat hearts and the effects of coronary artery
occlusions. Br J Pharmacol (submitted)
78. Sitsapesan R, Parratt JR, Mackenzie JE (1987) The effects of buprenorphine, a partial agonist
at opioid receptors, during acute myocardial ischaemia. In: Abstracts of the IVPHAR Congress,
Sydney, 0383
79. Slizgi R, Taylor J, Ludens H (1984) Effects of the highly selective kappa opioid, V-50,488, on
renal function in the anaesthetised dog. J Pharmacol Exp Ther 230: 641-645
80. Spampinato S, Goldstein A (1983) Immunoreactive dynorphin in rat tissues and plasma. Neuro-
peptides 3: 193-212
81. Spiegel K, Pasternak GW (1984) Meptazinol: a novel mu-1 selective opioid analgesic. J Pharma-
col Exp Ther 228:414-419
82. Starke K, Schoffel E, Illes P (1985) The sympathetic axons innervating the sinus node of the rab-
bit possess presynaptic opioid IC - but not Jl- or O-receptors. Naunyn Schmiedebergs Arch Phar-
macoI329:206-209
83. Watson JD, Varley JG, Tomlin SJ, Medbak S, Rees LH, Hinds CJ (1986) Biochemical charac-
terization of circulating Met-enkephalins in canine endotoxin shock. J EndocrinoI111: 329-334
84. Way EL (1985) Characterization of opiate receptors with excised tissues and organs. Fed Proc
44:2855-2857
85. Weihe E, McKnight AT, Corbett AD, Hartschuh W, Reinecke M, Kosterlitz HW (1983)
Characterisation of opioid peptides in guinea pig heart and skin. Life Sci 33 : 711-714
86. Weihe E, McKnight AT, Corbett AD, Kosterlitz HW (1985) Proenkephalin- and prodynorphin-
derived opioid peptides in guinea pig. Neuropeptides 5: 453-456
87. Willette RN, Punnen S, Krieger AJ, Sapru HN (1984) Hypertensive response following stimula-
tion of opiate receptors in the caudal ventrolateral medulla. Neuropharmacology 23:401-406
88. Wong TM, Lee A YS (1985) Cardiac antiarrhythmic evaluation of naloxone with or without
propranolol using a modified chloroform-hypoxia screening test in the rat. Clin Exp Pharmacol
Physiol 12: 379-385
89. Wright DJM (1981) The fall in circulating leucocyte and platelet counts after endotoxin: an
adrenergic-opioid interaction. Neuropeptides 1: 181-202
90. Zhan ZY, Lee AYS, Wong TM (1985) Naloxone blocks the cardiac effects of myocardial
ischaemia and reperfusion in the rat isolated heart. Clin Exp Pharmacol Physiol 12: 373-378
Chapter 3

Cardiac and Cerebral Effects of Local Anesthetics

R. HOTVEDT and H. REFSUM

Local anesthetic drugs affect excitable cellular membranes, and are widely used as
antiarrhythmics as well as in different forms of local and regional anesthesia. The
principal target organs for their toxic effects are the central nervous system (eNS)
and the cardiovascular system. Local anesthetics may also influence cardiovascular
function by inducing regional blockade of autonomic innervation. This may, for
example, be seen during high spinal or thoracic epidural anesthesia, in which cardiac
sympathetic nerve fibers may be blocked at their outlet from the spinal cord.

Thoracic Epidural Anesthesia and Cardiac Function

To investigate effects of thoracic epidural anesthesia (TEA) on cardiac electro-

physiology, metabolism, and hemodynamics, we carried out a series of experiments
in pentobarbital anesthetized dogs [19, 20]. Intracardiac impulse conduction times
were measured by His bundle electrography, and refractoriness was determined by
programmed electrical stimulation. Monophasic action potentials were recorded
from the right ventricle by suction electrode technique. Left ventricular blood pres-
sure was recorded by micromanometer-tipped catheter. Arterial plasma concentra-
tions of free fatty acids (FFA) were determined by radiochemical method [15]. TEA
was induced by injecting a local anesthetic agent, bupivacaine, into the epidural
space at vertebral level T2.

Electrophysiological Effects

Thoracic epidural anesthesia reduced heart rate (Fig. 1). TEA increased ventricular
refractoriness (Fig. 2) and the duration of the monophasic action potential (Fig. 3).
The intraatrial and His-Purkinje impulse conduction time and the QRS width were
not significantly influenced. Atrioventricular nodal conduction time (Fig. 4) and AV
nodal refractoriness were markedly prolonged by TEA. Following a high dose of
bupivacaine, TEA induced second degree A V nodal conduction block in most experi-
ments during high pacing frequencies.

Hemodynamic and Metabolic Effects

TEA markedly reduced mean aortic blood pressure (Fig. 5), left ventricular systolic
blood pressure and left ventricular dP/dt max (Fig. 6). TEA also reduced the plasma
38 R.Hotvedt and H.Refsum

200

150

Heart rate .::-=:::::: - ---;

(beats. min-') --.---
i~'~i
100 .-_.-.
:~.~:

.----.~
! i
2 3 4 5 Hours

Fig.t. Effect of thoracic epidural anesthesia induced with two doses of bupivacaine (TEAl: 0.7 mgt
kg, TEA 2 : 3.3mg/kg) on spontaneous heart rate. Values for individual dogs and median are given
[from 20]

concentration of FFA (Fig. 7). It should be noted that the FFA-Iowering effect was
greatest when the FFA values were the highest.

Mechanisms for Cardiac Effects

Our experiments show that the local anesthetic, bupivacaine, when used in regional
anesthesia techniques such as TEA, has significant effects on cardiac function. To
elucidate the mechanisms for these effects, we investigated in the same experimental
model the effects of TEA following fi-adrenoceptor blockade, achieved by intra-
venous injection of the cardioselective fi-blocker atenolol [21]. The combined effects
of f3-blockade and TEA are interesting also from a clinical point of view since use of
TEA may be considered in patients treated with fJ-blockers.
Atenolol reduced heart rate, prolonged A V nodal conduction time and refractor-
iness, prolonged ventricular refractoriness and action potential duration, and de-
creased left ventricular dP/dt max. Addition of TEA further reduced heart rate, pro-
longed A V nodal conduction time and refractoriness, decreased left ventricular dPI
dt max and mean aortic blood pressure, but had no effect on atrial and ventricular
electrophysiology. Induction of TEA with a local anesthetic drug thus had additive
depressive effects on sinoatrial and A V nodal functions as well as on left ventricular
inotropy [21].
Our studies indicate that the cardiac effects of TEA induced with local anesthetic
are mainly caused by decreased beta-adrenoceptor stimulation [21], which follows
blockade of cardiac sympathetic nerves by the local anesthetic drug in the epidural
space.
Cardiac and Cerebral Effects of Local Anesthetics 39

./:
ms

~.
170
.
./.
:>Z.
150
;~: 170.min-1

130
,
,,
.---.

~
160
. .
140

120
~.
· ~7L: 200'min-1

.---. "

~
150

130
~.
·
230·min-1 Fig. 2. Effect of thoracic epidural an-
esthesia induced with two doses of
110 bupivacaine (TEAl and TEA 2) on
ventricular effective refractory pe-
riod. Cl indicates data obtained be-
150 fore TEAl was administered, and C2
indicates data obtained immediately

130
~..
·· ~~
:~: 260.min-1
before administration of TEA2
210 min after TEAl' Data from
individual dogs are given at pacing
frequencies 170, 200, 230 and 260
110 beats min- 1 [from 19)
C1 TEAl C2 TEA2

TEA Fig. 3. Monophasic action potential record-

ing at spontaneous heart rate before (control)
and after induction of thoracic epidural an-
···-----90%rep------------ -- esthesia (TEA) with bupivacaine. The pro-
longations of 50% and 90% repolarization
(rep) times are indicated [from 19)
lOOms

However, increased vagal nervous activity also contributes. Cardiac vagal ef-
ferent activity has been shown to be modulated by changing activity in afferent fibers
from cardiac receptors. Such afferent fibers travel in sympathetic nerves towards
the CNS [50,51]. It has been demonstrated that stimulation of sympathetic afferent
fibers inhibits vagal efferent activity [6, 40]. Correspondingly, cardiac sympathetic
40 R. Hotvedt and H. Refsum

150

100

AH-time
(ms)

50

20
2 3 4 5 Hours

Fig.4. Effect of thoracic epidural anesthesia induced with two doses of bupivacaine (TEAl and
TEAl) on AV nodal conduction time (AH-time). Data from individual dogs, pacing frequency 200
beats min-i. In one dog, the second TEA induced an AV block of second degree [from 19]

:~ .-------.
150
:~\-------.:/.----'
~~:
---.-- ----. ~===
PAo
(mmHg)

100
"".~.
'"".
<::::::::::::::::.==--.:::---- .--........-----.
.--.---------.
. ~~
50
.-.-----
2 3 4 5 Hours

Fig.s.Effect of thoracic epidural anesthesia induced with two doses of bupivacaine (TEAl and
TEAl) on mean aortic blood pressure (PAO)' Values for individual dogs and median are given
[from 20]
Cardiac and Cerebral Effects of Local Anesthetics 41

4000

3000

LVdP/dt max
(mmHg sec-1)
.--'
2000

1000

2 3 4 5 Hours

Fig.6. Effect of thoracic epidural anesthesia induced with two doses of bupivacaine (TEAl and
TEA 2 ) on the maximum rate of change of left ventricular pressure (L V dPldtmax). Values for indi-
vidual dogs and median are given [from 20]

nervous blockade induced by a local anesthetic agent in the epidural space may in-
crease the cardiac vagal efferent activity.
Thoracic epidural analgesia induced with morphine can also have significant car-
diac electrophysiological and hemodynamic effects [18]. In another series of experi-
ments, we have shown that injection of morphine into the epidural space in dogs re-
duced heart rate, prolonged A V nodal conduction time and refractoriness, and re-
duced ventricular contractility. These effects were due to increased vagal activity and
could be reversed by vagotomy. We have suggested that these effects are due to mor-
phine which via the cerebrospinal fluid reaches central opioid receptors involved in
cardiovascular regulation [18].

Clinical Implications

We have shown that TEA induced with local anesthetic prolongs action potential
duration and refractoriness in the ventricles, without influencing the conduction of
impulses through the ventricles. Theoretically, TEA may thereby protect the heart
against ventricular arrhythmias. This should be true when the arrhythmia is caused
either by reentry circuit [3, 32], or by increased automaticity. There is abundant evi-
dence from studies in animals and man that prolongation of action potential duration
42 R. Hotvedt and H. Refsum

800

700

7-
E
.------.
600

\Ll--
(5
E
E
(I)
c 500

/
c::;
~

~~
>
f-
f- 400
~
u.
w
w

V~
a:
u. 300

200
~=y
'-.
.:=:::::::.~.
:~ :=--C:::::: :
Fig. 7. Effect of thoracic epidural anes-
thesia induced with two doses of bupi-
vacaine (TEAl and TEA 2 ) on plasma
100 ~..-.-.-~ . '
concentrations of free fatty acids. Values
50
.----..----. for individual dogs and median are given
, [from 20]

• •
I I I I
0 2 3 4 5 Hours

TEA1 TEA2

and refractoriness, without change in impulse conduction, is an important antiar-

rhythmic mechanism [7,12,35,37-39,53].
In animal models and in man, increased cardiac sympathetic nervous tone in-
creases the occurrence of ventricular arrhythmias [10, 29, 30]. On the other hand, re-
duced sympathetic nervous tone protects against ventricular arrhythmias [13, 27].
Further, there are experimental and clinical studies indicating that increased vagal
nervous tone may terminate ventricular arrhythmias [4, 25, 34, 55, 57]. Thus, experi-
mental and clinical evidence support the view that TEA induced with local anesthetic
may protect the heart against ventricular arrhythmias.
Another clinical electrophysiological implication is that TEA induced with local
anesthetic may cause A V nodal block, as it prolongs A V nodal conduction time and
refractoriness. TEA should therefore be used with caution in patients with A V nodal
conduction disturbances.
Heart rate, left ventricular systolic blood pressure and dP/dt max, and mean
aortic blood pressure are major determinants of myocardial oxygen consumption
[46]. Therefore, by reducing these variables, TEA reduces myocardial oxygen de-
mand. Experimentally, it has been shown that TEA induced with bupivacaine re-
duces the severity of acute myocardial ischemic injury [54]. However, a too pro-
nounced fall in diastolic blood pressure, caused by too great a dose of local anesthetic
spreading in the epidural space, may jeopardize coronary blood perfusion and there-
by possibly worsen the balance between myocardial oxygen supply and demand.
Cardiac and Cerebral Effects of Local Anesthetics 43

Myocardial oxygen demand is also influenced by the metabolic substrate supply.

During physiological conditions the major sources for energy production are glucose
and fatty acids. High plasma levels with increased utilization of FFA increase the
oxygen requirement of the heart, and seem to increase the severity of acute myocar-
dial ischemic injury [26, 33, 36]. We found that TEA induced with bupivacaine re-
duced high levels of FFA. Thus, TEA reduces myocardial oxygen demand also by its
FFA-Iowering effect.

Systemic Effects of Local Anesthetics

When discussing the effects of epidural anesthesia and their mechanisms, possible
contributory effects from bupivacaine absorbed into the systemic circulation have to
be taken into consideration. The main target organs for therapeutic as well as toxic
systemic effects of local anesthetics are the heart and the brain. The effects depend
on the concentration of the drug in the target organ, which again depends on the af-
ferent arterial plasma concentration of the drug.

Pharmacokinetics

Following paravascular injection, the arterial plasma level seems to be determined

primarily by the total dose, and not by the concentration, of the injected drug. The
site of injection is also important. Absorption of the drug from any site depends on
the blood supply, richly supplied areas favoring rapid absorption and high plasma
levels. A rich blood supply also increases the risk of accidental intravascular injec-
tion.
When the local anesthetic drug is injected intravenously, the speed of injection is
critical to the determination of the arterial plasma concentration. This is especially
important in conditions with low cardiac output. In such conditions not only is the
drug mixed with smaller blood volumes in the first circulation, but the proportion of
the cardiac output going to the heart and the brain is also increased.
In the distribution of the local anesthetic, the lungs have a special function. The
lungs are the first organ to be exposed to the drug once it has entered the systemic
circulation, and they have been shown to take up lidocaine, bupivacaine, and
mepivacaine in both animals and humans [1, 24, 45, 49]. The capacity is consider-
able, and lung uptake occurs rapidly. For example, Jorfeldt et al. [23] have shown in
healthy volunteers that during the first 5 s after intravenous injection of a bolus dose
of 0.5 mg lidocaine/kg body weight, more than 90% of the dose is taken up by the
lungs. The local anesthetic does not seem to be metabolized in the lungs, and is sub-
sequently released into the systemic arterial circulation. This organ thereby provides
an important buffer function of temporary sequestration of a large quantity of drug.
Hence, the arterial plasma drug concentration which hits the target organs, the heart
and the brain, is considerably lower compared to the drug concentration in the pul-
monary artery (Fig. 8) [49].
44 R. Hotvedt and H. Refsum

...E
.....
ii
6

E
......
51
.
.!
c
4

·5 Fig. 8. Plasma lidocaine levels in a subject following

0
;g cuff release after intravenous regional anesthesia with
...
::>.
2 3mg kg-l lidocaine hydrochloride (0.5% solution) .
Interval from injection to cuff-release was 45 minutes .
• , Pulmonary artery; 0, brachial artery (contralateral);
0 . , antecubital vein (contralateral) [from 49]
0 10 20 30
time after cuff-release(min)

Local anesthetics are transported in plasma mainly bound to alpharacid glyco-

protein. This protein binding is highly dependent on the acid-base status of the
blood. Low pH reduces the affinity of the binding sites, and increases the concentra-
tion of the free fraction of the drug [5, 9]. On the other hand, alphal-acid glyco-
protein is an acute phase reactant, and its concentration may increase in patients
with, for example, myocardial infarction, renal disease, or cancer, or during post-
operative stress. Thereby the binding capacity may be increased, and the therapeutic
and the toxic effects may be reduced [44].

Cerebral and Cardiac Toxicity

The cardiovascular system is generally considerably more resistant than the eNS to
toxic effects of local anesthetics. As blood and brain concentrations of the drug in-
crease, a well-described sequence of symptoms and signs from the eNS occurs [47].
At low concentrations, local anesthetics have sedative and analgesic, as well as anti-
convulsive, properties. At higher concentrations, the patients may feellight-headed-
ness, dizziness, numbness of lip and tongue, a metallic taste, visual and auditory dis-
turbances, nausea, and sleepiness. Parallel objective signs may include confusion,
dysarthria, nystagmus, and vomiting. At higher concentrations, objective signs pro-
gress to include shivering, choreiform movements, twitching, and ultimately, frank
convulsions. This may progress to generalized depression ofthe eNS, with coma and
ultimately death due to respiratory arrest.
Although still controversial, most investigators believe that local anesthetics gen-
erally inhibit neuronal activity, but that excitatory pathways are more resistant than
inhibitory pathways. Thus, at subtoxic doses, local anesthetics act as anticonvulsants;
at higher concentrations, resistant unopposed excitatory pathways cause convul-
sions; and at still higher concentrations, all pathways are inhibited [47].
In man, EEG recording has not been very useful in detecting the onset of toxicity
prior to convulsions. Local anesthetics have been given to volunteers to produce sub-
jective and objective signs of toxicity, including convulsions [11, 41, 52]. In none of
Cardiac and Cerebral Effects of Local Anesthetics 45

these experiments was any alteration detected in the EEG before onset of convul-
sions.
At high concentrations, local anesthetics have effects on the cardiovascular sys-
tem which involve both the heart and the peripheral blood vessels. With increasing
plasma levels, the sequence of signs and symptoms again follows a pattern of stimula-
tion, followed by depression [42, 56].

Systemic Cardiac Effects of Bupivacaine

It has generally been assumed that with local anesthetics there is a wide margin be-
tween central nervous effects and cardiovascular depression, and that acute cardio-
vascular toxicity of local anesthetics is proportional to anesthetic potency in vivo.
However, following an editorial in 1979 [2], the question was raised whether the
newer highly protein-bound and lipid-soluble local anesthetic drugs (bupivacaine
and etidocaine) were more cardiotoxic than other local anesthetics. Reports of serious
cardiovascular complications, including death of more than 20 patients, following the
use of bupivacaine and etidocaine in clinical doses were published [31, 43]. In some
patients, cardiac arrest was preceded by convulsions, in others, not. Several experi-
mental animal studies were carried out [31], some of which excellently simulated
human conditions [28]. The conclusion from these studies is that bupivacaine ap-
pears to be more cardiotoxic than lidocaine, and that this toxicity seems to be ag-
gravated by antecedent hypoxia-acidemia as well as by pregnancy.
Sudden cardiovascular collapse (ventricular fibrillation or tachycardia, cardiac
asystole, complete heart block) seen in some patients following regional anesthesia
with use of bupivacaine has been presumed to be caused by accidental intravascular
injection in most of the cases. However, intravascular injection has not been docu-
mented.
We found that thoracic epidural anesthesia induced with bupivacaine has sig-
nificant cardiac electrophysiological and hemodynamic effects, mainly caused by
blockade of the cardiac sympathetic nerves. However, the contributory effect of
bupivacaine absorbed systemically could not be ruled out. Therefore, in the same ex-
perimental animal model, we investigated the cardiac, electrophysiological, and
hemodynamic effects of bupivacaine at various plasma levels, especially plasma
levels in the range normally seen during clinical regional anesthesia [22]. After intra-
venous injection of bupivacaine, plasma concentrations below 1000 ng/ml had no sig-
nificant electrophysiological (Fig. 9) or hemodynamic effects. During thoracic epi-
dural anesthesia, we found that plasma concentrations of bupivacaine averaged 239
ng/ml in our dogs at the time of maximal electrophysiological effects. This indicates
that systemic responses to absorbed bupivacaine do not contribute to the cardiac
electrophysiological effects seen during this type of regional anesthesia.
At a plasma level of about 2000 ng/ml, a level occasionally achieved during other
forms of regional anesthesia in man in the absence of intravascular injection, bupi-
vacaine prolonged impulse conduction time in all parts of the heart, prolonged atrial
and AV nodal refractoriness, decreased left ventricular inotropy, but had no effect
on ventricular refractoriness or monophasic action potential duration [22]. These ef-
fects may enhance susceptibility to reentrant arrhythmias in predisposed patients.
46 R.Hotvedt and H.Refsum

ms

__--..-d-to " .O: -: : - : -: -~- - - - - +. . .- -;. . . . ,

:: 1:J01-"
25

20

15
L';AH
10

Fig. 9. Changes in intraatrial

(L1PA and L1StA intervals), AV
nodal (L1AH interval) and His-
Purkinje (L1HV interval) impulse
conduction times, and L1 QRS width
15i
10
at different plasma concentrations
ofbupivacaine. Values are mean ±

:-~------
L';QRS SEM. 0---0, Pacing at 230
beats min-I, n = 6; 0 - - 0 ,
spontaneous heart rate, n = 8;
I I I i
* P<O.OS [from 22]
o 1000 2000 3000
ng/ml

Bupivacaine induced arrhythmias have been associated with signs of depressed

cardiac conduction such as wide QRS complexes in the ECG [28], and in vivo as well
as in vitro studies [8] indicate that block of cardiac sodium channels may be an im-
portant mechanism underlying bupivacaine toxicity.

The Modulated Receptor Hypothesis

Depression of impulse conduction by local anesthetics in both nerve cells and cardiac
cells results from block of sodium channels. The blocking effect is both time and volt-
age dependent, i.e., block of sodium channels increases as the stimulation rate is in-
creased or membrane potential becomes more depolarized. The dependence upon
rate and membrane potential results from the fact that the blocking effect increases
in a time-dependent manner whenever sodium channels are inactivated, and de-
creases when the membrane potential is repolarized to potentials where channels be-
come rested. Such behavior can be described adequately by the modulated receptor
Cardiac and Cerebral Effects of Local Anesthetics 47

Fig. 10. The modulated receptor hypothesis. The 3 states of the Na channel are depicted (R = rested;
A = activated; f = inactivated). Transitions that occur with Hodgkin-Huxley kinetics (HH) are also
labelled. Transition to drug-bound (blocked) states are indicated by R', A', I'. Rate constants of
association and dissociation are indicated by K and f. Representative values for the rate constants
for lidocaine are: KR = O·4ms- I M- I; KA = 5 x 104 ms - I M-\ KJ = 50ms- I M- I ; h = 1ms- l ; fA =
1· 5 ms-I; h = 2 x 1O-3 ms -1 [from 17]

hypothesis [14,16]. According to this hypothesis (Fig. 10), the sodium channel block
results from interaction of drugs with a single specific receptor site associated with
the sodium channel. Drug affinity for this receptor site is dependent upon the state
of the channel, i.e., the rate constants defining association and dissociation of drug
from its receptor are different for each channel state and for different drugs. Drug-
associated channels do not conduct ions and they behave as if their voltage depen-
dence for inactivation is shifted to more negative potentials [8].

Medullary Effects of Local Anesthetics

Recently, an alternative hypothesis for the cardiovascular toxicity of local anesthetics

was proposed [48]. The authors focus on the well-known fact that the CNS can exert
dromotropic, inotropic, and chronotropic effects on the myocardium. Furthermore,
there are observations suggesting that abnormally high brain concentrations could
accompany convulsions produced by local anesthetics. Bupivacaine and lidocaine
were therefore injected at three vasomotor and cardioactive areas in the medulla of
rats. Both drugs significantly decreased mean arterial blood pressure, reduced heart
rate, and in 55% of the animals induced ventricular arrhythmias. In all the animals in
whom ventricular arrhythmias were associated with lidocaine the arrhythmias spon-
taneously reverted to normal sinus rhythm. In 50% of animals developing ventricular
arrhythmias after bupivacaine, the arrhythmias resulted in death. These experiments
suggest that the cardiovascular effects of local anesthetics also may be secondary to
effects on the CNS.

Conclusions

Local anesthetic drugs affect excitable cellular membranes, and are widely used as
antiarrhythmics as well as in different forms of local and regional anesthesia. The
48 R.Hotvedt and H.Refsum

principal target organs for their toxic effects are the CNS and the cardiovascular
system. Recent studies indicate that local anesthetics may influence cardiovascular
function by direct effect on cardiac cell membranes, by effects on cardiovascular
regulative sites in the CNS, and by interference with cardiovascular autonomic inner-
vation.

References

1. Aberg G (1972) Toxicological and local anesthetic effects of optically active isomers of two local
anesthetic compounds. Acta Pharmacol Toxieol31: 273-286
2. Albright GA (1979) Cardiac arrest following regional anesthesia with etidocaine or bupivacaine.
Anesthesiology 51: 285-287
3. Allessie MA, Bonke FIM, Schopman FJG (1977) Circus movement in rabbit atrial muscle as a
mechanism of tachycardia. III. The "leading circle" concept: a new model of circus movement
in cardiac tissue without the involvement of an anatomical obstacle. Circ Res 41 : 9-18
4. Amlie JP, Refsum H (1981) Vagus-induced changes in ventricular electrophysiology of the dog
heart with and without beta-blockade. J Cardiovasc Pharmacol 3: 1203-1210
5. Apfelbaum JL, Gross JB, Shaw LM, Spaulding BC (1984) Changes in lidocaine protein binding
may explain its increase in CNS toxicity at elevated CO2 tensions. Anesthesiology 61: A213
6. Armour JA, Wurster RD, Randall WC (1977) Cardiac reflexes. In: Randall WC (ed) Neural
regulation of the heart. Oxford University Press, New York, pp 157-186
7. Bigger JT, Jaffe CC (1971) The effect of bretyllium tosylate on the electrophysiological proper-
ties of ventricular muscle and Purkinje fibers. Am J Cardiol 27: 82-92
8. Clarkson CW, Hondeghem LM (1985) Mechanism for bupivacaine depression of cardiac con-
duction: fast block of sodium channels during the action potential with flow recovery from block
during diastole. Anesthesiology 62: 396-405
9. Coyle DE, Denson DD, Thompson GA, Myers JA, Arthur GR, Bridenbaugh PO (1984) The
influence of lactic acid on the serum protein binding of bupivacaine: species differences. Anes-
thesiology 61: 127-133
10. Fitzgerald JD (1982) The effects of beta-adrenoceptor blocking drugs on early arrhythmias in ex-
perimental and clinical myocardial ischaemia. In: Parratt JR (ed) Early arrhythmias resulting
from myocardial ischaemia: mechanisms and prevention by drugs. Macmillan, London, pp 295-
315
11. Foldes FF, Molloy R, McNall PG, Koukal LR (1960) Comparison of toxicity of intravenously
given local anesthetic agents in man. JAMA 172: 1493-1498
12. Greene HL, Werner JA, Gross BW, Sears GK, Trobaugh GB, Cobb LA (1983) Prolongation
of cardiac refractory times in man by clofilium phosphate, a new antiarrhythmic agent. Am
Heart J 106: 492-501
13. Harris AS, Estandia A, Tillotson RF (1951) Ventricular ectopic rhythms and ventricular fibrilla-
tion following cardiac sympathectomy and coronar occlusion. Am J Physiol165: 505-512
14. Hille B (1977) Local anesthetics: hydrophilic and hydrophobic pathways for the drug-receptor
reaction. J Gen PhysioI69:497-515
15. Ho RJ (1970) Radiochemical assay of long chain fatty acids using 63Ni as a tracer. Anal Biochem
36: 105-113
16. Hondeghem LM, Katzung BG (1977) Time- and voltage-dependent interactions of antiar-
rhythmic drugs with cardiac sodium channels. Biochim Biophys Acta 472:373-398
17. Hondeghem LM, Katzung BG (1980) Test of a model of antiarrhythmic drug action. Effect of
quinidine and lidocaine on myocardial conduction. Circulation 61: 1217-1224
18. Hotvedt R, Refsum H (1986) Cardiac effects of thoracic epidural morphine caused by increased
vagal activity in the dog. Acta Anaesthesiol Scand 30: 76-83
19. Hotvedt R, Platou ES, Refsum H (1983) Electrophysiological effects of thoracic epidural
analgesia in the dog heart in situ. Cardiovasc Res 17: 259-266
Cardiac and Cerebral Effects of Local Anesthetics 49

20. Hotvedt R, Platou ES, Refsum H (1984) Effects of thoracic epidural analgesia on cardiovascular
function and plasma concentration of free fatty acids and cathecholamines in the dog. Acta
Anaesthesiol Scand 28: 132-137
21. Hotvedt R, Refsum H, Platou ES (1984) Cardiac electrophysiological and hemodynamic effects
of beta-adrenoceptor blockade and thoracic epidural analgesia in the dog. Anesth Analg 63 :
817-824
22. Hotvedt R, Refsum H, Helgesen KG (1985) Cardiac electrophysiologic and hemodynamic ef-
fects related to plasma levels of bupivacaine in the dog. Anesth Analg 64: 388-394
23. Jorfeldt L, Lewis DH, LOfstrom JB, Post C (1979) Lung uptake of lidocaine in healthy volun-
teers. Acta Anaesthesiol Scand 23 : 567 - 57 4
24. Keenaghan JB, Boyes RN (1972) The tissue distribution, metabolism and excretion of lidocaine
in rats, guinea pigs, dogs and man. J Pharmacol Exp Ther 180: 454-463
25. Kent KM, Smith ER, Redwood DR, Epstein SE (1973) Electrical stability of ischemic myo-
cardium. Influences of heart rate and vagal stimulation. Circulation 47: 291-298
26. Kjekshus JK, Mjj1js OD (1973) Effect of inhibition of lipolysis on infarct size after experimental
coronary occlusion. J Clin Invest 52: 1770-1778
27. Kliks BR, Burgess MJ, Abildskov JA (1975) Influence of sympathetic tone on ventricular fibril-
lation threshold during experimental coronary occlusion. Am J Cardiol 36: 45-49
28. Kotelko DM, Shnider SM, Dailey PA, Brizgys RV, Levinson G, Shapiro WA, Koike M, Rosen
MA (1984) Bupivacaine-induced cardiac arrhythmias in the sheep. Anesthesiology 60: 10-18
29. Lown B, Verrier RL (1976) Neural activity and ventricular fibrillation. N Engl J Med 294:
1165-1170
30. Malliani A, Schwartz PJ, Zanchetti A (1980) Neural mechanisms in life-threathening ar-
rhythmias. Am Heart J 100: 705-715
31. Marx GF (1984) Cardiotoxicity of local anesthetics - the plot thickens. Anesthesiology 60: 3-5
32. Mines GR (1913) On dynamic equilibrium in the heart. Proc Physiol Soc (London) 46: 349-383
33. Mjj1js OD, Kjekshus JK, Lekven J (1974) Importance of free fatty acids as a determinant of myo-
cardial oxygen consumption and myocardial ischemic injury during norepinephrine infusion in
dogs. J Clin Invest 53 : 1290-1299
34. Myers RW, Pearlman AS, Hyman RM, Goldstein RA, Kent KM, Goldstein RE, Epstein SE
(1974) Beneficial effects of vagal stimulation and bradycardia during experimental acute myo-
cardial ischemia. Circulation 49: 943-947
35. Nademenee K, Singh BN (1982) Advances in antiarrhythmic therapy. JAMA 247:217-222
36. Opie LH, Tansey M, Kennelly BM (1977) Proposed metabolic vicious circle in patients with
large myocardial infarcts and high plasma-free-fatty-acid concentrations. Lancet 2: 890-892
37. Platia EV, Reid PR (1984) Dose-ranging studies of clofilium, an antiarrhythmic quaternary am-
monium. Clin Pharmacol Ther 35: 193-202
38. Platou ES, Refsum H (1982) Class III antiarrhythmic action in experimental atrial fibrillation
and flutter in dogs. J Cardiovasc Pharmacol 4: 839-846
39. Platou ES, Refsum H, Hotvedt R (1986) Class III antiarrhythmic action linked with positive
inotropy: antiarrhythmic, electrophysiological, and hemodynamic effects of the sea-anemone
polypeptide ATX II in the dog heart in situ. J Cardiovasc Pharmacol 8: 459-465
40. Schwartz PJ, Pagani M, Lombardi F, Malliani A, Brown AM (1973) A cardiocardiac sympatho-
vagal reflex in the cat. Circ Res 32: 215-220
41. Scott DB (1975) Evaluation of the toxicity of local anaesthetic agents in man. Br J Anaesth 47:
56-61
42. Scott DB (1984) Toxicity caused by local anaesthetic agents. Editorial. Br J Anaesth 56: 435-436
43. Scott DB (1984) Toxicity due to local anaesthetics. Ann Chir Gynaecol 73: 153-157
44. Shand DG (1984) Alpharacid glycoprotein and plasma lidocaine binding. Clin Pharmacokinet
[Suppl 1] 9:27-31
45. Sjostrand U, Widman B (1973) Distribution of bupivacaine in the rabbit under normal and
acidotic conditions. Acta Anaesthesiol Scand [Suppl] 50: 1-24
46. Sonnenblick EH, Ross J Jr, Braunwald E (1968) Oxygen consumption of the heart. Newer con-
cepts of its multifactorial determination. Am J Cardiol 22: 328-336
47. Steen PA, Michenfelder JD (1979) Neurotoxicity of anesthetics. Anesthesiology 50:437-453
48. Thomas RD, Behbehani MM, Coyle DE, Denson DD (1986) Cardiovascular toxicity of local
anesthetics: an alternative hypothesis. Anesth Analg 65: 444-450
50 R. Hotvedt and H. Refsum: Cardiac and Cerebral Effects of Local Anesthetics

49. Tucker GT, Boas RA (1971) Pharmacokinetic aspects of intravenous regional anesthesia. Anes-
thesiology 34: 538-549
50. Uchida Y, Kamisaka K, Murao S, Ueda H (1974) Mechanosensitivity of afferent cardiac sym-
pathetic nerve fibers. Am J Physiol 226: 1088-1093
51. Ueda H, Uchida Y, Kamisaka K (1969) Distribution and responses of the cardiac sympathetic
receptors to the mechanically induced circulatory changes. Jpn Heart J 10: 78-81
52. Usubiaga JE, Wikinski J, Ferrero R, Usubiaga LE, Wikinski R (1966) Local anesthetic-induced
convulsions in man - an electroencephalographic study. Anesth Analg 45 : 611-620
53. Vaughan Williams EM (1984) A classification of antiarrhythmic actions reassessed after a decade
of new drugs. J Clin Pharmacol 24: 129-147
54. Vik-Mo H, Ottesen S, Renck H (1978) Cardiac effects of thoracic epidural analgesia before and
during acute coronary artery occlusion in open-chest dogs. Scand J Clin Lab Invest 38: 737-746
55. Waxman MB, Wald RW (1977) Termination of ventricular tachycardia by an increase in cardiac
vagal drive. Circulation 56: 385-391
56. Wiklund L (1984) Cardiovascular and central nervous system toxicity of local anaesthetics. Ann
Chir Gynaecol 73: 123-130
57. Zuanetti G, De Ferrari GM, Priori SG, Schwartz PJ (1987) Protective effect of vagal stimulation
on reperfusion arrhythmias in the cat. Circ Res 61 : 429-435
Chapter 4
Neurotoxins as Tools
in Studying Cardiac Excitation-Contraction Coupling
V.RAVENS and E.WETIWER

Common Features of Brain and Heart

At first sight, the brain as top manager of the human body does not seem to have
much in common with the heart as one of its most industrious laborers. However,
both cerebral and cardiac tissue share the fundamental properties of electrical excit-
ability and of impulse conduction, which are the basis for signal transmission be-
tween and coordination of the individual cells in these two organs.
In the first section of this article, we will briefly summarize our understanding
of the electrical excitation process and how results obtained with certain toxins
applied to nervous tissue have contributed to the elucidation of these events at the
molecular level. In the second part, we shall describe experimental results of some
of our own research with neurotoxins, which we have used as tools for the study of
cardiac excitation-contraction coupling.

Excitation Process (Sodium Channels)

Most excitable cells possess a resting membrane potential of approximately - 90 mV,

which is mainly a potassium diffusion potential. Large ionic gradients for Na+ and
K+ are maintained by means of an adenosine triphosphate (ATP)-consuming pump,
the Na+-K+-ATPase. During rest, the membrane lacks an appreciable permeability
to Na+ ions. In response to an appropriate stimulation, the cell membrane suddenly
becomes highly permeable to Na+ which, flowing down their electrochemical gra-
dient, result in an ionic current that further depolarizes the cell and thus gives rise to
a propagated action potential. The membrane repolarizes because the Na+ perme-
ability declines again and K+ leave the cell.
The transient changes in membrane permeability have been explained by Hodgkin
and Huxley on the assumption of ion-specific channels in the membrane which may
open and close, respectively, in a voltage- and time-dependent manner [17, 18].
Toxic compounds from plants or animals interfere specifically with the excitation
process and thus provide useful tools for its study.
Tetrodotoxin (TTX), for example, is the toxic substance contained in the buffer
fish which, in spite of the danger of intoxication, is a well-esteemed delicacy in
Japan. (Less than 1 mg of the toxin is lethal in man!) The toxin was classified as a
neurotoxin because it inhibits propagated action potentials in nerves.
52 V.Ravens and E. Wettwer

v
B
'1
c 1amp potentia ---lI IL--ss----l Q:lms L - -
current
Fig. I A, B. Schematic explanation of the principle of the patch clamp technique developed by Neher
and his coworkers [24, 25, 34, 35]. A From left to right: tip of the glass pipette (shaded area) ap-
proaches the myocyte; establishment of the tight seal between the glass wall of the pipette and the
cell membrane by applying suction to the interior of the pipette (whole-cell configuration); mem-
brane patch removed from the cell (inside-out patch configuration). B Reconstruction of typical
signals of sodium channel openings obtained with the patch clamp technique in the whole-cell con-
figuration

Tetrodotoxin was first investigated by Hille [15] in the node of Ranvier with the
voltage clamp technique. This electrophysiological method allows the study of the
size and the time course of membrane currents in response to a membrane potential
controlled by an electronic feedback circuit. ITX completely blocked the inwardly
directed current component that depends on the extracellular Na+ concentration
without affecting the outwardly directed K+ current. Therefore, the current compo-
nents flowing during the excitation process could be separated by the application of
ITX. This finding strongly confirmed the concept of Hodgkin and Huxley [18] that
Na+ and K+ pass through independent membrane channels. The effect is a very
specific one; 1 nmolllITX is sufficient to block half of all neuronal sodium channels
[15].
The concept of membrane channels is further supported by the recently developed
technique of patch clamping [24, 25, 34, 35]; a schematic presentation of the princi-
ple is given in Fig. 1. A glass pipette with a rounded-off wall at the opening of 1-2JlIIl
in diameter is brought within the close vicinity of the membrane of an isolated cell.
When it just touches the membrane, a small negative pressure (suction) is applied to
the interior of the pipette so that the membrane is sucked gently into the tip aperture
and forms a tight seal ("giga-ohm seal") with the glass wall. The small "patch" of
membrane under investigation lends its name to the method. For investigation, it
may be left within or tom out of the cell membrane. With appropriate saline solution
in the pipette sudden conductance changes are recorded at different patch voltages.
These are interpreted as openings and closings of single membrane channels. Besides
channels for conducting Na+, multiple other types of channels have been discovered,
estimated by some authors to a total number of about 50 [16].
Neurotoxins as Tools in Studying Cardiac Excitation-Contraction Coupling 53

A B

'- Inac tivation gate

Fig. 2 A, B. Model of the sodium channel. A A large protein (dotted area) spans the phospholipid
barrier of the plasmalemma. The selectivity filter controls the ion species that can pass and lies with-
in the hydrophilic pore. At its intracellular side, the channel possesses an "inactivation gate" which
is responsible for the closing of the channel at maintained depolarization (inactivation). This subunit
of the channel can be cleaved selectively with pronase [1] . Negative signs mark fixed surface charges.
B Location of different binding sites for various neurotoxins (white areas) at the sodium channel.
1, Tetrodotoxin (lTX); 2, lipid soluble toxins such as batrachotoxin (BTX); 3, polypeptide toxins
modifying inactivation (sea anemone toxins, ATX, and a-scorpion toxins, ScTX); 4, polypeptide
toxins modifying activation ("J3-scorpion toxins"); the site for local anesthetics is marked with a posi-
tive charge (Redrawn after Hille [16])

Sodium channels may also be characterized by means of various neurotoxins

which have characteristic effects on channel "gating," i.e., on the activation and in-
activation processes. Four separate binding sites have been described for these toxins
(Fig. 2) . They include site 1 for guanidinium toxins that block at the selectivity filter,
e.g., TTX or saxitoxin; site 2 for lipid soluble toxins, e.g., batrachotoxin, veratridine,
aconitine, gryanotoxins, pyrethroids; site 3 for polypeptide toxins, e.g., sea anemone
toxins or North African a-scorpion toxins; and site 4 also for polypeptides, e.g.,
American fJ-scorpion toxins [7, 11, 16,21,36]. Some of these toxins bind with such
high affinity and specificity that their radio labeled derivatives have been employed
in the isolation and purification of channel molecules (for review, see Catterall [8]).
When purified channel molecules are reconstituted into phospholipid membranes,
voltage-dependent conductance changes are observed just as in natural membranes
from excitable cells [13, 23] .
According to our current knowledge, the sodium channel of vertebrate brain or
muscle cells consists of one large membrane-spanning glycoprotein, the a-unit, and
two smaller units, the /31 and Ih. units, which are exposed to the extracellular surface
of the membrane [8, 16]. The a-unit contains four homologous subunits composed of
six a-helical segments and forms the major ion-conducting part of the channel (for
review, see Catterall [8]).
Recently, the primary structure of the sodium channel protein from Electrophorus
electricus and from rat brain has been elucidated [27, 28]. Using cloning techniques,
it is also possible to transcribe specific mRNAs of the sodium channel from cloned
cDNA. When these mRNAs are injected into large oocytes from Xenopus, they in-
duce the synthesis of sodium channels which are incorporated into the cell mem-
brane of the oocyte. This ingenious experimental approach not only provides a
model for genetic expression of sodium channels but at the same time also supplies a
convenient system to study their electrophysiological properties [37, 38].
54 U. Ravens and E. Wettwer

Neurotoxins Acting ou Cardiac Sodium Channels

Excitability of cells from both the brain and the heart depends on the existence of
sodium channels. However, these channels are not identical in both tissues as can be
judged from differences in their sensitivity to certain neurotoxins.
Tetrodotoxin is also effective in cardiac cells (Fig. 3A). It has little effect on the
overall shape of the action potential, but depresses the maximum rate of depolariza-
tion which serves as an indirect estimate of the Na+ current [9]. In the presence of
3 ttmolll TTX, the maximum rate of depolarization decreases from 270 Vis at control
to 130V/s. Since the general shape of the action potential does not change much, it
may be assumed that TTX neither affects the K+ current nor the Ca2 + current, i.e.,

lOOms 5ms
A

5mN BOVls

Control TTX (3 ~ mol/I)

B lOOms

Control ATX II (20nmolltl

Fig. 3. A Effects of tetrodotoxin (3llmolll) in a papillary muscle from the guinea pig heart. The left
side of each frame shows action potentials (upper trace) and force of contraction (lower trace); the
right side contains recordings of the rapid depolarization phase of the action potential at an ex-
panded time scale (upper trace) and the first derivative of voltage against time (lower trace). Calibra-
tions are as indicated. Experimental conditions: frequency of stimulation, 1 Hz; temperature, 35°C;
modified Tyrode solution containing Ca2+ 1.8mmolll, K+ 5.4mmolll; pH 7.4. Action potentials were
recorded with conventional glass microelectrodes containing KCI, 3molll; isometric contractions
were recorded via a Statham UC-2 force transducer. B Effect of ATX II on electrical and mechanical
activity in guinea pig papillary muscle. Original recordings of action potentials (upper trace) and
force of contraction (lower trace) after an initial equilibration period of 90 min (control) and after
60 min of exposure to A TX II, 20 nmolll (right frame). Experimental conditions as in A; the 50 mV
calibration bar is also valid for A
Neurotoxins as Tools in Studying Cardiac Excitation-Contraction Coupling 55

[
ATXII (20 nmolll)
mV
Control
r-~~':":";'''':;''':'-~-''
ATX II (20nmoll\) + TT X (60 ~m ol /l )
40

o r+--~~----~
- 40

-80~~~==:::::::::::::::l
o 80 160 ~ 320 o 80 160 240 320 o 80 160 240 320 ms

Fig. 4. Effects of A TX II in cardiac myocytes. Upper part: myocyte impaled by a patch pipette (dark
shadow). Scale bar = 20 JlIll. Lower part from left to right: action potentials under control condi-
tions; in the presence of ATX II, 20nmolll; after addition of ATX II and tetrodotoxin, 60,umolll.
Calibrations are as indicated . Myocytes were isolated according to Isenberg and Klockner [19) and
investigated in a small perfusion chamber with modified tyrode solution [for experimental details,
see 4, 10,20, 31). Stimulation frequency, 1 Hz; filling solution for the patch electrode, K glutamate
140mmolll adjusted with morpholinopropane sulfonic acid (MOPS)/KOH to pH 7.2; temperature
35°C

the second inward current characteristic for cardiac cells. In conclusion, TTX also in-
hibits the Na+ channels of heart muscle; however, higher concentrations than in
nerve [15] are required. The selectivity of TTX for sodium channels in fact justifies
the conclusion that sodium channels are involved if a biological effect is reversed in
the presence of TTX. Although TTX is of no therapeutic value, it nevertheless has
proven a useful pharmacological tool.
While tetrodotoxin inhibits the flow of Na+ current, other neurotoxins such as
various alkaloids (e.g., ceveratrum alkaloids, batrachotoxin, aconitine) or polypep-
tides (from scorpions, sea anemones, corals) promote the entry of Na+ into the cells.
Of the neurotoxins that enhance the entry of Na+ via channels, we have chosen the
sea anemone toxin A TX II for further study. In shore crabs this polypeptide com-
pound causes repetitive firing in motor nerves leading to spastic paralysis of the ani-
mal [3, 29]. We have shown previously that ATX II prolongs the action potential
duration in isolated papillary muscle of the guinea pig heart [2, 30]. Similar effects
are observed in cardiac myocytes. When single cells isolated from the heart by en-
zymatic dispersion [5] are exposed to ATX II (Fig. 4), the action potential duration
56 V.Ravens and E. Wettwer

ATX II +
Control ATXII (100nmolll) TTX (60~mol/l)
nA
n~:J -f'- 0
A -2 -2
-4 -4
-6 -85--50mV -65 --50mV -6
-85 --50mV
-8 -8
0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 IToS

OPI+
Control OPI (3j.1mol/l) TTX (60 j.lmol/t)
nA nA
0
+5
B 0 r -5
r
-5 -85 - -20mV
-83 - -15mV
-10 -10
lOilrii"s lOOms 20ms

Fig. 5. A Effect of A TX II on membrane currents in a cardiac myocyte (guinea pig). Membrane cur-
rents were recorded in response to a 250ms long depolarizing clamp pulse from -85 to -50mV
(only initial40ms after depolarization are shown) in normal Tyrode solution (left frame). After addi-
tion of ATX II, 100nmolll (middle frame), net membrane current remains inwardly directed during
the whole pulse (0.3nA at 250ms). Exposure to tetrodotoxin, TTX, 6OJIIIlolll, reverses this effect
(right frame). Single electrode voltage clamp technique with a patch electrode (Hamill et al. [12]).
B Effects of OP! 201-106 on membrane currents in cardiomyocytes (guinea pig). Net membrane cur-
rent in response to a 320ms long clamp step from -83 to -15mV in normal Tyrode solution (left
frame) and in the presence of OP! 201-106, 3 JIIIlolll (middle frame); right frame, current tracing
from a different cell in response to 160ms long clamp steps from -85 to -20mV in the presence of
OP! 201-106 (3 JIIIlolll) alone and with both OP! 201-106 and tetrodotoxin (60 JIIIlolll); calcium-free,
nickel-containing medium (Ni2+, 3.6mmolll) to suppress calcium currents (Buggisch et al. [6])

is markedly prolonged, and this effect is readily reversible after addition of TIX in-
dicating that sodium channels are involved. In heart cells the effective concentration
of A TX II is 100 times lower than in nerve cells [20].
Figure 5A shows net membrane currents in response to a depolarizing clamp step
from a holding potential near the resting membrane potential of -85 mV to - 50 mY.
Under control conditions net membrane current is inwardly directed during the first
few milliseconds, inactivating rapidly. After exposure to ATX II, a slowly decaying
net inward current occurs which disappears again in the presence of TIX. These
findings suggest that A TX II slows down the closing process of the sodium channel.
The delayed flow of inward current explains the prolongation in action potential
duration.
Schreibmayer and co-workers [33] described the effects of A TX II at the single
channel level; the result of one of their experiments is depicted in Fig. 6. The cell-
attached patch was depolarized by 70mV from a holding potential20mV negative to
the resting potential. In the control traces, brief openings are observed in the begin-
ning of the clamp step. With ATX II in the pipette solution, the average channel
openings are much longer and the channels may even reopen again after closing.
Neurotoxins as Tools in Studying Cardiac Excitation-Contraction Coupling 57

Control ATX II (SOnmol/l

12 """'1.......,......,,'II<M~~N.J.<...,.,""'*'~.,.....,vIJo.A~

13~·
14 3

15
16

17

18

19 8~~~~~~~~

20 9f#o.~~
21 10

3 pA ]
24 ms

Fig. 6. Single-channel recordings of the effects of A TX II on the sodium channel in rat cardiomyo-
cytes. Activation of openings of single sodium channels following suprathreshold voltage pulses from
a potential of 20mV negative to the resting potential (RP) to 50mV positive to the RP. Left column,
control; right column, pipette solution contains 50nmolll ATX II (plus BaCh, Immolll, to block
potassium channels). Current and time calibrations are as indicated by the bars. Cell-attached mem-
brane patch, analogue low-pass filtering at 1.8 kHz; for further experimental details, see [33]. (Cour-
tesy of Schreibmayer, unpublished experiment)

This behavior of the channels explains the inhibition of inactivation of the macro-
scopic sodium current on a molecular level.

Neurotoxins Influencing Cardiac Contraction

Electrical excitation of the heart muscle initiates its mechanical function and Ca 2 +
ions are required for activation of the contractile proteins. However, Na+ ions may
indirectly influence cardiac contraction because they are involved in the maintenance
of the intracellular Ca2 + homeostasis via the Na/Ca exchange. The activity of this
transport system is dependent on the transmembrane concentration gradients for
Na + and Ca2 + and on membrane potential (for review, see Noble [26]). Since sodium
channels control the entry of Na+ into the cell, their modification is expected to in-
fluence contractile force.
58 U. Ravens and E. Wettwer

Fig. 7. Diagram of the pro-

posed sequence of events in
the action of the neurotoxin
ATXII

When ATX II is added to an isolated cell (or a multicellular preparation,

Fig. 3B), the prolongation in action potential duration is indeed accompanied by an
increase in contractility [30]. However, long-lasting depolarizations may by them-
selves enhance contractility. To establish that the ATX II induced increase in con-
tractility is related to the increase in Na + entry, we carried out the following experi-
ment (for details , see also Ravens and Wettwer [31]): ATX II was applied to a single
myocyte depolarized at regular intervals with voltage clamp steps of constant dura-
tion (320 ms) from a holding potential near the resting membrane potential to -15
mY. In this experiment the duration of depolarization could not change because it
was kept constant. After exposure to ATX II, contractility was enhanced all the
same. Therefore , prolongation of the action potential per se cannot be the only
determinant of the contractility enhancing effect, but an additional mechanism must
exist which is probably related to the cellular sodium load .
The suggested sequence of events in the action of A TX II is illustrated schemati-
cally in Fig. 7. Sodium ions that enter the cell with the action potential prolonged by
ATX II decrease the concentration gradient for Na+ across the membrane and there-
fore reduce the driving force for the Na/Ca exchange. The exchange carrier con-
sequently transports less Ca2+ out of the cell. The resulting elevation of the cytosolic
Ca2+ concentration will enhance the filling of cellular calcium stores from where it
may be released to increase the amplitude of subsequent contractions.
The positive inotropic effect of ATX II could be beneficial in the failing heart.
Unfortunately, the potential use of A TX II for the treatment of heart failure is ham-
pered by its polypeptide structure , which is liable to cause immunological reaction.
However, smaller molecules with a similar mechanism of action may prove useful.

DPI 201-106 as an Example of Drug Development from This Basic Research

Recently, the novel compound DPI 201-106 (for structural formula , see Fig. 8) was
synthesized, which also affects the inactivation of the sodium channels . Like ATX II,
Neurotoxins as Tools in Studying Cardiac Excitation-Contraction Coupling 59

OH
I ,.......,
O-CH2-CH-CHrN.....jI-CH
9
...
1

6i H
CN 6
DPI 201-106
Fig. 8. Chemical structure and effects of DPI 201-106

lL J'\~ 1L [~mv on action potentials and contractions of an isolated

cardiomyocyte (guinea pig). Action potentials (upper
traces) and unloaded shortening (lower traces) mea-
sured with an optoelectrical device. Left recording,
control; middle, after 5 min of exposure to DPI 201-
106, 1 pmolll; right, after 2 min of exposure to DPI
Control 201-106, 3 pmolli. (For experimental details, see [31])

the new compound D PI 201-106 prolongs the action potential and enhances the force
of contraction in multicellular heart tissue [6] as well as in single cells (Fig. 8). Both
effects are blocked by TIX and are therefore related to the sodium channels (not
shown).
The effect of DPI 201-106 on membrane currents is shown in Fig. 5B. DPI 201-
106 also induces a slowly decaying inward current in response to a depolarizing
clamp step from -80 to + 15 mY. Compared with the ATX II-induced retardation of
Na+ current inactivation, the effect of DPI 201-106 (3 llmol/l) is much larger. TIX
(60 llmol/l) blocks the effects on membrane currents confirming the involvement of
sodium channels. The single-channel recordings of the effects of DPI 201-106 show
similar prolonged openings and also reopenings of the channels as observed with
ATX II [22].
All these results taken together strongly suggest that DPI 201-106 also increases
the sodium load of the cell and enhances the force of contraction by a similar mecha-
nism as outlined for ATX II. However, it should be pointed out that DPI 201-106
also sensitizes the contractile proteins for Ca2 + and that this effect may contribute to
the positive inotropic action of the agent [14, 32]. First investigations in human vol-
unteers have shown that the drug is well tolerated, and preliminary results from pa-
tients demonstrate that the compound is effective in heart failure [39].

Conclusions

The brain and the heart share the manifestation of excitability which is linked to the
function of sodium channels. The physiological properties of the sodium channels,
e.g., the rapid opening in the range of milliseconds upon depolarization or the volt-
age dependence of the gating processes, are similar in heart and nerve cells. How-
ever, their pharmacological properties differ: cardiac sodium channels are less sensi-
tive to the blocking action of tetrodotoxin than neuronal ones, whereas they respond
to lower concentrations of sea anemone toxins. Compounds primarily discovered as
neurotoxins have provided valuable information about the excitation-contraction
coupling in heart muscle. Besides being useful pharmacological tools, the study of
60 V.Ravens and E. Wettwer

their mechanism of action has even led to the discovery of new compounds with a
potential for therapeutic use.

Acknowledgments. This work was supported by the Deutsche Forschungsgemeinschaft (Ra 222/6).
The authors gratefully acknowledge the skillful technical assistance of Ms. Barbara Langer and Iris
Manthey. Part of the experimental work in isolated cells was carried out in the laboratory of Prof.
Dr. Gerrit Isenberg whom we wish to thank for his advice and criticism.

References

1. Armstrong CM, Bezani1la F, Rojas E (1973) Destruction of sodium conductance inactivation in

squid axons perfused with pronase. J Gen Physiol 62: 375-391
2. Beress L, Ritter R, Ravens V (1982) The influence of the rate of electrical stimulation on the
effects of the Anemonia sulcata toxin ATX II in guinea pig papillary muscle. Eur J Pharmacol
79:265-272
3. Beress L, Beress R (1971) Reinigung zweier krabbenliihmender Toxine aus der Seeanemone
Anemonia sulcata. Kieler Meeresforsch 27: 117-127
4. Borchard M, Ravens V (1986) Intracellularly applied sodium mimics the effects of ouabain in
single cardiac myocytes. Eur J Pharmacol131 : 269-272
5. Bendukidze Z, Isenberg G, Klockner V (1985) Ca++-tolerant guinea-pig ventricular myocytes
as isolated by pronase in the presence of 250 JlM free calcium. Basic Res Cardiol [Suppl] 80:
13-18
6. Buggisch D, Isenberg G, Ravens V, Scholtysik G (1985) The role of sodium channels in the ef-
fects of the cardiotonic compound DPI 201-1()6 on contractility and membrane potentials in iso-
lated mammalian heart preparations. Eur J PharmacoI118:303-311
7. Catterall WA (1980) Neurotoxins that act on voltage-sensitive sodium channels in excitable
membranes. Ann Rev Pharmacol Toxicol20: 15-43
8. Catterall WA (1986) Structure of voltage-sensitive sodium and calcium channels. In: Liittgau
HC (ed) Membrane control of cellular activity. Gustav Fischer, Stuttgart, pp 3-27
9. Cohen CJ, Bean BP, Tsien RW (1984) Maximal upstroke velocity as an index of available
sodium conductance. Comparison of maximal upstroke velocity and voltage clamp measure-
ments of sodium current in rabbit Purkinje fibers. Circ Res 54: 636-651
10. Cragoe E, Ravens V, Wettwer E (1987) Amiloride derivatives that block Na/Ca exchange in-
hibit spontaneous inward currents in sodium loaded cardiac myocytes. Eur J Pharmacol 140:
113-116
11. Darbon H, Jover E, Couraud F, Rochat H (1983) Photoaffinity labeling of a- and {J-scorpion
toxin receptors associated with rat brain sodium channels. Biochem Biophys Res Commun 115:
415-422
12. Hamill OP, Marty A, Neher E, Sakmann B, Sigworth F (1981) Improved patch-clamp tech-
niques for high-resolution current recording from cells and cell-free membrane patches. Pfliigers
Arch 391 : 85-100
13. Hartshorne RP, Keller BV, Talvenheimo JA, Montal M (1984) Functional reconstitution of the
purified brain sodium channel in planar lipid bilayers. Proc Nat! Acad Sci VSA 82: 240-244
14. Herzig JW, Quast V (1984) Increase in Ca++ sensitivity of myocardial contractile structures by
DPI 201-106. J Mol Cell Cardiol [Suppl 3]16: 6
15. Hille B (1966) Common mode of action of three agents that decrease the transient change in
sodium permeability in nerves. Nature 210: 1220-1222
16. Hille B (1984) Modifiers of gating. In: Ionic channels of excitable membranes. Sinauer, Sunder-
land, pp 303-328
17. Hodgkin AL, Huxley AF (1952a) Currents carried by sodium and potassium ions through the
membrane of the giant axon of Loligo. J Physiol (Lond) 116: 449-472
18. Hodgkin AL, Huxley AF (1952b) A quantitative description of membrane current and its appli-
cation to conduction and excitation in nerve. J Physiol (Lond) 117: 500-544
Neurotoxins as Tools in Studying Cardiac Excitation-Contraction Coupling 61

19. Isenberg G, Klockner U (1982) Calcium tolerant ventricular myocytes prepared by preincuba-
tion in a "KB medium". Pfliigers Arch 395:30-41
20. Isenberg G, Ravens U (1984) The effects of the Anemonia sulcata toxin (ATX IT) on membrane
currents of isolated mammalian myocytes. I Physiol (Lond) 357: 127-149
21. lover E, Couraud F, Rochat H (1980) Two types of scorpion neurotoxins characterized by their
binding to two separate receptor sites on rat brain synaptosomes. Biochem Biophys Res Com-
mUll 95:1697-1714
22. Kohlhardt M, Frobe U, Herzig JW (1986) Modification of single cardiac Na+ channels by DPI
201-106. I Membrane Bioi 89: 163-172
23. Krueger BK, Worley IF III, French IR (1983) Single sodium channels from rat brain incorporated
into planar lipid bilayer membranes. Nature 303: 172-175
24. Neher E, Sakmann B (1976) Single-channel currents recorded from membrane of denervated
frog muscle fibres. Nature 260: 779-802
25. Neher E, Sakmann B, Steinbach IH (1978) The extracellular patch clamp: A method for resolv-
ing currents through individual open channels in biological membranes. Pfliigers Arch 375:
219-228
26. Noble D (1985) Sodium-calcium exchange and its role in generating electric current. In: Nathan
RD (ed) Cardiac muscle: the regulation of excitation and contraction. Academic, London, pp
171-199
27. Noda M, Shimizu S, Tanabe T, Takai T, Kayano T, Ikeda T, Takahashi H, Nakayama H,
Kanoaka Y, Minamino N, Kangawa K, Matsuo H, Raftery MA, Hirose T, Imayama S, Haya-
shida H, Miyata T, Numa S (1984) Primary structure of Electrophorus electricus sodium channel
deduced from cDNA sequence. Nature 312: 121-127
28. Noda M, Ikeda T, Kayano T, Suzuki H, Takeshima H, Kurasaki M, Takahashi H, Numa S
(1986) Existence of distinct sodium channel messenger RNAs in rat brain. Nature 320: 188-192
29. Rathmayer W, lessen B, Beress L (1975) Effect of toxins from sea anemones on neuromuscular
transmission. Naturwissenschaften 62: 538
30. Ravens U (1976) Electromechanical studies of an Anemonia sulcata toxin in mammalian cardiac
muscle. Naunyn-Schmiedeberg's Arch PharmacoI296:73-78
31. Ravens U, Wettwer E (1988) Pharmacological studies with cardiac myocytes. Agents that influ-
ence contraction. In: Piper HM, Isenberg G (eds) Isolated adult cardiac myocytes. CRC Press,
Boca Raton (in press)
32. SchoJtysik G, Salzmann R, Berthold R, Herzig IW, Quast U, Markstein R (1985) DPI 201-206,
a novel cardioactive agent. Combination of cAMP-independent positive inotropic, negative
chronotropic, action potential prolonging and coronary dilatory properties. Naunyn-Schmiede-
berg's Arch Pharmacol329: 316-325
33. Schreibmayer W, Kazerani H, Tritthart H (1987) A mechanistic interpretation of the action of
toxin II from Anemonia sulcata on the cardiac sodium channel. Biochim Biophys Acta (in press)
34. Sigworth FI (1986) The patch clamp is more useful than anyone had expected. Fed Proc 45:
2673-2677
35. Sigworth FI, Neher E (1980) Single Na+ channel currents observed in cultured rat muscle cells.
Nature 287:447-449
36. Strichartz G, Rando T, Wang GK (1987) An integrated view of the molecular toxinology of
sodium channel gating in excitable cells. Annu Rev Neurosci 10: 237-267
37. Sumikawa K, Houghton M, Emtage IS, Richards BM, Barnard EA (1981) Active multisubunit
ACh receptor assembled by translation of heterologous mRNA in Xenopus oocytes. Nature
292:862-864
38. Sumikawa K, Parker I, Miledi R (1986) Xenopus oocytes as a tool for molecular cloning of the
genes coding for neurotransmitter receptors and voltage-operated channels. In: Liittgau HC
(ed) Membrane control of cellular activity. Fischer, Stuttgart, pp 127-139
39. Thormann I, Kramer W, Kindler M, Kremer FP, Schlepper M (1986) Comparative efficacy of
the new cardiotonic agent DPI 201-106 versus dobutamine in dilated cardiomyopathy: Analysis
by serial pressure/volume relations and "on-line" MV02 assessment. I Cardiovasc Pharmacol
8:749-757
Chapter 5
Adenosine and ATP Interactions
with Autonomic Neural Control of the Heart
A.PELLEG

Introduction

The purine nucleoside adenosine and its nucleotide adenosine 5' -triphosphate (ATP)
are biological compounds found in every cell of the human body. Almost five de-
cades ago Drury and Szent-Gyorgyi [34] demonstrated that in addition to their well-
established role in cellular metabolism, adenosine and related compounds exert pro-
nounced effects on the mammalian heart. These include coronary vasodilation,
negative chronotropic effects on cardiac pacemakers, negative dromotropic effects
on atrioventricular (AV) nodal conduction, and negative inotropic effect on myo-
cardial contractility (for recent reviews, see [8, 11, 103].
The electrophysiological effects of adenosine and A TP have led to their use as
antiarrhythmic agents for the acute management of paroxysmal supraventricular
tachycardia involving the AV node [10, 11]. In addition, it has recently been sug-
gested that adenosine could be effective in terminating a certain type ventricular
tachycardia observed in the clinical setting [64, 65]. However, since adenosine and
ATP are released from myocardial cells under physiological and pathophysiological
conditions, it has been suggested that adenosine and possibly also ATP are involved
in the mechanism of certain cardiac arrhythmias characterized by sinus node dys-
function and/or AV nodal conduction disturbances [6,112]. Indeed, recent observa-
tions in the clinical setting support this hypothesis, at least with regard to the AV
node [99, 113].
The mechanisms of the cardiac actions of adenosine and ATP have been studied
extensively in the last decade. Important information has been gathered with regard
to the receptors which mediate these actions and the postreceptor reaction cascade
leading to alterations of cell membrane characteristics. In addition, data have been
obtained suggesting a complex interaction between adenosine and ATP and the auto-
nomic nervous system. This interaction involves pre- and postsynaptic effects which
can modulate the autonomic neural control of the heart. Thus, it could be hypo-
thesized that the physiological roles of adenosine and ATP released by the heart in-
clude modulation of both sympathetic and parasympathetic input to the heart. The
present chapter reviews the literature dealing with studies which give support to this
hypothesis.
Adenosine, ATP and Cardiac Neural Control 63

Adenosine-Vagal Interaction

Since the original work of Brown and Eccles [16], many studies have established that
dynamic vagal control of sinus node pacemaker activity depends mainly on the tem-
poral relationship between vagal discharge and the pacemaker period [32, 59, 66].
Thus, factors which affect either vagal input to the sinus node or sinus cycle length
(or both) can potentially modulate the relationship between sinus pacemaker activity
and vagal input to the sinus node. Adenosine, which exerts a pronounced direct
negative chronotropic action on the pacemaker activity of the sinus node, is there-
fore a prime candidate for a biological vagal modulation.
In their early study, Drury and Szent-Gyorgyi [34] found that in the cat adenosine
induced early and delayed bradycardia. The latter, which they termed "secondary
bradycardia," was abolished by atropine. In 1938 Drury et al. [33] reported that
physostigmine occasionally potentiated the depressant effect of adenosine on AV
nodal conduction in the guinea pig. This finding, which was confirmed recently [80],
was interpreted by Drury et al. as suggesting modulation by vagal tone of the ani-
mal's "sensitivity" to adenosine [33]. Almost 30 years later a brief anecdotal report
by Rossi et al. [91] described accentuation of vagal bradycardia by adenosine in the
rat heart. Data obtained more recently further implicated the vagus in the electro-
physiological action of adenosine. For example, the negative dromotropic action of
adenosine on the canine AV node under conditions of chloralose anesthesia was sig-
nificantly more pronounced [72] than under pentobarbital anesthesia [85], which is
known to suppress vagal tone [73]. In addition, atropine significantly attenuated the
effect of adenosine under the former but not the latter conditions [85]. These data
suggest either additive or potentiating interaction between adenosine and the vagus.
Indeed, in recent studies, it was found that infusion of 2-chloroadenosine, a potent
analogue of adenosine, into the sinus nodal artery consistently caused significant en-
hancement of the negative chronotropic action of vagus nerve stimulation [82]. In
addition, elevated vagal tone (simulated by continuous right vagus nerve stimula-
tion) resulted in either accentuation or attenuation of the negative chronotropic ac-
tion of exogenous adenosine depending on the degree of vagal tone elevation (i.e.,
depending on the intensity of vagal stimulation as well as the phase of the sinus cycle
length at which it was performed) [82].
Since adenosine is released from myocardial cells in the presence of variable
background vagal tone, these findings suggest that the depressant action of adenosine
on sinus node can be modulated by the parasympathetic input to the heart.
The cellular mechanisms which mediate adenosine-vagus interaction are not fully
known, and only one preliminary report dealing with this topic has been published
so far [79]. In this study, which employed postganglionic vagal stimulation, it was
found that adenosine augmented vagal-induced lengthening of sinus cycle length.
The predominant mechanism responsible for this appeared to be a decrease in dia-
stolic depolarization [79]. Furthermore, there was no evidence for adenosine-depen-
dent attenuation of vagal effects in these cells [79].
Different results with respect to adenosine-vagus interaction were obtained by
Monteiro and Ribeiro [70]. They have recently reported that infusion of adenosine
in anesthetized cats and rats attenuated the bradycardiac effect induced by vagal
stimulation in both animals [70]. These results were interpreted by Monteiro and
64 A.Pelleg

Ribeiro as evidence for adenosine inhibition of acetylcholine release from vagal

nerve terminals [70]. While this preliminary report did not exclude other mecha-
nisms, the modulation of acetylcholine release by adenosine in vivo is a viable possi-
bility. Indeed, studies with preparations other than myocardial tissue have indicated
that adenosine can attenuate the release of acetylcholine from nerve terminals [50,
53, 101]. However, at least one study has shown potentiation by adenosine of acetyl-
choline effects in extracardiac tissue [49].
Thus, it seems that adenosine can exert dual modulating effects on vagal-
mediated actions: a prejunctional inhibition due to attenuation of acetylcholine re-
lease from nerve terminals and postjunctional potentiation due to an as yet unknown
mechanism [49]. With regard to the latter, it is interesting to note that ATP was re-
ported to increase acetylcholine receptor sensitivity in different preparations [1, 2,
42]. Another factor which could be involved in the effects of adenosine on cholin-
ergic response is the action of adenosine on choline acetyltransferase [116]. Hypo-
thetically, enhanced activity of this enzyme could influence the amount of acetyl-
choline available for release. It seems, therefore, that adenosine-vagal interaction is
a complicated process. It depends to a large extent on the dose and mode of adminis-
tration of adenosine which directly affect adenosine concentration at its different
functional compartments.
The issue of adenosine-vagal interaction becomes more complicated when the
nucleoside is infused over a prolonged period of time into a peripheral vein. This is
due, at least in part, to various actions of adenosine on different organs and tissues
throughout the body which trigger autonomic reflexes. For example, exogenous
adenosine directly excited the carotid chemoreceptors of the anesthetized cat [69];
in the dog with intact renal nerves, intrarenal adenosine produced hypertension by
activating the sympathetic nervous system [61]; in rabbits under light, stable halo-
thane anesthesia, continuous infusion of adenosine produced immediate hypoten-
sion without tachycardia [47]. Furthermore, it has been shown that the net chrono-
tropic action of exogenous adenosine (given as a bolus injection into a peripheral
vein) depends on its direct depressant action on the sinus node as well as baroreflex-
mediated, vasodilation-triggered sinus tachycardia [55].
Thus, it can be concluded that the adenosine-vagal interaction possibly involves
several physiological factors including direct actions of adenosine on target cells,
direct action of adenosine on vagal nerve terminals and nondirect chemo- and baro-
receptor-mediated reflexes (see below). More studies are required before the
physiological importance of endogenous adenosine-vagal interaction is fully deter-
mined and exogenous adenosine-vagal interaction is quantitated under different
pathophysiological conditions.

ATP-Vagal Interaction

Early studies in canine and feline models showed that atropine attenuates the elec-
trophysiological actions of ATP, thus suggesting vagal involvement in these actions
of the nucleotide [20, 38, 110]. A more recent systematic study has confirmed these
observations and established the ability of exogenous ATP to trigger a strong vagal
Adenosine, ATP and Cardiac Neural Control 65

reflex [47]. This reflex mediates to a large extent the electrophysiological effects of
ATP and is responsible for the differential potency of ATP and adenosine when the
two compounds are administered rapidly at equimolar doses into the canine right
atrium [7S]. Indeed, under conditions of either muscarinic cholinergic blockade or
bilateral cervical vagotomy, ATP and adenosine are equipotent with respect to their
negative chronotropic and dromotropic actions [78]. Similar vagal involvement in the
electrophysiological action of ATP has also been observed in humans [57, 100, 110].
Neither the trigger mechanism, nor the site at which ATP initiates the afferent vagal
volley, are known. The sites at which ATP triggers this vagal reflex are not present
uniformly in the cardiovascular system. Thus, when administered into the canine
sinus nodal artery (in the presence of intact autonomic nervous system), ATP and
adenosine are equipotent [60], or the latter is more potent than the former [S3]. The
major vagal component in the electro physiological actions of A TP explains the
potentiation of its actions by interventions which are known to enhance the activity
of acetylcholine. For example, physostigmine, an inhibitor of acetylcholinesterase,
markedly potentiated the negative chronotropic and dromotropic actions of ATP in
the canine sinus and AV nodes, respectively [SO]. In addition, elevation of plasma
Ca2 + levels, which is known to enhance release of acetylcholine from nerve terminals
[102, 107], also significantly potentiated the negative chronotropic action of ATP
[77]. In view of these data, atrial fibrillation induced by ATP [9] could be considered
a manifestation of vagal-mediated atrial arrhythmias [23].
In contrast to the actions of ATP on the sinus and AV nodes, its negative chrono-
tropic action on ventricular escape rhythm is not mediated by the vagus [Sl]. This is
consistent with an earlier study, which indicated that vagal stimulation has only a
small depressant effect on ventricular escape rhythms [l05].
Further studies are required to determine the threshold concentration of ATP
which triggers vagal afferent traffic. This will be instrumental in evaluating the
physiological significance of this phenomenon.

Adenosine Interactions with Catecholamine and Histamine

Almost a decade ago Schrader et al. [95] reported that infusion of adenosine into the
coronary artery of isolated guinea pig hearts produced a dose-dependent inhibition
of the positive inotropic effect caused by bolus injections of isoproterenol. In addi-
tion, adenosine inhibited some of the biochemical effects of isoproterenol including
the rise in myocardial levels of cyclic adenosine 3,'5'-monophosphate (cAMP) and
adenyl ate cyclase activity [95]. These observations were later confirmed by Dobson
and his colleagues in a series of studies (for review, see reference 30). These and
more recent studies [2S, 29, 31, 43, 63] led Dobson and his colleagues to hypothesize
that adenosine "may naturally serve as a negative feedback modulator that normally
regulates the magnitude of the myocardial response to catecholamine stimulator"
[27].
Adenosine also attenuates cardiac responses to histamine [4, 52]. Infusion of
adenosine into the coronary arteries of isolated guinea pig hearts produced a dose-
dependent inhibition of histamine-induced contractile force development as well as
66 A.Pelleg

a rise in tissue levels of cAMP [4]. In a later study Hattori and Levi demonstrated
that adenosine attenuates histamine actions which are mediated by H 2- and not H 1-
receptors in perfused guinea pig hearts [52]. This selective antagonism which un-
masks the Hrmediated effects of histamine has also been shown in pectinate muscles
isolated from human right atrium [48]. Further studies will determine the clinical im-
plication of adenosine-histamine interaction, in particular with respect to patho-
physiological conditions associated with increased histamine stimulation and adeno-
sine release. Adenosine-catecholamine antagonism also affects the electrophysio-
logical actions of adenosine. Adenosine antagonized the positive chronotropic effect
of norepinephrine in spontaneously beating isolated rat atria [94]. Analogous to its
interaction with histamine, the attenuation of norepinephrine effects by adenosine
was specific to ,B-adrenergic-receptor-mediated actions resulting in unmasking of a-
mediated responses [41, 52]. In potassium-depolarized guinea pig atrial muscle fibers,
adenosine abolished isoproterenol-induced actin potentials [92, 96]. Adenosine also
antagonized the electrophysiological action of isoproterenol in embryonic chick right
ventricular strips [7], canine Purkinje fibers [87], and isolated bovine myocytes [5].
Based on the above-mentioned findings, it is tempting to speculate that the anti-
adrenergic effect of adenosine is mediated by its action on the adenylate cyclase sys-
tem resulting in decreased myocardial level of cAMP. Indeed, such a decrease was
recently shown by Belardinelli et al. [7].
This hypothesis is supported by recent studies on adenosine-forskolin interaction.
Forskolin, a diterpene extracted from the roots of coleus forskohlii [13], activates the
adenylate cyclase system without interacting with either an extracellular receptor or
an intracellular guanyl nucleotide-binding protein [24]. Thus, the action of forskolin
mimics those of ,B-adrenoceptor agonists with regard to the activation of the adenylate
cyclase system [67, 90]. Adenosine has been shown to inhibit catecholamine-like
effects of forskolin in extracardiac tissues [46, 117]. In addition, in a recent study
West et al. [60] have found that adenosine antagonized the forskolin-induced in-
crease in left ventricular pressure and dp/dt in isolated perfused guinea pig hearts.
Adenosine also antagonized the effects of forskolin on the action potential and sarco-
mere shortening in guinea pig isolated ventricular myocytes [115]. Furthermore, the
latter actions of forskolin were associated with an increase in the level of cAMP, which
was reduced by the potent adenosine analogue N6-R-phenyl-isopropryladenosine
[115].
Further support is derived from the effects of pertussis toxin (i.e., islet-activating
protein, lAP) on the actions of adenosine. This toxin is able to modulate receptor-
mediated generation of cAMP by inhibiting Ni, the guanine nucleotide-binding pro-
tein that couples membrane receptors to the catalytic unit of adenylate cyclase [108].
Adenosine attenuation of isoproterenol-induced positive inotropic action and in-
crease in cAMP in isolated rat left atria was prevented by pretreatment with lAP
[39]. Similar results were obtained with regard to rate of contraction of isolated rat
right atria [40].
In contrast to the above-mentioned data, other studies found evidence against
cAMP involvement in adenosine-catecholamine antagonism. For example, Schutz
and Tuisl [97] reported that two potent analogues of adenosine did not influence
basal or isoproterenol-stimulated cyclase activity in guinea pig ventricular membrane
preparations. Furthermore, it was found in a different study that adenosine com-
Adenosine, ATP and Cardiac Neural Control 67

pletely abolished and attenuated the positive inotropic effect of isoprenaline in iso-
lated, electrically driven guinea pig atrial and ventricular muscle preparations, re-
spectively [15]; these actions of adenosine as well as its shortening of the atrial action
potential were not associated with a decrease in cAMP content [15]. In addition, the
direct negative inotropic effect of adenosine and its shortening of action potential
duration in isolated electrically driven left auricles from the guinea pig were also not
accompanied by a change of cAMP content [17].
These inconsistent results probably reflect methodological differences since
adenosine can both stimulate and inhibit the catalytic activity of adenylate cyclase
and, furthermore, adenosine in both intracellular and extracellular compartments
may be involved in this modulation of adenylate cyclase activity [111]. It seems that
the direct electrophysiological actions of adenosine are not associated with change in
cAMP levels; however, the indirect anti adrenergic ations of adenosine involve mod-
ulation of the adenylate cyclase and changes in cAMP levels [40]. Thus, there are still
important unanswered questions with regard to adenosine-catecholamine antagonism,
and additional studies are required to fully elucidate the mechanism behind this
phenomenon.
All of the above-mentioned studies were carried out in vitro, and adenosine-cate-
cholamine antagonism in vivo has not been shown conclusively. Two recent studies
in anesthetized [98] and conscious dogs [86] did not find evidence for adenosine-
catecholamine antagonism in vivo. In the first study adenosine failed to influence the
positive inotropic and chronotropic effects of intracoronary infusions or bolus injec-
tions of isoproterenol [98]. In the second study, neither adenosine receptor competi-
tive blockade with 8-phenyltheophylline nor the intracellular incactivation of
adenosine by homocysteine had any influence on exercise-induced increases in myo-
cardial contractility and heart rate [86]. In contrast, Lerman and his colleagues [64,
65] have recently reported on exercise- and isoproterenol-induced ventricular tachy-
cardia which was terminated by adenosine. In addition, propranolol was also effec-
tive in preventing and terminating these tachycardias [64, 65]. These data were inter-
preted as evidence for adenosine suppression of catecholamine-dependent, cAMP-
mediated triggered ventricular activity in these patients. However, additional studies
are necessary to establish adenosine-catecholamine antagonism in vivo and its physio-
logical role.

Autonomic Reflexes

Due to the rapid breakdown of ATP to adenosine and the transport of the latter into
cells, the half-life of these compounds is very short (i.e., < 10 s). Thus, bolus injec-
tions of relatively small doses of adenosine and ATP are characterized mainly by
transient direct actions in the mammalian heart. However, large doses or continuous
infusion of these compounds resulting in elevated plasma adenosine levels might
have systemic effects which could modulate the direct electrophysiological actions of
adenosine. For example, the net effect of adenosine on sinus nodal activity is depen-
dent on its direct negative chronotropic action as well as baroreceptor-mediated
autonomic input to the sinus node due to adenosine-induced decrease in systemic
68 A.Pelleg

arterial blood pressure [55]. Consequently, variable results were obtained following
infusions of adenosine and ATP in animal models as well as human subjects. For
example, in rats anesthetized with halothane, infusion of ATP, which caused a pro-
nounced fall in blood pressure, was not associated with a significant change in heart
rate [56]. However, in conscious, unrestrained, normotensive rats adenosine caused
hypotension and either increased or unchanged heart rate (depending on site of ad-
ministration: intra-arterial vs intravenous, respectively) [75]. In rabbits anesthetized
with halothane, infusion of adenosine and ATP resulted in similar marked hypoten-
sion and a mild drop in heart rate [62]. In dogs anesthetized initially with thiopental
and nitrous oxide followed by phenoperidine and droperidol, adenosine and ATP in-
duced pronounced hypotension associated with marked decrease in heart rate [104].
In contrast, adenosine infusion in conscious human subjects resulted in increased
heart rate and no change in blood pressure [22]. This effect of adenosine was poten-
tiated by dipyridamole [22]. In a more recent study, also in conscious normal sub-
jects, continuous infusion of adenosine markedly increased heart rate and induced a
small increase and decrease in systolic and diastolic blood pressures, respectively
[14]. These changes were associated with increased plasma levels of norepinephrine
and epinephrine [14]. A delayed sinus tachycardia (which followed transient sinus
bradycardia) was also observed in healthy adults following intravenous bolus doses
of adenosine [109]. In this report, the tachycardia was explained by a chemoreceptor
reflex (see below) and not by hypotension. However, since blood pressure was not
monitored in this study, it is difficult to accept or reject this hypothesis.
Adenosine can also act directly on the chemoreceptors [89]; an action similar to
that was noted with ATP [98]. Adenosine increased the spontaneous activity of the
chemoreceptors by acting at extracellular receptor sites located probably on the
nerve endings [89]. A similar effect on renal chemoreceptor nerve terminals is re-
sponsible for the enhanced afferent renal nerve activity resulting in increased sym-
pathetic tone following acute intrarenal administration of adenosine [61]. However,
different results were obtained by chronic elevation of intrarenal adenosine [51], and
therefore the renal actions of adenosine are not fully understood.
It should be mentioned here that adenosine can exert both excitatory and inhibi-
tory effects on respiration. The ability of adenosine to induce respiration has been
shown in experimental models, as well as in man [19, 26, 109]. Recent data obtained
in a rat model have indicated that carotid body chemoreceptors mediate this action
of adenosine [71]. Preliminary studies in cats showed that adenosine can also stimu-
late respiration by affecting pulmonary vagal nerve terminals [21, 93]. In contrast to
its peripheral action, the central respiratory effects of adenosine are inhibitory [25,
37,114].
Whether these actions of adenosine can affect, directly or indirectly, its cardiac
electrophysiology and electropharmacology is not known.

Inhibition of Neurotransmission

Considerable data have accumulated suggesting that adenosine and ATP act in the
central and peripheral nervous system as neurotransmitters released, at least in part,
Adenosine, ATP and Cardiac Neural Control 69

from purinergic nerves [3, 18, 36, 106]. In addition, adenosine and its related nucleo-
tides inhibit the release of neurotransmitters from nerve endings in various organs
and tissues [35, 45,84,88] including the mammalian heart [54, 68]. This mechanism
could be involved in the adenosine-catecholamine antagonism discussed above. In-
deed, a recent study found that the effects of adenosine on systemic arterial blood
pressure and heart rate were significantly different in normotensive rats compared
with spontaneously hypertensive rats [75]. It was hypothesized that these differences
were due, at least in part, to adenosine's inhibition of neurotransmitter release from
sympathetic nerve terminals and the elevated sympathetic tone in the spontaneously
hypertensive rats [75]. While the hypothesis that the electrophysiological actions of
adenosine are significantly modulated by its inhibition of neurotransmission is very
attractive, it has not been studied systematically and therefore the physiological im-
portance of this modulation is still to be determined.

Summary and Conclusions

Studies reviewed in this chapter indicate that adenosine and ATP, and the auto-
nomic nervous system, can interact both at the pre- and postsynaptic level. This
interaction is a complex one and can affect the autonomic neural control of the heart
as well as certain cardiac actions of adenosine and ATP. Figure 1 schematically shows
several mechanisms of interaction between adenosine (ADN) and ATP and the auto-
nomic nervous system. The hexagon in the center upper part represents myocardial
cell and its purine metabolism (Purine Metab.). Adenosine and ATP are transported
across the cell membrane (into and out of the cell) and probably can also leak (L) out
of the cell under pathophysiological conditions [44, 58]. Evidence for a specific trans-
port system for adenosine [76], which is blocked by certain agents (e.g., dipyridamole,

Coronaries Coronaries
'------.-----'..,. Blood Flowt

Vasodi 1 at ion Pur i ne

Metab.

~* ~[:
-
~~~.
<-~ . -!f-
RVN
fiRTP ~
Parasymp. + Rv PG's

~~symp.
~ ~tr---
- ~----
+

Fig.1. Schematic representation of adenosine and ATP interactions with autonomic control of the
heart (see text for details)
70 A.Pelleg

nitrobenzylthioinosine, and calcium channel blockers), has been found in several dif-
ferent cell types. However, the mechanism of this system is not fully elucidated. Less
is known about the transport system of ATP.
The potent vasodilatory action of adenosine and ATP on the coronary arteries
(Coronaries) is also shown. This action can influence cellular metabolism by affect-
ing nutrients and O 2 supply via blood flow (i.e., the adenosine hypothesis [12]). The
rapid breakdown of ATP to adenosine by ectoenzymes (Es) is also depicted in the
upper part of the figure.
The sympathetic (Symp.) and parasympathetic (Parasymp.) limbs of the auto-
nomic nervous system are represented, respectively, by two nerve terminals con-
taining neurotransmitter vesicles in the center lower part of the figure. An arrow
labeled with a minus sign, leading from each terminal to the other, symbolizes the
sympathetic-parasympathetic interaction at the presynaptic level. Three major com-
ponents of the specialized tissues of the heart are shown to the left as three rectangles.
These are: the sinus node (SN), the AV node (AVN), and infranodal pacemakers
(INP). Arrows leading from adenosine to these components represent the negative
chronotropic and dromotropic actions of adenosine. The asterisk above each rectangle
symbolizes the potential postsynaptic interactions of adenosine with the sympathetic
and/or parasympathetic neural input to these structures (arrows leading from the two
nerve terminals; a plus sign indicates positive and a minus sign negative chronotropic
or dromotropic action). Arrows leading from adenosine to the two nerve terminals
indicate presynaptic action, i.e., attenuation of neurotransmitter release.
Finally, vagal reflex (Rv) triggered by ATP directly, or by ATP and adenosine via
triggered synthesis and release of prostaglandin (PG's) [74, 118], is depicted on the
right lower part of the figure.
It can be concluded that the nucleoside adenosine and its nucleotide ATP can
interact with the autonomic nervous system in a complex manner. These interactions
can modulate the autonomic neural control of the heart and also the cardiac electro-
physiological actions of adenosine and ATP. Further studies are required to fully
elucidate the mechanisms of these interactions and determine their physiological
importance.
Acknowledgments. The support, encouragement, and helpful comments by Drs. Leonard S. Dreifus
and Eric L. Michelson, and the assistance of Rose Marie Wells and Marilyn Lucianetti in preparing
this manuscript are gratefully acknowledged. Original work was supported in part by grants from the
American Heart Association, Pennsylvania Affiliate.

References

1. Akasu T, Hirai K, Koketsu K (1981) Increase in acetylcholine-receptor sensitivity by adenosine

triphosphate: A novel action of A TP on ACh-sensitivity. Br J Pharmacol 74: 505-507
2. Akasu T, Koketsu K (1985) Effect of adenosine triphosphate on the sensitivity of the nicotinic
acetylcholine-receptor in the bullfrog sympathetic ganglion cell. Br J Pharmacol 84: 525-531
3. Akasu T, Shinnick-Gallagher P, Gallagher JP (1984) Adenosine mediates slow hyperpolarizing
synaptic potential in autonomic neurons. Nature 311: 62-65
4. Baumann G, Schrader J, Gerlach E (1981) Inhibitory action of adenosine on histamine- and
dopamine-stimulated cardiac contractility and adenylate cyclase in guinea pigs. Circ Res 48:
259-266
Adenosine, ATP and Cardiac Neural Control 71

5. Belardinelli L, Isenberg G (1983) Actions of adenosine and isoproterenol on isolated mam-

malian ventricular myocytes. Circ Res 53: 287 -297
6. Belardinelli L, Mattos E, Berne RM (1981) Evidence for adenosine mediation of atrioventricu-
lar block in the ischemic canine myocardium. J Clin Invest 68: 195-205
7. Belardinelli L, Vogel S, Linden J, Berne RM (1982) Antiadrenergic action of adenosine on
ventricular myocardium in embryonic chick hearts. J Mol Cell Cardiol 14: 291-294
8. Belardinelli L, West A, Crampton R, Berne RM (1983) Chronotropic and dromotropic effects
of adenosine. In: Berne RM, Rail TW, Rubio R (eds) Regulatory function of adenosine.
Nijhoff, Boston, pp 377-396
9. Belhassen B, Pelleg A, Shoshani D, Laniado S (1984) Atrial fibrillation induced by adenosine
triphosphate. Am J Cardiol54: 1405-1406
10. Belhassen B, Pelleg A (1985) Adenosine triphosphate and adenosine: Perspectives in the acute
management of paroxysmal supraventricular tachycardia. Clin Cardiol 8: 460-464
11. Belhassen B, Pelleg A (1986) Electrophysiologic effects of adenosine triphosphate and adenosine
on the mammalian heart: clinical and experimental aspects. J Am Coli CardioI2:414-424
12. Berne RM (1963) Cardiac nucleotides in hypoxia: Possible role in regulation of coronary blood
flow. Am J PhysioI204:317-322
13. Bhat SV, Bajwa BS, Domauer H, DeSouza NJ (1977) Structures and stereochemistry of new
lambdane diterpenoids from coleus forskohlii brig. Tetrahedron Lett 19: 1669-1672
14. Biaggioni I, Qurot J, Kincaid D, Hollister AS, Robertson D (1985) Humoral and hemodynamic
effects of adenosine infusion in man (Abstr). Clin Res 33 : 280 A
15. Bohm M, Bruckner R, Hackbarth I, Haubitz B, Linhart R, Meyer W, Schmidt B, Schmitz W,
Scholz H (1984) Adenosine inhibition of catecholamine-induced increase in force of contraction
in guinea-pig atrial and ventricular heart preparations. Evidence against a cyclic AMP- and
cyclic GMP-dependent effect. J Pharmacol Exp Ther 230: 483-492
16. Brown GL, Eccles JC (1934) The action of a simple vagal volley on the rhythm of the heart
beat. J Physiol 82: 211
17. Bruckner R, Genner A, Wilfried M, Nobis TM, Schmitz W, Scholz H (1985) Cardiac effects of
adenosine and adenosine analogs in guinea-pig atrial and ventricular preparations: Evidence
against a role of cyclic AMP and cyclic GMP. J Pharmacol Exp Ther 234:766-774
18. Bumstock G, Campbell G, Satchell D, Smythe A (1970) Evidence that adenosine triphosphate
or a related nucleotide is the transmitter substance released by nonadrenergic inhibitory nerves
in the gut. Br J Pharmacol 40 : 668-688
19. Buss DC, Routledge PA, Watt AH (1986) Intravenous adenosine stimulates respiration in
conscious adult rabbits. Br J Pharmacol 87: 182P
20. Cardenas M, Aceves J, Alarcon G (1964) Efecto del acido adenosintrifosforico sobre las
propiedades fisiologicas del corazon. Arch Inst Cardiol Mex 34: 485-494
21. Chemiack NS, Runold M, Rrabhakar NR, Mitra J (1987) Effect of adenosine on vagal sensory
pulmonary afferents (Abstr). Fed Proc 46: 825
22. Conradson TB, Dixon CMS, Clarke B, Barnes PJ (1986) Dipyridamole potentiates the cardio-
vascular effects of adenosine in man (Abstr). Pfliigers Arch 407: S55
23. Coumel P, Attuel P, Lavallee JP, Flammang D, Leclerq JF, Slama R (1978) Syndrome d'arythmie
auriculaire d'origine vagale. Arch Mal Coeur 71: 645-656
24. Daly JW (1984) Forskolin adenylate cyclase and cell physiology: An overview. Adv Cycl
Nucleotide Protein Phosphorylation Res 17: 81-89
25. Darnall RA Jr (1985) Aminophylline reduces hypoxic ventilatory depression: possible role of
adenosine. Pediatr Res 19: 706-710
26. Dixon CMS, Fuller RW, Hughes JMB, Maxwell DL, Nolop KB (1986) Hypoxic and hyper-
capnic ventilatory responses during adenosine infusion in man. J Physiol (Lond) 376: 51P
27. Dobson JG Jr (1983) Adenosine reduces catecholamine contractile response in oxygenated and
hypoxic atria. Am J Physiol245: H468-H474
28. Dobson JG Jr (1983) Interaction between adenosine and inotropic interventions in guinea pig
atria. Am J Physiol245 :H475-H480
29. Dobson JG Jr (1983) Mechanism of adenosine inhibition of catecholamine-induced responses
in the heart. Circ Res 52: 151-160
30. Dobson JG Jr, Fenton RA (1983) Antiadrenergic effects of adenosine in the hearts. In: Berne
RM, Rail TW, Rubio R (eds) Regulatory function of adenosine. Nijhoff, Boston, pp 363-376
72 A.Pelleg

31. Dobson JG Jr, Schrader J (1984) Role of extracellular and intracellular adenosine in the attenu-
ation of catecholamine evoked responses in guinea pig heart. J Mol Cell Cardiol 16: 813-822
32. Dong E Jr, Reitz BA (1970) Effects of timing of vagal stimulation on heart rate in the dog. Circ
Res 27: 635-646
33. Drury AN, Lutwak-Mann C, Solandt OM (1938) The inactivation of adenosine by blood with
special reference to cat's blood. Q J Exp Physiol27: 215-236
34. Drury AN, Szent-Gyorgyi A (1929) The physiological activity of adenine compounds with
especial reference to their action upon the mammalian heart. J Physiol (Lond) 68: 213-237
35. Dunwiddie TV (1985) The physiological role of adenosine in the central nervous system. Int
Rev NeurobioI27:63-132
36. Editorial (1977) Purinergic nerves. Lancet 2: 1331-1332
37. Eldridge FL, Millhom DE, Kiley JP (1985) Antagonism by theophylline of respiratory inhibi-
tion induced by adenosine. J Appl Physiol59: 1428-1433
38. Emmelin N, Feldberg W (1948) Systemic effects of adenosine triphosphate. Br J Pharmacol
Chemother 3: 273-284
39. Endoh M, Maruyama M, Taira N (1983) Modification by islet-activating protein of direct and
indirect inhibitory actions of adenosine on rat atrial contraction in relation to cyclic nucleotide
metabolism. J Cardiovasc Pharmacol 5 : 131-142
40. Endoh M, Maruyama M, Taira N (1983) Adenosine-induced changes in rate of beating and
cyclic nucleotide levels in rat atria: modification by islet-activating protein. In: Daly JW,
Kuroda Y, Phillis JW, Shimizu H, Vi M (eds) Physiology and pharmacology of adenosine
derivatives. Raven, New York, pp 127-141
41. Endoh M, Yamashita S (1980) Adenosine antagonizes the positive inotropic action mediated
via f3-, but not a-adrenoceptors in the rabbit papillary muscle. Eur J Pharmacol 65 : 445-448
42. Ewald DA (1976) Potentiation of postjunctional cholinergic sensitivity of rat diaphragm muscle
by high energy phosphate adenine nucleotides. J Membr Bioi 29: 47 -65
43. Fenton RA, Dobson JG Jr (1984) Adenosine and calcium alter adrenergic-induced intact heart
protein phosphorylation. Am J Physiol 246: H559-H565
44. Forrester T, Williams CA (1977) Release of adenosine triphosphate from isolated adult heart
cells in response to hypoxia. J Physiol 268: 371-390
45. Fredholm BB, Hedqvist P (1980) Modulation of neurotransmission by purine nucleotides and
nucleosides. Biochem Pharmacol 29: 1635-1643
46. Fredholm BB, Jonzon B, Lindstrom K (1986) Effect of adenosine receptor agonists and other
compounds on cyclic AMP accumulation in forskolin-treated hippocampal slices. Arch Phar-
macol 332 : 173-178
47. Fukunaga AF, Flacke VE, Bloor BC (1982) Hypotensive effects of adenosine and adenosine
triphosphate compared with sodium nitroprusside. Anesth Analg 61: 273-278
48. Genovese A, Levi R (1987) Adenosine modulates histamine-induced stimulation of human
atrial myocardium: protection by Hrreceptor blockade. In: Pelleg A, Michelson EL, Dreifus
LS (eds) Cardiac electrophysiology and pharmacology of adenosine and A TP: basic and clinical
aspects. Liss, New York, pp 345-360
49. Gustafsson L (1981) Influence of adenosine on responses to vagal nerve stimulation in the anes-
thetized rabbit. Acta Physiol Scand 111: 263-268
50. Gustafsson L, Hedqvist P, Fredholm BB, Lundgren G (1978) Inhibition of acetylcholine re-
lease in guinea pig ileum by adenosine. Acta Physiol Scand 104:469-478
51. Hall JE, Granger JP (1986) Renal hemodynamics and arterial pressure during chronic intra-
renal adenosine infusion in conscious dogs. Am J Physiol250: F32-F39
52. Hattori Y, Levi R (1984) Adenosine selectively attenuates H 2- and beta-mediated cardiac re-
sponses to histamine and norepinephrine: an unmasking of HI and alpha-mediated selective re-
sponses. J Pharmacol Exp Ther 231 :215-223
53. Hayashi E, Mori M, Yamada S, Kunitomo M (1978) Effects of purine compounds on cholin-
ergic nerves. Specificity of adenosine and related compounds on acetylcholine release in electri-
cally stimulated guinea pig ileum. Eur J Pharmacol48: 297-307
54. Hedqvist P, Fredholm BB (1979) Inhibitory effect of adenosine on adrenergic neuroeffector
transmission in the rabbit heart. Acta Physiol Scand 105: 120-122
55. Hintze TH, Belloni FL, Harrison JE, Shapiro GE (1985) Apparent reduction in baroreflex sen-
sitivity to adenosine in conscious dogs. Am J Physiol249: H554-H559
Adenosine, ATP and Cardiac Neural Control 73

56. Hoffman WE, Satinover I, Miletich DJ, Albrecht RF, Gans BJ (1982) Cardiovascular changes
during sodium nitroprusside or adenosine triphosphate infusion in the rat. Anesth Analg 61:
99-103
57. Hugues FC, Jan Y, Bars V, Baubion N (1980) Etude de l'action chronotrope de l'ATP chez
l'homme. Coeur Med Inter 19: 227-234
58. Imai S, Riley AL, Berne RM (1964) Effect of ischemia on adenine nucleotides in cardiac and
skeletal muscle. Circ Res 15: 443-450
59. Jalife J, Moe GK (1979) Phasic effects of vagal stimulation on pacemaker activity of the isolated
sinus node of the young cat. Circ Res 45: 595-607
60. James TN (1965) The chronotropic action of ATP and related compounds studied by direct per-
fusion of the sinus node. J Pharmacol Exp Ther 149: 233-247
61. Katholi RE, Whitlow PL, Hageman GR, Woods WT (1984) Intrarenal adenosine produces
hypertension by activating the sympathetic nervous system via the renal nerves in the dog.
J Hypertension 2: 349-359
62. Lagerkranser M, Irestedt L, Sollevi A, Andreen M (1984) Central and splanchnic hemo-
dynamics in the dog during controlled hypotension with adenosine. Anesthesiology 60: 547-552
63. LaMonica DA, Frohloff N, Dobson JG (1985) Adenosine inhibition of catecholamine-stimu-
lated cardiac membrane adenylate cyclase. Am J Physiol248:H737-H744
64. Lerman BB, Belardinelli L (1986) Effects of adenosine on ventricular tachycardia. In: Pelleg
A, Michelson EL, Dreifus LS (eds) Cardiac electrophysiology and pharmacology of adenosine
and ATP: basic and clinical aspects. Liss, New York, pp 301-314
65. Lerman BB, Belardinelli L, West GA, Berne RM, DiMarco JP (1986) Adenosine sensitive
ventricular tachycardia: evidence suggesting cyclic AMP mediated triggered activity. Circula-
tion 74: 270-280
66. Levy MN, Martin PJ, Iano T, Zieske H (1970) Effects of single vagal stimulus on heart rate and
atrioventricular conduction. Am J Physiol 281 : 1256
67. Lindner E, Dohadwalla AN, Bhattacharya BK (1978) Positive inotropic and blood pressure
lowering activity of a diterpene derivative isolated from coleus forskohlii: Forskolin. Arznei-
mitteiforschung 28: 284-289
68. Lokhandwala MF (1979) Inhibition of cardiac sympathetic neuro-transmission by adenosine.
Eur J Pharmacol 60: 353-357
69. McQueen DS, Ribeiro JA (1981) Effect of adenosine on carotid chemoreceptor activity in the
cat. Br J Pharmacol 74: 129-136
70. Monteiro EC, Ribeiro JA (1986) Influence of adenosine on heart responses to vagal nerve
stimulation. Br J Pharmacol 87: 184P
71. Monteiro EC, Ribeiro JA (1987) Ventilatory effects of adenosine mediated by carotid body
chemoreceptors in the rat. Arch Pharmacol335: 143-148
72. Munoz A, Leenhardt A, Sassine A, Puech P (1985) Atropine antagonizes the effect of adeno-
sine on atrioventricular conduction in closed chest dogs. Circulation 72: 111-241
73. Murthy VS, Zagar ME, Vollmer RR, Schmidt DH (1982) Pentobarbital-induced changes in
vagal tone and reflex vagal activity in rabbits. Eur J Pharmacol84:41-50
74. Needleman P, Minkes MS, Douglas JR (1974) Stimulation of prostaglandin biosynthesis by
adenine nucleotides. Profile of prostaglandin release by perfused organs. Circ Res 34: 455-460
75. Ohnishi A, Biaggioni I, Deray G, Branch RA, Jackson EK (1986) Hemodynamic effects of
adenosine in conscious spontaneously hypertensive and normotensive rats. Hypertension 8:
391-398
76. Paterson ARP, Harley ER, Cass CE (1985) Measurement and inhibition of membrane trans-
port of adenosine. In: Paton DM (ed) Methods in adenosine research. Plenum, New York, pp
165-180
77. Pelleg A, Michelson EL (1987) Role of the vagus in modulation by Ca2 + of the depressant
action of adenosine and adenosine 5 ' -triphosphate on the canine sinus node in vivo. J Auton
Pharmacol7:127-134
78. Pelleg A, Belhassen B, Ilia R, Laniado S (1985) Comparative electrophysiologic effects of ATP
and adenosine in the canine heart: influence of atropine, propranolol, vagotomy, dipyridamole
and aminophylline. Am J Cardiol 55: 571-576
79. Pelleg A, Mazgalev T, Price R, Dreifus LS, Michelson EL (1985) Adenosine modulation of
vagal effects in the rabbit sino-atrial node. Fed Proc 44: 468
74 A.Pelleg

80. Pelleg A, Mitamura H, Michelson EL (1985) Evidence for vagal involvement in the electro-
physiologic actions of exogenous adenosine and adenosine triphosphate in the canine heart.
J Auton Pharmacol5: 207-212
81. Pelleg A, Mitamura H, Mitsuoka T, Michelson EL, Dreifus LS (1986) Effects of adenosine and
ATP on ventericular escape rhythm in the canine heart. J Am Coil Cardiol8: 1145-1151
82. Pelleg A, Mitsuoka T, Mazgalev T, Michelson EL (1987) Interacting negative chronotropic
effects of adenosine and the vagus nerve on the canine sinus node. Cardiovasc Res (in press)
83. Pelleg A, Mitsuoka T, Michelson EL, Menduke H (1988) Adenosine mediates the negative
chronotropic action of adenosine 5 ' -triphosphate in the canine sinus node. J Pharmacol Exp
Ther 242: 791-795
84. Phillis JW, Barraco RA (1985) Adenosine, adenylate cyclase, and transmitter release. Adv
Cyclic Nucleotide Protein Phosphorylation Res 19: 243-257
85. Puech P, Sassine A, Munoz A, Masse C, Effelmaier F, Leenhardt A, Yoshimura H (1986)
Electrophysiologic effects of purines: clinical applications. In: Zipes DP, Jalife J (eds) Cardiac
electrophysiology and arrhythmias. Grune and Stratton, Orlando, pp 443-450
86. Raberger G, Fischer G, Krumpl G, Schneider W, Stroibrig H (1986) Further evidence against
an adenosine-catecholamine antagonism in vivo: Investigations with treadmill-exercise in dogs.
Pfliigers Arch 407: S16
87. Rardon DP, Bailey JC (1983) Adenosine attenuation of the electrophysiologic effects of isopro-
terenol on cardiac Purkinje fibers. J Pharmacol Exp Ther 228: 792-798
88. Ribeiro JA (1979) Purinergic modulation of transmitter release. J Theor Bioi 80: 259-270
89. Ribeiro JA, McQueen DS (1983) On the neuromuscular depression and chemoreceptor activa-
tion caused by adenosine. In: Daly JW, Kuroda Y, Phillis JW, Shimizu H, Vi M (eds) Physiol-
ogy and pharmacology of adenosine. Raven, New York, pp 179-188
90. Rodger IW, Shahid M (1984) Forskolin, cyclic nucleotides and positive inotropism in isolated
papillary muscles of the rabbit. Br J Pharmacol 81 : 151-159
91. Rossi A, Piery Y, Kreher P (1967) Modification par les nucleotides adenyliques de l'effect
chronotrope de la stimulation du vague sur Ie coeur du rat. CR Soc Bioi (Paris) 161: 860-861
92. Rubio R, Belardinelli L, Thompson CI, Berne RM (1979) Cardiac adenosine: electrophysio-
logic effects, possible significance in cell function, and mechanism controlling its release. In:
Baer HP, Drummond GI (eds) Physiological and regulatory functions of adenosine and adenine
nucleotides. Raven, New York, pp 167-182
93. Runold M, Prabhakar NR, Mitra J, Cherniack NS (1987) Adenosine stimulates respiration by
acting on vagal receptors. Fed Proc 46: 825
94. Samet MK, Rutledge CO (1984) Antagonism of the positive chronotropic effect of norepi-
nephrine by purine nucleosides in rat atria. J Pharmacol Exp Ther 232: 106-110
95. Schrader J, Baumann G, Gerlach E (1977) Adenosine as an inhibitor of myocardial effects of
catecholamines. Pfliigers Arch 372: 29-35
96. Schrader J, Rubio R, Berne RM (1975) Inhibition of slow action potentials of guinea pig atrial
muscle by adenosine: a possible effect on Ca2+ influx. J Mol Cell CardioI7:427-433
97. Schutz W, Tuisl E (1981) Evidence against adenylate cyclase-coupled adenosine receptors in
the guinea pig heart. Eur J PharmacoI76:285-288
98. Seitelberger R, Schutz W, Schlappack 0, Raberger G (1984) Evidence against the adenosine-
catecholamine antagonism under in vivo conditions. Arch Pharmacol 4325: 234-239
99. Shah PK, Nalos P, Peter T (1987) Atropine resistant post infarction complete AV block: Pos-
sible role of adenosine and improvement with aminophylline. Am Heart J 113: 194-195
100. Sharma AD, Milstein S, Rattes M, Klein GJ (1987) Disparate effects of adenosine triphosphate
on sinus node automaticity and A V nodal conduction in man. J Am Coil Cardiol 9246A
101. Shinozuka K, Cheng JT, Hayashi E (1985) Effects of adenosine on contractile response of
circular muscle in electrically stimulated guinea pig ileum. Jpn J Pharmacol 38: 361-365
102. Silinksy EM (1985) The biological pharmacology of calcium-dependent acetylcholine secretion.
Pharmacol Rev 37:81-132
103. Sollevi A (1986) Cardiovascular effects of adenosine in man; possible clinical implications. Prog
Neurobiol27: 319-349
104. Sollevi A, Lagerkranser M, Andreen M, Irestedt L (1984) Relationship between arterial and
venous adenosine levels and vasodilation during ATP- and adenosine-infusion in dogs. Acta
Physiol Scand 120: 171-176
Adenosine, ATP and Cardiac Neural Control 75

105. Spear JF, Moore EN (1973) Influence of brief vagal and stellate nerve stimulation on pacemaker
activity and conduction within the atrioventricular conduction system of the dog. Circ Res 32:
27-41
106. Stone TW (1981) Physiological roles for adenosine and adenosine 5'-triphosphate in the
nervous system. Neuroscience 6: 523-555
107. Timour-Chah Q, Bertrix L, Lang J, Bouzouita K, Faucon G (1983) Enhancement by Ca2 + ions
of cholinergic effects on the canine heart in situ. Arch Pharmacol 322: 59-64
108. Vi M (1984) Islet-activating protein, pertussis toxin: a probe for functions of the inhibitory
guanine nucleotide regulatory component of adenylate cyclase. Trends Pharmacol Sci 5: 277-
279
109. Watt AH, Routledge PA (1985) Adenosine stimulates respiration in man. Br J Clin Pharmacol
20:503-506
110. Wayne EJ, Goodwin JF, Stoner HB (1949) The effect of adenosine triphosphate on the electro-
cardiogram of man and animals. Br Heart J 11 : 55-67
111. Webster S, Olsson RA (1981) Adenosine regulation of canine cardiac adenylate cyclase. Bio-
chern PharmacoI30:369-373
112. Welsley RC Jr, Boykin MT, Belardinelli L (1986) Role of adenosine as mediator of bradyar-
rhythmias during hypoxia in isolated guinea pig hearts. Cardiovasc Res 20: 752-759
113. Welsley RC Jr, Lermann BB, DiMarco JP, Berne RM, Belardinelli L (1986) Mechanism of
atropine-resistant atrioventricular block during inferior myocardial infarction: possible role of
adenosine. J Am Coll Cardiol 8: 1232-1234
114. Wessberg P, Hedner J, Hedner T, Presson B, Jonasson J (1985) Adenosine mechanisms in the
regulation of breathing in the rat. Eur J Pharmacol 109: 659-667
115. West GA, Isenberg G, Belardinelli L (1986) Antagonism of forskolin effects by adenosine in
isolated hearts and ventricular myocytes. Am J Physiol250: H769-H777
116. Wu N, Armstrong I, Wagner J (1984) Genetic evidence that chloroadenosine increases the
specific activity of choline acetyltransferase in PC 12 cells via modulation of an adenosine-
dependent adenylate cyclase. Neuroscience 13: 1365-1371
117. Zafirov DH, Palmer JM, Wood JD (1985) Adenosine inhibits forskolin-induced excitation in
myentric neurons. Eur J Pharmacol113: 143-144
118. Zehl U, Ritter CH, Forster W (1976) Influence of prostaglandin upon adenosine release and of
adenosine upon prostaglandin release in the isolated rabbit heart. Acta Bioi Med Germ 35:
K77-K82
Part II
Autonomic Nervous System and Arrhythmias
Chapter 6

Sympathetic Influences on Arrhythmogenesis

in the Ischemic Heart
K.A. YAMADA, G.P. HEATHERS, S.M.POGWIZD, and P. B.CORR

Introduction

Myocardial ischemia frequently results in activation of the sympathetic nervous sys-

tem [146], with clinical manifestations including sinus tachycardia and hypertension
[158]. Of more concern than these hemodynamic alterations produced by sympathetic
neural activation is the electrical instability which ensues after an ischemic event and
during evolving myocardial infarction. Several lines of evidence suggest that in-
creases in sympathetic neural activity induced by ischemia may be of primary impor-
tance in arrhythmogenesis and that stimulation of both a- and ,B-adrenergic receptors
in the myocardium contributes to the electrophysiological derangements leading to
malignant ventricular tachycardia or ventricular fibrillation resulting in sudden car-
diac death. Prior to considering the electrophysiological effects of catecholamines
and their relation to arrhythmogenesis in the ischemic heart, the anatomic and neuro-
physiological substrates underlying this relationship and the mechanisms responsible
for sympathetic activation will briefly be discussed. The present review will concen-
trate solely on the sympathetic branch of the autonomic nervous system. The reader
is referred to other reviews [31, 35] for a more detailed account of the interactions
between the sympathetic and parasympathetic nervous systems.

Mechanisms Responsible for Sympathetic Activation During Ischemia

Sensory Receptors and AlTerent Activation

Three types of receptors reside in the reflexogenic areas of the cardiovascular system:
baroreceptors, chemoreceptors, and mechanoreceptors [11, 14, 24, 40, 59, 112].
These sensory transducers are dispersed widely throughout the heart and vascula-
ture, and communicate with the central nervous system via afferent nerve fibers.
They are responsible for initiating a variety of cardiocardiac and cardiovascular re-
flexes, and thus are sub served by both myelinated and unmyelinated sympathetic
and vagal neurons which comprise the afferent limb of spinal and supraspinal reflex
arcs [127]. Cardiac sympathetic afferents have their cell bodies in the dorsal root
ganglia ofthe T1-T6 segments ofthe spinal cord and synapse in the dorsal hom. Re-
flexes involving these afferents are generally considered to be excitatory [90, 91,115]
and produce increased efferent sympathetic outflow.
80 K. A. Yamada et al.

Baroreceptors transduce tonic deformation stimuli from the carotid sinus and
aortic arch, various sites along the common carotid artery, and within the chambers
of the heart itself into electrical signals which are relayed to the central nervous sys-
tem via the carotid sinus and aortic depressor nerves. The baroreceptor reflex which
originates with this transduction may serve as a sensitive indicator of autonomic
neural status and, in fact, has been used to assess cardiac electrical instability in pa-
tients with myocardial infarction [130].
Sympathetic afferents originating in atrial tissue appear to be activated by atrial
contraction and stretch [92,147], as well as by chemical stimuli [148, 149]. Sympathetic
afferents arising from ventricular mechanosensitive and chemosensitive receptors
appear to be activated by ventricular contraction and distension [21], coronary oc-
clusion [146], increases in ventricular pressure [92, 150], bradykinin [148], and vera-
tridine [91]. These sympathosympathetic excitatory reflexes, aside from playing a
contributing role in arrhythmogenesis, may also mediate cardiac pain, particularly
pain associated with myocardial ischemia [15,16,93]. Thus, ischemia is likely to acti-
vate numerous afferent sympathetic fibers resulting in a net increase in efferent sym-
pathetic neural activity to the heart.

Central Nervous System Control

Net autonomic efferent neural outflow is dependent upon integration of the incom-
ing afferent information with the background preexisting sympathetic tone and other
central influences such as those of the forebrain. The first synapse in the cardiovascu-
lar reflex pathways occurs in the medulla at the nucleus tractus solitarius (NTS) as
demonstrated by denervation studies [32] and anatomic [9] and electrophysiological
[33] procedures. There are nine distinct subnuclei of the NTS, several of which re-
ceive primary cardiovascular afferent input [85]. Sympathetic efferent fibers to the
heart originate in the intermediolateral cell column of the spinal cord. The pre-
ganglionic fibers, whose cell bodies reside in the intermediolateral cell column, ter-
minate in the stellate ganglion which contains the cell bodies of postganglionic ef-
ferent fibers projecting to the heart and vasculature. The multivariate central con-
nections between the first synapse at the NTS and exiting efferents in the inter-
mediolateral cell column appear to involve the ventral surface of the medulla [4,
166], the Al or C1 catecholaminergic cell groups [12, 70, 84, 121, 125], and the in-
ferior olivary complex [138]. Other synapse sites which have been anatomically iden-
tified [84] are likely important as well. In addition, medullary neurons in the raphe
nuclei [5], lateral reticular nucleus, and nucleus gigantocellularis [17] have been anti-
dromically activated by stimulation of the intermediolateral cell column of the spinal
cord and may also be sites of physiological significance in the pathway [63]. Further
electrophysiological and neurophysiological studies of bulbospinal neurons are
needed to characterize the functional significance of anatomically defined pathways
contributing to sympathetic outflow. The spinal cord itself may be an important site
for integration of ascending and descending pathways and may influence the ultimate
magnitude and pattern of sympathetic neural discharge [7].
Based on anatomic [126, 141] and electrophysiological [57] data, forebrain re-
gions also appear to influence medullary pattern generators [50, 64] which determine,
Sympathetic Influences on Arrhythmogenesis 81

in part, background sympathetic tone. Forebrain areas such as the hypothalamus,

limbic system, and parabrachial nucleus may have a more direct role in modulating
cardiovascular control and the associated pathways involved in emotion and other
behavioral states. Several studies have demonstrated associations between central
neural influences and arrhythmogenesis [98, 142]. For example, stress during acute
myocardial ischemia and infarction may be a major contributor to the genesis of
malignant ventricular arrhythmias [136, 137, 152, 153]. Likewise, psychological fac-
tors and environmental stress may be prominent risk factors for ventricular fibril-
lation and sudden death in man [38, 44, 88, 116].

Efferent Nerve Traffic

Several lines of evidence indicate that efferent as well as afferent sympathetic nerves
are activated during myocardial ischemia. For example, occlusion of either the LAD
or circumflex coronary artery in cats enhances efferent sympathetic neural activity
from the T3 ramus coincident with alterations in the ST segment and T wave of the
surface electrocardiogram [93]. Likewise, increased preganglionic efferent sym-
pathetic activity occurs in cats subjected to left main coronary occlusion, associated
with the development of malignant ventricular arrhythmias [51]. Ischemia may result
in a disparate response in overall cardiac sympathetic neural discharge with both in-
creases and decreases in activity in different nerve fibers [26, 81]. This disparate or
asynchronous activation may be particularly arrhythmogenic because of nonuniform
effects on excitability, refractoriness, and conduction in the ischemic heart [60].
Stimulation of cardiac sympathetic nerves has been shown to elicit ventricular ar-
rhythmias [58], including ventricular fibrillation [154]. Conversely, sympathetic de-
nervation has been shown, in several cases, to be protective against ventricular ar-
rhythmias secondary to myocardial ischemia [86, 95, 113, 131]. Specifically, left as
opposed to right stellate ganglion blockade has been shown to be antiarrhythmic
[131]. Furthermore, left stellate ganglionectomy has been shown to increase the ven-
tricular fibrillation threshold [78]. Most reports indicate that increased left stellate
neural activity contributes more to deleterious sympathetic input to the heart during
myocardial ischemia than does increased right stellate neural activity. However,
there have been studies in which either acute or chronic bilateral stellate ganglionec-
tomy did not prevent the occurrence of ventricular fibrillation induced by coronary
artery occlusion [42, 47]. This apparent contradiction may be explained by differing
extents of denervation resulting in variable residual stores of intramyocardial cate-
cholamines which can be released in response to ischemia independent of enhanced
efferent neural activity.

Norepinephrine Release During Ischemia

Although ischemia is associated with enhanced sympathetic neural activity to the

heart, there is conflicting experimental evidence regarding norepinephrine release
within the ischemic zone per se. Although several reports have shown that
norepinephrine may be released directly within the ischemic zone [1,41,65,67,80,
82 K. A. Yamada et al.

133, 164] independent of neural traffic, others have shown that catecholamines are
not released into the local coronary venous effluent during myocardial ischemia
[94, 97, 114, 123, 124]. During the first few minutes after induction of myocardial
ischemia, there does not appear to be a marked increase in catecholamine release
within the ischemic region [2], although release of adrenal catecholamines is promi-
nent [76, 139]. Thus, despite an increase in plasma and urinary catecholamines dur-
ing acute myocardial infarction [36, 46, 49, 55, 96,107,122,155,156], local release
with overflow into the coronary venous effluent is not prominent during early isch-
emia. Likewise, a temporal relationship between catecholamine release and ar-
rhythmogenesis is not usually apparent [53, 122, 156]. However, since reuptake of
norepinephrine into the adrenergic nerve terminal is not altered during early isch-
emia, it is not surprising that enhanced release into the venous effluent does not
occur even though enhanced neural release of catecholamines locally may be sub-
stantial. In addition, venous effluents sampled from the ischemic region may be sig-
nificantly diluted by venous blood from nonischemic regions. There may also be a re-
distribution of catecholamines within the ischemic zone of the dog heart [105]. Myo-
cardial tissue damage and a greater reduction in the percent volume of catechol-
amine-containing nerve terminals than in total tissue catecholamines were observed
after 1 and 3h of coronary occlusion [105]. Finally, catecholamine depletion in iso-
lated hearts reduced the incidence of arrhythmias occurring during global ischemia
and reperfusion [34], suggesting that endogenous myocardial catecholamines contri-
bute to the electrophysiological alterations and arrhythmias during myocardial isch-
emia and reperfusion.

Factors Contributing to Nonuniform Sympathetic InOuences During Ischemia

As previously mentioned, asynchronous sympathetic activation may be particularly

arrhythmogenic because of nonuniform effects on excitability, regional refractory
periods, or conduction in the ischemic heart. After reviewing the sympathetic neural
substrate involved, it is clear that there are several factors which may contribute to
the nonuniform sympathetic influences on cellular electrophysiology during isch-
emia.
Unusual central neural contributions elicited by stressful stimuli may alter normal
sympathetic outflow such that forebrain influences are manifest as sympathetic imba-
lances. Several reflexes induced by ischemia can likewise alter central sympathetic
outflow. Nonuniform sympathetic outflow may then elicit a host oflocal effects rang-
ing from heterogeneous, enhanced catecholamine release, enhanced adrenergic re-
ceptor activation, local tissue lesions, local blood flow alterations, biochemical and
metabolic derangements. These factors will add to those pathological effects pro-
duced by ischemia directly, such as mechanical demise and dyskinesis, which may,
with stretch, stimulate mechanoreceptors setting up additonal reflex activations or
inflammatory cell responses.
These interactions are complex. For example, the ischemic milieu may contribute
to enhanced catecholamine release independent of increases in efferent sympathetic
outflow. Ischemia-induced catecholamine release may be influenced by ionic alter-
ations such as the extent of regional extracellular accumulation of K+ [87]. Extracel-
Sympathetic Influences on Arrhythmogenesis 83

lular K+ concentrations may increase heterogeneously across the ischemic tissue

[66], thus contributing further to the nonuniformity of the sympathetic response.
Sympathetic contributions to the cellular electrophysiology of the ischemic heart
have yet to be systematically explored. These questions will be addressed in the near
future, however, as new insights into the understanding of the mechanisms respon-
sible for myocardial ischemia- and reperfusion-induced arrhythmias have been re-
cently elucidated. Work from our laboratory reveals that both reentrant and non-
reentrant pathways contribute to the initiation and maintenance of ventricular ar-
rhythmias during brief periods of myocardial ischemia without reperfusion [119].
The data were obtained using a computerized three-dimensional mapping system
[163] with which 240 simultaneous intramural sites in the cat heart were examined.
In all but one animal, initiation of ventricular tachycardia occurred in the subendo-
cardium, adjacent to the site of delayed subendocardial and midmyocardial activa-
tion. In 76% of the cases, intramural reentry was responsible for the initiation of
ventricular tachycardia. The remainder of cases demonstrated nonreentry as the
mechanism for initiation as evidenced by the lack of intervening electrical activity be-
tween the end of the preceding sinus beat and the initiation of the tachycardia. Once
initiated, the tachycardia could be maintained or terminated by the same or an alter-
nate mechanism, and both mechanisms could occur during initiation of the same beat
or during the same run of ventricular tachycardia.
In contrast, it was found that ventricular tachycardia during reperfusion occurred
largely due to nonreentrant mechanisms originating in the subendocardium and sub-
epicardium [118]. Although both reentrant and nonreentrant mechanisms usually
maintained the tachycardia, in those cases in which ventricular tachycardia degener-
ated to fibrillation, the mechanism for the transition was due to a rapid acceleration
of a nonreentrant mechanism arising from both the subendocardium and subepicar-
dium. As total activation time actually exceeded the coupling interval of the tachy-
cardia, multiple wave fronts were observed simultaneously. Further studies are on-
going to determine the nature of the nonreentrant mechanisms, the hypothesis being
that early or delayed afterdepolarizations leading to triggered activity may be in-
volved.

Electrophysiological and Arrhythmogenic Effects

of Adrenergic Receptor Stimulation

p-Adrenergic Receptor Stimulation

The physiological effects of sympathetic stimulation in the heart are mediated

primarily through ,B-adrenergic receptors. These effects include acceleration of sinus
node rate [144], increase in AV node conduction velocity [157], decrease in A V junc-
tional and His bundle refractoriness, shortening of Purkinje fiber and ventricular
muscle action potential duration, and increases in automaticity in the His-Purkinje
network [52, 68]. As stated above, particularly in the ischemic heart, the reduction
in refractory period may contribute to arrhythmogenesis. Likewise, an increase in
84 K. A. Yamada et al.

sinus node rate or A V node conduction velocity can enhance the extent of functional
conduction delay in the ischemic tissue.
Experimental evidence obtained over the past several years has delineated the
role of fJ-adrenergic receptor stimulation in maintaining slow conduction and con-
tributing to conduction block through activation of the slow inward current (lsi). De-
polarization of ischemic tissue may inactivate the fast sodium channel, with the result
being that depolarization and conduction may then be dependent exclusively on the
lsi. Catecholamines can increase the magnitude of the lsi and decrease its threshold
for excitation, which apparently occurs through stimulation of ,B-adrenergic recep-
tors and increases in intracellular cyclic AMP [82].
A recent study by Janse and colleagues indicates that left stellate stimulation ac-
centuates the decrease in T-Q depression recorded by DC electrograms within isch-
emic tissue in vivo, suggesting an accentuation of the decrease in resting membrane
potential [74]. It appears likely, therefore, that enhanced sympathetic tone to the left
ventricle during early ischemia contributes to the reduction in resting membrane
potential, although the adrenergic receptor mediating this effect is unknown. Unfor-
tunately, no data are yet available to indicate the precise cellular electrophysiological
effects of ,B-adrenergic stimulation in the ischemic heart in vivo, or the precise mech-
anisms responsible for the beneficial effects of fJ-adrenergic blockade.
Other factors contributing to the arrhythmogenic potential of ,B-adrenergic stimu-
lation include the enhancement of delayed afterdepolarizations and subsequent
triggered activity [161]. These phenomena, characterized in vitro, have yet to be
characterized in the in situ heart or identified as being causally related to arrhythmo-
genesis in vivo. Recently, we demonstrated that lysophosphatidylcholine, which ac-
cumulates in ischemic myocardium primarily in the extracellular compartment [3,
28], could induce delayed afterdepolarizations and triggered activity in isolated tis-
sue, even in the presence of increased extracellular K+ and decreased pH [120].
Finally, direct effects of ,B-adrenergic stimulation to increase myocardial oxygen de-
mand may contribute to the extension of ischemic damage and thereby contribute to
arrhythmogenesis.
Since 1974, several double-blind, randomized trials have evaluated the effective-
ness of ,B-adrenergic blockade on sudden death in postinfarction patients, although
the definition of sudden death was variable. Pooled data from seven of these studies
revealed that treatment led to a 28% reduction in mortality, a 33% reduction in sud-
den cardiac death, and a 20% reduction in nonsudden cardiac death [48], which sug-
gest a primary antiarrhythmic effect of ,B-adrenergic blockade. These are the first re-
ports in which a pharmacological intervention has enhanced survival in the period
after myocardial infarction, demonstrating that increased sympathetic neural activity
mediated through ,B-adrenergic receptor stimulation contributes to mortality associ-
ated with coronary artery disease. Although the precise mechanisms responsible for
the reduction in cardiovascular mortality with chronic fJ-adrenergic blockade are un-
known, they likely include reduction of the extent of myocardial ischemia, a direct
antiarrhythmic effect, and improvement in left ventricular dysfunction. The ob-
served beneficial effects appear to be directly related to the ,B-adrenergic blocking
properties of the drugs rather than to other properties of the compounds, since bene-
ficial effects have been noted with several different agents and doses possessing dif-
ferent nonspecific properties.
Sympathetic Influences on Arrhythmogenesis 85

a-Receptor Stimulation

Under normal physiological conditions, a-adrenergic stimulation results in a de-

crease in automaticity in the sinus node and latent ventricular pacemakers, lengthen-
ing of refractory period, and an increase in the ventricular fibrillation threshold, ef-
fects which would be expected to be primarily antiarrhythmic. However, in depressed
tissue, arrhythmogenic effects of a-adrenergic stimulation may become manifest.
Several lines of experimental evidence indicate that under certain pathophysiological
conditions, the effect of a-adrenergic stimulation may be enhanced. For example,
central neural stimulation with picrotoxin results in ventricular arrhythmias which
may be blocked by the a-adrenergic antagonists phentolamine and tolazoline, and by
denervation, but not by j3-adrenergic antagonists [39]. Both a- and j3-adrenergic
mechanisms appear to contribute to the evolution of ventricular fibrillation in ani-
mals treated with probucol [43]. In dogs with regional ischemia, either epinephrine
or stellate nerve stimulation induces a positive inotropic response mediated through
stimulation of j3-adrenergic receptors in the normal myocardium but through a-
adrenergic receptor stimulation in the ischemic myocardium [72]. al"Adrenergic
blockade has been shown to induce potent antiarrhythmic effects during both coro-
nary occlusion and reperfusion in the cat, independent of hemodynamic alterations
or changes in regional myocardial blood flow [135]. The beneficial antiarrhythmic ef-
fects of al"adrenergic blockade with prazosin have been confirmed in the cat, dog,
and pig [6, 8, 37]. In the dog, systemic administration of phentolamine (0.2mg/kg)
is antiarrhythmic during reperfusion and blocks the associated increase in the effec-
tive refractory period [140]. Intracoronary administration of phentolamine in the dog
at the time of reperfusion completely attenuates the incidence of ventricular fibrilla-
tion [160]. More recent evidence using left stellate ganglion stimulation within a brief
2-min ischemic interval indicates that al-adrenergic blockade with prazosin (100 JIg/
kg) reduces the incidence of ventricular complexes but is less effective in reducing
ventricular fibrillation [132, 151]. Several studies in the rat also indicate that a-
adrenergic blockade is antiarrhythmic during both coronary occlusion and reperfu-
sion [74, 143]. In contrast, one report indicates that neither prazosin (1 mg/kg) nor
phentolamine (5 mg/kg) significantly altered the incidence of ventricular fibrillation
associated with reperfusion in the dog [13]. This apparent discrepancy may be due to
the relatively high doses of a-adrenergic blocking agents for the dog used in this lat-
ter study resulting in a marked decrease in systemic pressure necessitating aortic con-
striction to maintain perfusion. Recently, a protective effect of al"adrenergic blockade
has been demonstrated in a conscious canine model of sudden cardiac death [159].
Prazosin (100-500 JIg/kg) significantly reduced the incidence of ventricular fibrilla-
tion during acute posterolateral ischemia in conscious dogs with a previous 5-day-old
anterior myocardial infarction. Although prazosin did not alter inducible ventricular
tachycardia elicited during baseline programmed stimulation, and therefore had no
effect on normal or necrotic tissue, al-adrenergic blockade did prevent ventricular
fibrillation during the subsequent ischemic interval.
Although the mechanisms responsible for the protective effect of al"adrenergic
blockade during ischemia appear to be complex, current research from our laboratory
and others has provided important clues. For example, in rabbit papillary muscle de-
polarized by high [K+]o (to partially simulate ischemic conditions), a-adrenergic
86 K.A. Yamada et al.

mechanisms appear to increase the lsi and maintain depolarization [100]. Indeed,
phenylephrine increased the magnitude of the Isimax and decelerated the decay of the
lsi in bovine hearts, without altering outward potassium currents [20]. In vivo, the
augmentation of contractility in response to stellate nerve stimulation may be medi-
ated through fJ-adrenergic receptors in the normal zone, but by a-adrenergic recep-
tors in the ischemic region [72]. Both of these effects may involve modulation of
intracellular calcium. The mechanism whereby a-adrenergic stimulation activates
calcium influx in the heart is not clear. It has been known for a number of years that
myocardial at-adrenergic receptors, in several species including man, mediate posi-
tive inotropic influences independent of increases in cAMP [19,56, 128, 129].
Reperfusion in the cat after coronary occlusion results in an increased idioven-
tricular rate, which can be abolished by a-adrenergic blockade or catecholamine de-
pletion with 6-hydroxydopamine but not by ,B-adrenergic blockade [135]. In addi-
tion, the idioventricular rate can be enhanced during reperfusion by regional infu-
sion of methoxamine, an a-agonist, into the reperfused zone [135]. This finding is in
contrast to the fJ-mediated effects on idioventricular rate in nonischemic animals.
Thus, at-adrenergic stimulation may result in electrophysiological derangements
during ischemia and reperfusion which are not evident in normal tissue. Additional
work from our laboratory suggests that a-adrenergic mechanisms may mediate a
large portion of the increase in intracellular calcium observed during reperfusion
[134]. In untreated animals subjected to 35 min of coronary occlusion followed by re-
perfusion, there is a twofold increase in total tissue calcium. Using measurements of
the vascular space etCr-red blood cells) and the interstitium eH-inulin), the increase
was shown to be largely intracellular [134]. Ultrastructural localization of calcium in
the reperfused myocardium was confirmed by pyroantimonate precipitation of the
calcium and X-ray microprobe analysis [134]. a-Adrenergic blockade, even 2 min
prior to reperfusion, completely prevented the increase in intracellular calcium. This
calcium accumulation during reperfusion in reversibly injured tissue is not a passive
process but appears to be modulated by at-adrenergic stimulation. This may be im-
portant not only in the progression of cellular damage but also in arrhythmogenesis.
Using microelectrodes and membrane patches derived from cultured myocytes, an
inward current has been characterized that is activated by increased cytosolic calcium
[25]. In vivo, during ischemia or reperfusion, a-adrenergic stimulation may enhance
this inward current by increasing cytosolic calcium. This inward current may contri-
bute to depolarization, which is likely to be slow, and may also contribute to the
development of delayed afterdepolarizations and triggered activity. This type of re-
sponse may be involved in the accelerated idioventricular rates seen during reperfu-
sion in both animals [135] and man [54].
One explanation for the enhanced a-adrenergic responsivity seen in vivo during
reperfusion [135] may be related to the twofold increase in aradrenergic receptor
density confined to the ischemic region, with no alteration in receptor affinity [29].
The increase in arreceptor density returns to control values during sustained reper-
fusion with a time course corresponding to the decrease in a-adrenergic responsivity
[29, 135]. Similar increases in at-adrenergic receptors have been observed in the
guinea pig heart in response to ischemia [89]. In the dog, both at-adrenergic [104]
and fJ-adrenergic receptor number [103] increases in ischemic myocardium. The in-
crease in ,B-receptor number does not occur until 1 h of ischemia and is associated
Sympathetic Influences on Arrhythmogenesis 87

with an increased tissue production of cAMP and phosphorylase kinase activity in re-
sponse to isoproterenol. However, several problems exist with the present methods
ofreceptor analysis. They include: (a) disruption of normal cellular morphology dur-
ing membrane preparation preceding classic radioligand binding assays, (b) contami-
nation of membrane preparations with membrane fractions from nonmyocytic cells,
(c) inability to measure ligand binding to cell surface receptors on intact myocytes,
and (d) the inability to localize the receptor change on or within the myocyte.
To characterize and evaluate the potential mechanisms responsible for the isch-
emia-induced alterations in adrenergic receptors observed in vivo, we have recently
developed procedures using isolated adult canine myocytes which have been exten-
sively characterized by light and electron microscopy, electrophysiological and bio-
chemical parameters, and intracellular enzyme release [61]. Radioligand binding as-
says are readily performed on these suspensions of intact, calcium-tolerant myo-
cytes, thus precluding several of the problems outlined above.
Previous work from our laboratory indicates that exogenous delivery of long-
chain acy1carnitines results in electrophysiological abnormalities in vitro [30] analog-
ous to those alterations seen during ischemia in vivo [69, 83]. Likewise, hypoxia in
cultured neonatal myocytes results in a marked increase in total long-chain acyl-
carnitines (LCA) [79] similar to that seen during ischemia in vivo [69, 83], and a 70-
fold increase in LCA content in the sarcolemma as determined by electron-micro-
scopic autoradiography of myocytes prelabeled with 3H-carnitine [79]. Most impor-
tantly, pretreatment of myocytes with sodium 2-[5-( 4-chlorophenyl)-pentyl]-oxirane-
2-carboxylate (POCA), a potent inhibitor of carnitine acyltransferase I, prevents not
only the sarcolemmal accumulation of LCA but also dramatically attenuates the
electrophysiological derangements induced by hypoxia [79]. Since LCA are potent
amphiphiles which can perturb membrane function in a variety of systems, we rea-
soned that their sarcolemmal accumulation may alter the exposure of adrenergic re-
ceptors residing within or near the sarcolemma. Using the isolated adult myocyte
preparation described above, we have recently shown that al-adrenergic receptor
number increases approximately threefold after only 10 min of hypoxia at 37°C, prior
to any evidence of irreversible damage as assessed by release of intracellular en-
zymes [61]. The increased density of al-receptors was reversible after lOmin of re-
oxygenation. The concentration of LCA also increased threefold on exposure to
hypoxia. Inhibition of carnitine acyltransferase I with POCA completely inhibited
the hypoxia-induced increase in LCA as well as the increase in aradrenergic recep-
tors [61]. Finally, incubation of cells with exogenous palmitoyl carnitine (lpM) for
10 min also increased aradrenergic receptor number under normoxic conditions in
the presence or absence of POCA [61]. These data indicate that sarcolemmal ac-
cumulation of endogenous LCA is critical to the hypoxia-induced increase in al-
adrenergic receptors on adult canine myocytes and may be the mechanism whereby
al-adrenergic receptors are increased during ischemia in vivo. Further avenues of in-
vestigation which are presently being explored include the origin and migration of the
newly exposed aradrenergic receptors in the isolated ventricular myocytes exposed to
hypoxia. In vitro autoradiographic binding has been used for localizing a-adrenergic
receptors on cultured myocytes [106] and could be applied to these questions as well.
The mechanisms whereby the newly exposed al-adrenergic receptors may be
coupled to intracellular events have yet to be elucidated. There is substantial evi-
88 K.A. Yamada et al.

dence to suggest that a-adrenergic stimulation in many tissues can lead to increased
intracellular calcium [27]. The likely mechanism involves aI-adrenergic receptor-
mediated hydrolysis of plasma membrane phosphatidylinositol 4,5-bisphosphate
(PIP2) to produce the calcium-mobilizing second messenger inositol1,4,5-trisphos-
phate (IP3) and diacylglycerol [10, 22, 23, 99]. The role of diacylglycerol in signal
transduction has been reviewed extensively elsewhere [108, 109] and will not be con-
sidered here. Additionally, we have recently reviewed the molecular mechanisms in-
volved in the coupling of both al- and P.adrenergic receptor stimulation to the intra-
cellular responses mediating the electrophysiological alterations responsible for ar-
rhythmogenesis [62].
To date, definitive data implicating IP3 as the second messenger in cardiac tissue
do not exist. Several studies have demonstrated that aI-adrenergic stimulation will
increase eH]inositol incorporation into inositol I-phosphate (IP I) or IP3 in cardiac
tissue [18, 73, 111, 117]. However, when measurement of [3H]IPI accumulation is
used to assess phosphatidylinositol turnover, it is possible that agonist-stimulated
hydrolysis of phosphatidylinositol is being measured, without any increase in IP3
concentration from PIP2 hydrolysis. Similarly, measurement of [3H]IP3, in the pres-
ence of Liel to inhibit inositol I-phosphatase, does not permit measurement of the
rapid and transient response to aI-adrenergic stimulation.
To circumvent these problems, we have recently developed a technique to mea-
sure the mass of inositol phosphates after short periods of catecholamine stimulation
in the isolated adult canine myocytes. Pilot studies have demonstrated 80% recovery
through successive ion-exchange column separations and consistent background
levels of myoinositol as detected by gas chromatography with chiroinositol as the
internal standard. Preliminary results have revealed a fourfold increase in IP3 con-
tent after only 60s of norepinephrine stimulation, after which a rapid decline to con-
trollevels occurs. Thus, IP3 production does appear to occur in cardiac myocytes in
response to al-adrenergic receptor stimulation.
Our previous data demonstrate that aI-adrenergic receptor blockade at the time
of reperfusion prevents intracellular [134], including intramitochondrial [134, 165],
calcium accumulation. We are currently investigating the role of aI-adrenergic recep-
tor activation in transsarcolemmal calcium influx or intracellular calcium mobiliza-
tion in the isolated adult myocytes. Recently two groups of investigators have re-
ported that IP3 enhanced release of intracellular calcium from the sarcoplasmic re-
ticulum in permeabilized cardiac cells [45, 110], while another group failed to dem-
onstrate a stimulatory effect of IP3 on calcium release [101] based on quin2 fluores-
cence.

Role of Calcium in the Arrhytbmogenic Effects of a-Adrenergic Stimulation

in the Ischemic Heart

Several lines of evidence outlined above suggest that the effects of aI-adrenergic
receptor activation in ischemic myocardium are mediated by an influx of calcium into
the myocyte or mobilization of calcium from a store within the cell. Our data have
shown that aI-adrenergic stimulation mediates the movement of calcium into the
myocyte in response to reperfusion of reversibly injured tissue [134, 165]. We have
Sympathetic Influences on Arrhythrnogenesis 89

also shown that a-adrenergic blockade (lmg/kg Lv. phentolamine) just 2 min prior
to reperfusion completely prevented mitochondrial calcium accumulation and at-
tenuated reperfusion-induced mitochondrial dysfunction as compared with non-
treated control animals [165]. The data demonstrate that blockade of a-adrenergic
receptors during reperfusion in reversibly injured tissue enhances mitochondrial
functional recovery and prevents mitochondrial calcium accumulation independent
of changes occurring during the antecedent ischemic interval.
It has also been reported that an increase in intracellular calcium may lead to a
transient inward current, carried primarily by Na+, leading to delayed afterdepolari-
zations [145]. Agents which decrease the intracellular calcium concentration either
directly, or by reducing influx via the Ish decrease the transient inward current [75]
and may attenuate the amplitude of delayed afterdepolarizations [162]. Conversely,
aI-adrenergic receptor agonists, which also increase intracellular calcium levels in
ischemic tissue, may contribute to the occurrence of delayed afterdepolarizations.
The enhancement of delayed afterdepolarizations by catecholamines has been shown
to be inhibited by ,B-adrenergic blockade in normoxic tissue in vitro, but by a-adren-
ergic blockade during hypoxia [102]. In addition, a-adrenergic receptor activation
can augment triggered activity arising from delayed afterdepolarizations in Purkinje
fibers overlying a healed infarct [77]. Thus, in the ischemic myocardium in which
newly exposed aI-adrenergic receptors are unmasked, enhanced activation appears
to lead to increases in intracellular calcium. The enhanced calcium may lead to acti-
vation of the transient inward current which may, in tum, elicit membrane oscilla-
tions resulting in delayed afterdepolarizations and triggered activity which could be
the electrophysiological substrate responsible for the initiation and maintenance of
malignant ventricular arrhythmias occurring during both myocardial ischemia and
reperfusion.

Conclusions and Future Directions

It appears that both a- and fJ-adrenergic receptor stimulation contributes to ar-

rhythmogenesis during myocardial ischemia. The a-adrenergic component appears
to depend on exposure of latent receptors which in tum may be coupled via IP3 pro-
duction to an increase in intracellular calcium. Current studies aimed at elucidating
the biochemical and molecular events involved, particularly with respect to phos-
phoinositide metabolism and the modulation of the aradrenergic receptor response,
should enhance our understanding of the mechanisms responsible for ischemia-induced
arrhythmias. In particular, the question of whether there is a synergistic effect of
diacylglycerol-activated protein kinase C activity and IP3-mediated increases in intra-
cellular calcium should be addressed. Future studies should be directed at delineat-
ing the precise in vivo electrophysiological derangements elicited by a- and ,B-adren-
ergic stimulation in the ischemic heart, independent of alterations in regional coro-
nary flow. The time course, regional transmural distribution, and sarcolemmal/intra-
cellular localization of the changes in adrenergic receptors during ischemia as well as
reperfusion require clarification as well. Further investigations will be required to
answer the following questions:
90 K.A. Yamada et al.

1. What are the membrane mechanisms responsible for the alterations in adrenergic
receptors during ischemia? How does sarcolemmal accumulation of LeA unmask
latent receptors? Do other metabolites such as lysophosphoglycerides induce the
same effects on receptor density?
2. Does the increase in cytosolic calcium during reperfusion also occur during isch-
emia in response to al-adrenergic stimulation?
3. Are the electrophysiological effects of al-adrenergic stimulation during ischemia
and reperfusion mediated through a transmembrane influx of calcium?
4. Are sympathetic influences responsible for the occurrence of delayed after-
depolarizations in vivo? Are early or delayed afterdepolarizations responsible for
the nonreentrant mechanisms leading to initiation and maintenance of ventricular
tachycardia and ventricular fibrillation during myocardial ischemia and reperfusion?
Finally, as increasing evidence implicates al-adrenergic receptor activation as a
factor in the induction and maintenance of malignant arrhythmias, the question of
whether a-adrenergic blockade decreases the incidence of sudden cardiac death in
man warrants investigation.
Acknowledgments. Research from the authors' laboratory was supported by National Institutes of
Health Grants HL-17646, SCOR in Ischemic Heart Disease, grant HL-28995 and by grant HL-
36773. This work was done during the tenure of an Established Investigatorship (Dr. Corr) of the
American Heart Association and with funds contributed in part by the Missouri Heart Affiliate.
Dr. Heathers is the recipient of a Research Fellowship of the American Heart Association, Missouri
Affiliate, and Dr. Pogwizd was supported under a research training grant from the National Insti-
tutes of Health, HL-07275.

References

1. Abrahamsson T, Almgren 0, Holmgren S (1982) Effects of ganglionic blockade on noradrena-

line release and cell injury in the acutely ischemic rat myocardium. J Cardiovasc Pharmacol4:
584-591
2. Abrahamsson T, Holmgren S, Almgren 0 (1982) Noradrenaline release in acute myocardial
ischaemia, a fluorescence-histochemical and biochemical study. In: Parratt JR (ed) Early ar-
rhythmias resulting from myocardial ischaemia. Mechanisms and prevention by drugs. Mac-
millan, London, pp 153-169
3. Akita H, Creer MH, Yamada KA, Sobel BE, Corr PB (1986) Electrophysiologic effects of
intracellular lysophosphoglycerides and their accumulation in cardiac lymph with myocardial
ischemia in dogs. J Clin Invest 78:271-280
4. Amendt K, Czachurski J, Dembowsky K, Seller H (1978) Neurones within the "chemosensitive
area" on the ventral surface of the brainstem which project to the intermediolateral column.
Pfliigers Arch Ges Physiol 375: 289-292
5. Amendt K, Czachurski J, Dembowsky K, Seller H (1979) Bulbospinal projections to the inter-
mediolateral cell column; a neuroanatomical study. J Auton Nerv Syst 1: 103-117
6. Aubry ML, Davey MJ, Petch B (1985) Cardioprotective and antidysrhythmic effects of aj"
adrenoceptor blockade during myocardial ischaemia and reperfusion in the dog. J Cardiovasc
Pharmacol [SuppI6] 7:S93-S102
7. Barman SM (1984) Spinal cord control of the cardiovascular system. In: Randall WC (ed)
Nervous control of cardiovascular function. Oxford University Press, New York, pp 321-345
8. Benfey BG, Elfellah MS, Ogilvie RI, Varma DR (1984) Anti-arrhythmic effects of prazosin
and propranolol during coronary artery occlusion and reperfusion in dogs and pigs. Br J Phar-
macoI82:717-725
Sympathetic Influences on Arrhythmogenesis 91

9. Berger AJ (1979) Distribution of carotid sinus nerve afferent fibers to solitary tract nuclei of the
cat using transganglionic transport of horseradish peroxidase. Neurosci Lett 14: 153-158
10. Berridge MJ, Irvine RF (1984) Inositol trisphosphate, a novel second messenger in cellular
signal transduction. Nature 312: 315-321
11. Bishop VS, Malliani A, Thoren P (1983) Cardiac mechanoreceptors. In: Shepherd JT, Abboud
FM (eds) Handbook of physiology, sect 2, vol 3. American Physiological Society, Bethesda, pp
677-689
12. Blessing WW, Reis DJ (1982) Inhibitory cardiovascular function of neurons in the caudal
ventrolateral medulla of the rabbit: relationship to the area containing Al noradrenergic cells.
Brain Res 253: 161-171
13. Bolli R, Fisher DJ, Taylor AA, Young JB, Miller RR (1984) Effect of a-adrenergic blockade
on arrhythmias induced by acute myocardial ischemia and reperfusion in the dog. J Mol Cell
CardioI16:1101-1117
14. Brown AM (1979) Cardiac reflexes. In: Berne RM (ed) Handbook of physiology, sect 2, vol 1.
American Physiological Society, Bethesda, pp 677-689
15. Brown AM (1968) Excitation of afferent cardiac sympathetic nerve fibers during myocardial
ischaemia. J Physiol (Lond) 190: 35-53
16. Brown AM, Malliani A (1971) Spinal sympathetic reflexes initiated by coronary receptors.
J Physiol (Lond) 212: 685-705
17. Brown DL, Guyenet PG (1984) Cardiovascular neurons of brain stem with projections to spinal
cord. Am J Physiol 247: RlO09-RlO16
18. Brown JH, Buxton IL, Brunton LL (1985) arAdrenergic and muscarinic cholinergic stimula-
tion of phosphoinositide hydrolysis in adult rat cardiomyocytes. Circ Res 57: 532-537
19. Bruckner R, Meyer W, Mugge A, Schmitz H, Scholz H (1984) Alpha-adrenoceptor-mediated
positive inotropic effect of phenylephrine in isolated human ventricular myocardium. Eur J
PharmacoI99:345-347
20. Bruckner R, Scholz H (1984) Effects of alpha-adrenoceptor stimulation with phenylephrine in
the presence of propranolol on force of contraction, slow inward current and cyclic AMP
content in the bovine heart. Br J Pharmacol 82: 223-232
21. Casati R, Lombardi F, Malliani A (1979) Afferent sympathetic unmyelinated fibers with left
ventricular endings in cats. J Physiol (Lond) 292: 135-148
22. Charest R, Blackmore PF, Berthon B, Exton JH (1983) Changes in free cytosolic Ca2+ in
hepatocytes following a)-adrenergic stimulation. Studies on quin-2-loaded hepatocytes. J Bioi
Chern 258: 8769-8773
23. Charest R, Prpic V, Exton JH, Blackmore PF (1985) Stimulation of inositol trisphosphate for-
mation in hepatocytes by vasopressin, adrenaline and angiotensin II and its relationship to
changes in cytosolic free Ca2+. Biochem J 227 : 79-90
24. Coleridge JG, Coleridge H (1979) Chemoreflex regulation of the heart. In: Berne RM (ed)
Handbook of physiology, sect 2, vol 1. American Physiological SOCiety, Bethesda, pp 653-676
25. Colquhoun D, Neher E, Reuter H, Stevens CF (1981) Inward current channels activated by
intracellular Ca in cultured cardiac cells. Nature 294: 752-754
26. Constantin L (1963) Extracardiac factors contributing to hypotension during coronary occlu-
sion. Am J Cardiol 11 : 205-217
27. Corr PB, Sharma AD (1984) a-Adrenergic-mediated effects on myocardial calcium. In: Opie
LH (ed) Calcium antagonists and cardiovascular disease. Raven, New York, pp 193-204
28. Corr PB, Gross RW, Sobel BE (1984) Amphipathic metabolites and membrane dysfunction in
ischemic myocardium. Circ Res 55: 135-154
29. Corr PB, Shayman JA, Kramer JB, Kipnis RJ (1981) Increased a-adrenergic receptors in
ischemic cat myocardium: a potential mediator of electrophysiological derangements. J Clin
Invest 67 : 1232-1236
30. Corr PB, Snyder DW, Cain ME, Crafford WA, Gross RW, Sobel BE (1981) Electrophysio-
logical effects of amphiphiles on canine Purkinje fibers: implications for dysrhythmia secondary
to ischemia. Circ Res 49: 354-363
31. Corr PB, Yamada KA, Witkowski FX (1986) Mechanisms controlling cardiac autonomic func-
tion and their relation to arrhythmogenesis. In: Fozzard HA, Haber E, Jennings KB, Katz AM
(eds) The heart and cardiovascular system. Raven, New York, pp 1343-1403
32. Cottle MK (1979) Degeneration studies of primary afferents of the IXth and Xth cranial nerves
in the cat. J Comp Neuro1122: 329-343
92 K. A. Yamada et al.

33. Crill WE, Reis DJ (1968) Distribution of carotid sinus and depressor nerves in cat brain stem.
Am J PhysioI214:269-276
34. Culling W, Penny WJ, Lewis MJ, Middleton K, Sheridan DJ (1984) Effects of myocardial cate-
cholamine depletion on cellular electrophysiology and arrhythmias during ischaemia and reper-
fusion. Cardiovasc Res 18: 675-682
35. Dampney RAL (1981) Functional organization of central cardiovascular pathways. Clin Exp
Pharmacol PhysioI8:241-259
36. Daugherty A, Frayn KN, Redfern WS, Woodward B (1986) The role of catecholamines in the
production of ischaemia-induced ventricular arrhythmias in the rat in vivo and in vitro. Br J
PharmacoI87:265-277
37. Davey MJ (1980) Relevant features of the pharmacology of prazosin. J Cardiovasc Pharmacol
2: S287-S301
38. DeSilva RA (1982) Central nervous system risk factors for sudden cardiac death. Ann NY
Acad Sci 382: 143-161
39. DiMicco JA, Prestel T, Pearle DL, Gillis RA (1977) Mechanism of cardiovascular changes pro-
duced in cats by activation of the central nervous system with picrotoxin. Circ Res 41: 446-451
40. Downing SE (1979) Baroreceptor regulation of the heart. In: Berne RM (ed) Handbook of
physiology, sect 2, vol 1. American Physiological Society, Bethesda, pp 621-652
41. Dutta SN, Booker WM (1970) Possible myocardial adaption to acute coronary occlusion: rela-
tion to catecholamines. Arch Int Pharmacodyn Ther 185 : 5-12
42. Ebert PA, Vanderbeek RB, Allgood RJ (1970) Effect of chronic cardiac denervation on ar-
rhythmias after coronary artery ligation. Cardiovasc Res 4: 141-147
43. Elharrar V, Watanabe AM, Molello J, Besch HR Jr, Zipes DP (1979) Adrenergically mediated
ventricular fibrillation in probucol-treated dogs: roles of alpha and beta adrenergic receptors.
PACE 2:435-443
44. Eliot RS, Buell JC (1985) Role of emotions and stress in the genesis of sudden death. J Am Coli
Cardiol 5 : 95B-98B
45. Fabiato A (1986) Inositol (1,4,5)-trisphosphate-induced release of Ca2+ from the sarcoplasmic
reticulum of skinned cardiac cells. Biophys J 49: 190a
46. Forssman 0, Hansson G, Jensen C (1952) The adrenal function in coronary thrombosis. Acta
Med Scand 142:441-449
47. Fowlis RAF, Sang CTM, Lundy PM, Ahuja SP, Calhoun H (1974) Experimental coronary
artery ligation in conscious dogs six months after bilateral cardiac sympathectomy. Am Heart J
88:748-757
48. Frishman WH, Furberg CD, Friedewald WT (1984) P.Adrenergic blockade for survivors of
acute myocardial infarction. N Engl J Med 310: 830-837
49. Gazes PC, Richardson JA, Woods EF (1959) Plasma catecholamine concentrations in myo-
cardial infarction and angina pectoris. Circulation 19:657-661
50. Gebber GL (1984) Brainstem systems involved in cardiovascular regulation. In: Randall WC
(ed) Nervous control of cardiovascular function. Oxford University Press, New York, pp 346-
368
51. Gillis RA (1971) Role of the nervous system in the arrhythmias produced by coronary occlusion
of the cat. Am Heart J 81 : 677-684
52. Giotti A, Ledda F, Mannaioni PF (1973) Effects of noradrenaline and isoprenaline, in combi-
nation with a- and p.receptor blocking substances, on the action potential of cardiac Purkinje
fibers. J Physiol (Lond) 229: 99-113
53. Godin D, Campeau N, Nadeau R, Cardinal R, de Champlain J (1985) Catecholamine release
and ventricular arrhythmias during coronary occlusion and reperfusion in the dog. Can J
Physiol Pharmacol 63 : 1088-1095
54. Goldberg S, Greenspon AJ, Urban PL, Muza B, Berger B, Walinsky P, Maroko PR (1983)
Reperfusion arrhythmias: a marker of restoration of antegrade flow during intracoronary
thrombolysis for acute myocardial infarction. Am Heart J 105: 26-32
55. Goldstein DS (1981) Plasma norepinephrine as an indicator of sympathetic neural activity in
clinical cardiology. Am J Cardiol48: 1147-1154
56. Govier WC (1967) A positive inotropic effect of phenylephrine mediated through alpha adren-
ergic receptors. Life Sci 6: 1361-1365
57. Grizzle WE, Johnson RN, Schramm LP, Gann DS (1975) Hypothalamic cells in an area
mediating ACTH release respond to right atrial stretch. Am J Physiol 228: 1039-1045
Sympathetic Influences on Arrhythrnogenesis 93

58. Hageman GR, Goldberg JM, Armour JA, Randall WC (1973) Cardiac dysrhythmias induced
by autonomic nerve stimulation. Am J Cardiol 32: 823-830
59. Hainsworth R, Kidd C, Linden RJ (1979) Cardiac receptors. Cambridge University Press,
Cambridge
60. Han J, Moe GK (1964) Nonuniform recovery of excitability in ventricular muscle. Circ Res 14:
44-60
61. Heathers GP, Yamada KA, Kanter EM, Corr PB (1987) Long-chain acylcarnitines mediate the
hypoxia-induced increase in alphal-adrenergic receptors on adult canine myocytes. Circ Res
61:735-746
62. Heathers GP, Yamada KA, Pogwizd SM, Corr PB (1988) The contribution of a- and fj-adren-
ergic mechanisms in the genesis of arrhythmias during myocardial ischemia and reperfusion. In:
Kulbertus HE, Franck G (eds) Neurocardiology. Futura Publishing, Mount Kisco, pp 143-178
63. Henry JL, Calaresu FR (1974) Excitatory and inhibitory inputs from medullary nuclei project-
ing to spinal cardioacceleratory neurons in the cat. Exp Brain Res 20: 485-504
64. Hilton SM, Spyer KM (1971) Participation of the anterior hypothalamus in the baroreceptor
reflex. J Physiol (Lond) 218:271-293
65. Hirche Hj, McDonald FM, Polwin W, Addicks K (1985) Vicious cycle of catecholamines and
K+ in cardiac ischemia. J Cardiovasc Pharmacol [Suppl 5] 7: S71-S75
66. Hirche Hj, Kebbel U, McDonald FM, Knopf H, Bischoff A, Barth A (1985) Measurement of
inhomogeneous changes of extracellular K+ concentration within the ischemic myocardium.
In: Kessler M, Harrison DK, Hoper J (eds) Ion measurements in physiology and medicine.
Springer, Berlin Heidelberg New York, pp 269-273
67. Hirche Hj, Addicks K, Deutsch HJ, Friedrich R, Griebenow R, McDonald FM, Zylka V
(1981) The effect of lignocaine on the release of K+ and of noradrenaline from ischemic pig
heart. Pfliigers Arch Ges Physiol 389: R5
68. Hordof AJ, Rose E, Danilo P Jr, Rosen MR (1982) a- and fj-adrenergic effects of epinephrine
on ventricular pacemakers in dogs. Am J Physiol 242: 677-682
69. Idell-Wenger JA, Grotyohann LW, Neely JR (1978) Coenzyme A and carnitine distribution in
normal and ischemic hearts. J BioI Chern 253: 4310-4318
70. Imaizumi T, Granata AR, Sved AF, Benarroch E, Reis DJ (1983) Lesions of the Al area of rat
ventrolateral medulla increase cardiac output and arterial pressure without modifying mechano-
receptor reflexes. Soc Neurosci Abstr 9: 182
71. Janse MJ, Schwartz PJ, Wilms-Schopman F, Peters RJG, Durrer D (1985) Effects of unilateral
stellate ganglion stimulation and ablation on electrophysiologic changes induced by acute myo-
cardial ischemia in dogs. Circulation 72: 585-595
72. Juhasz-Nagy A, Aviado DM (1976) Increased role of alpha-adrenoreceptors in ischemic myo-
cardial zones. Physiologist 19: 245
73. Kaku T, Lakatta E, Filburn CR (1986) Effect of al-adrenergic stimulation on phosphoinositide
metabolism and protein kinase C (PK-C) in rat cardiomyocytes. Fed Proc 45: 209
74. Kane KA, Parratt JR, Williams FM (1984) An investigation into the characteristics of reper-
fusion-induced arrhythmias in the anaesthetized rat and their susceptibility to antiarrhythmic
agents. Br J Pharmacol 82: 349-357
75. Kass RS, Lederer WJ, Tsien RW, Weingart R (1978) Role of calcium ions in transient inward
currents and aftercontractions induced by strophanthidin in cardiac purkinje fibres. J Physiol
(Lond) 281: 187-298
76. Kelliher GJ, Widmar C, Roberts J (1975) Influence of adrenal medulla on cardiac rhythm
disturbances following acute coronary artery occlusion. In: Roy PE, Rona G (eds) Recent
advances in cardiac structure and metabolism, vol 10. Park University Press, Baltimore, pp
387-400
77. Kimura S, Bassett AL, Kohya T, Kozlovskis PL, Myerburg RJ (1987) Automaticity, triggered
activity, and responses to adrenergic stimulation in cat subendocardial Purkinje fibers after
healing of myocardial infarction. Circulation 75: 651-660
78. Kliks BR, Burgess MJ, Abildskov JA (1975) Influence of sympathetic tone on ventricular fibril-
lation threshold during experimental coronary occlusion. Am J Cardiol 36: 45-49
79. Knabb MT, Saffitz JE, Corr PB, Sobel BE (1986) The dependence of electrophysiologic derange-
ments on accumulation of endogenous long-chain acyl carnitine in hypoxic rat myocytes. Circ
Res 58: 230-240
94 K. A. Yamada et al.

80. Lammerant J, Delterdt P, DeSchryver C (1966) Direct release of myocardial catecholamines

into the left heart chambers; the enhancing effect of acute coronary occlusion. Arch Int Phar-
macodyn Ther 163: 219-226
81. Lathers CM, Kelliher GJ, Roberts J, Beasley AB (1978) Nonuniform cardiac sympathetic
nerve discharge: mechanism for coronary occlusion and digitalis induced arrhythmia. Circula-
tion 57: 1058-1065
82. Li T, Sperelakis N (1983) Stimulation of slow action potentials in guinea pig papillary muscle
cells by intracellular injection of cAMP, Gpp(NH)p, and cholera toxin. Circ Res 52: 111-117
83. Liedtke AJ, Nellis S, Neely JR (1978) Effects of excess free fatty acids on mechanical and
metabolic function in normal and ischemic myocardium in swine. Circ Res 43 : 652-661
84. Loewy AD (1981) Descending pathways to sympathetic and parasympathetic preganglionic
neurons. J Auton Nerv Syst 3: 265-275
85. Loewy AD, Burton H (1978) Nuclei of the solitary tract. Efferent projections to the lower
brainstem and spinal cord of the cat. J Comp Neurol181: 421-450
86. Lombardi F, Verrier RL, Lown B (1983) Relationship between sympathetic neural activity,
coronary dynamics, and vulnerability to ventricular fibrillation during myocardial ischemia and
reperfusion. Am Heart J 105: 958-965
87. Lorenz RR, Vanhoutte PM (1975) Inhibition of adrenergic neurotransmission in isolated veins
of the dog by potassium ions. J Physiol (Lond) 246: 479-484
88. Lown B, DeSilva RA, Reich P, Murawski BJ (1980) Psychophysiologic factors in sudden
cardiac death. Am J Psychiatry 137: 1325-1335
89. Maisel AS, Motulsky HJ, Insel PA (1986) Receptor traffic in the myocardium: Comparison of
a- and fJ-adrenergic receptors. JAm Coll Cardiol [Suppl A] 7: 80A
90. Malliani A, Parks M, Tuckett RP, Brown AM (1973) Reflex increases in heart rate elicited by
stimulation of afferent cardiac sympathetic nerve fibers in the cat. Circ Res 32:9-14
91. Malliani A, Peterson DF, Bishop VS, Brown AM (1972) Spinal sympathetic cardiocardiac
reflexes. Circ Res 30: 158-166
92. Malliani A, Recordati G, Schwartz PJ (1973) Nervous activity of afferent cardiac sympathetic
fibers with atrial and ventricular endings. J Physiol (Lond) 29: 457-469
93. Malliani A, Schwartz PJ, Zanchetti A (1969) A sympathetic reflex elicited by experimental
coronary occlusion. Am J Physiol 217: 703-709
94. Marshall RM, Parratt JR (1980) The early consequences of myocardial ischaemia and their
modification. J Physiol (Paris) 76:699-715
95. Martin C, Meesmann W (1985) Antiarrhythmic effect of regional myocardial chemical sym-
pathectomy in the early phase of coronary artery occlusion in dogs. J Cardiovasc Pharmacol
[Suppl 5] 7: S76-S80
96. McDonald L, Baker C, Bray C, McDonald A, Restieaux N (1969) Plasma-catecholamines after
cardiac infarction. Lancet 2: 1021-1023
97. McGrath BP, Lim SP, Leversha L, Shanahan A (1981) Myocardial and peripheral catechol-
amine responses to acute coronary artery constriction before and after propranolol treatment
in the anaesthetized dog. Cardiovasc Res 15: 28-34
98. Melville KI, Blum B, Shister HE, Silver MD (1963) Cardiac ischemic changes and arrhythmias
induced by hypothalamic stimulation. Am J Cardiol 12: 781-791
99. Michell RH (1984) Polyphosphoinositide breakdown as the initiating reaction in receptor-
stimulated inositol phospholipid metabolism. Life Sci 32: 2083-2085
100. Miura Y, Inui J, Imamura H (1978) Alpha-adrenoceptor-mediated restoration of calcium
dependent potentials in the partially depolarized rabbit papillary muscle. Naunyn Schmiede-
bergs Arch Pharmacol 301: 201-205
101. Movsesian MA, Thomas AP, Selak M, Williamson JR (1985) Inositol trisphosphate does not
release Ca2+ from permeabilized cardiac myocytes and sarcoplasmic reticulum. FEBS Lett 185 :
328-332
102. Mugelli A, Cerbai E, Amerini S, Giotti A (1985) Altered responsiveness of cardiac alpha- and
beta-adrenoceptors during hypoxia and aging: relevance for arrhythmias. New Trends Ar-
rhythmias 1: 115-123
103. Mukherjee A, Bush LR, McCoy KE, Duke RJ, Hagler H, Buja LM, Willerson JT (1982)
Relationship between fJ-adrenergic receptor numbers and physiological responses during
experimental canine myocardial ischemia. Circ Res 50: 735-741
Sympathetic Influences on Arrhythmogenesis 95

104. Mukherjee A, Hogan M, McCoy K, Buja LM, Willerson JT (1980) Influence of experimental
myocardial ischemia on alphal-adrenergic receptors. Circulation [SuppI3] 64: 149
105. Muntz KH, Hagler HK, Boulas HJ, Willerson JT, Buja LM (1984) Redistribution of catechol-
amines in the ischemic zone of the dog heart. Am J Pathol114: 64-78
106. Muntz KH, Willerson JT, Buja LM (1983) Autoradiographic localization of alpha adrenergic
receptors on rat cardiac myocytes. Circulation [Suppl 3] 68: 59
107. Nazum FR, Bischoff F (1953) The urinary output of catechol derivatives including adrenaline
in normal individuals in essential hypertension and in myocardial infarction. Circulation 7:
96-101
108. Nishizuka Y (1984) The role of protein kinase C in cell surface signal transduction and tumor
promotion. Nature 308: 693-697
109. Nishizuka Y (1984) Turnover of inositol phospholipids and signal transduction. Science 225:
1365-1370
110. Nosek TM, Williams MF, Zeigler ST, Godt RE (1986) Inositol trisphosphate enhances calcium
release in skinned cardiac and skeletal muscle. Am J Physiol250: C807-C811
111. Otani H, Otani H, Das DK (1986) Evidence that phosphoinositide response is mediated by al-
adrenoceptor stimulation, but not linked with excitation-contraction coupling in cardiac muscle.
Biochem Biophys Res Commun 136: 863-869
112. Paintal AS (1972) Cardiovascular receptors. In: Neil E (ed) Handbook of sensory physiology,
vol 3. Enteroceptors. Springer, Berlin Heidelberg New York, pp 1-45
113. Pandey RC, Srivastava RD, Bhatnagar VM (1979) Effect of unilateral stellate ganglion blockade
and stimulation on experimental arrhythmias. Indian J Physiol Pharmacol23: 305-314
114. Parratt JR (1980) Beta-adrenoceptor blockade and early post-infarction dysrhythmias. In:
Burley DM, Birdwood GFB (eds) The clinical impact of beta-adrenoceptor blockade. Ciba
Laboratories, Horsham, Surrey, pp 29-49
115. Peterson DF, Brown AM (1971) Pressor reflexes produced by stimulation of afferent fibers in
the cardiac sympathetic nerves of the cat. Circ Res 28: 605-610
116. Podrid PJ (1984) Role of higher nervous activity in ventricular arrhythmia and sudden cardiac
death: implications for alternative antiarrhythmic therapy. Ann NY Acad Sci 432: 296-313
117. Poggioli J, Sulpice JC, Vassort G (1986) Inositol phosphate production following aI-adrenergic,
muscarinic or electrical stimulation in isolated rat heart. FEBS Lett 206: 292-298
118. Pogwizd SM, Corr PB (1987) Electrophysiologic mechanisms underlying arrhythmias due to
reperfusion of ischemic myocardium. Circulation 76: 404-426
119. Pogwizd SM, Corr PB (1987) Reentrant and nonreentrant mechanisms contribute to ar-
rhythmogenesis during early myocardial ischemia: Results using three-dimensional mapping.
Circ Res 61: 352-371
120. Pogwizd SM, Onufer JR, Kramer JB, Sobel BE, Corr PB (1986) Induction of delayed after-
depolarizations and triggered activity in canine Purkinje fibers by lysophosphoglycerides. Circ
Res 59:416-426
121. Reis DJ, Ross CA, Ruggiero DA, Granata AR, Joh TH (1984) Role of adrenaline neurons of
ventrolateral medulla (the Cl group) in the tonic and phasic control of arterial pressure. Clin
Exp Hypertens [A] 6:221-241
122. Richardson JA (1963) Circulating levels of catecholamines in acute myocardial infarction and
angina pectoris. Prog Cardiovasc Dis 6: 56-62
123. Rochette L, Didier JP, Moreau D, Bralet J (1980) Effect of substrate on release of myocardial
norepinephrine and ventricular arrhythmias following reperfusion of the ischaemic isolated
working rat heart. J Cardiovasc Pharmacol2: 267-279
124. Rogg H, Bucher UM (1979) Effects of an isolated myocardial ischaemia in dogs on plasma cat-
echolamines in the systemic and local coronary venous blood. Naunyn-Schmiedebergs Arch
Pharmacol307:R42
125. Ross CA, Ruggiero DA, Park DH, Joh TH, Sved AF, Fernandez-Pardal J, Saavedra JM, Reis
DJ (1984) Tonic vasomotor control by the rostral ventrolateral medulla: effect of electrical or
chemical stimulation of the area containing Cl adrenaline neurons on arterial pressure, heart
rate, and plasma catecholamines and vasopressin. J Neurosci 4: 474-494
126. Saper C, Loewy AD, Swanson LW, Cowan WM (1976) Direct hypothalamo-autonomic con-
nections. Brain Res 117: 305-312
127. Sato A, Schmidt RF (1973) Somatosympathetic reflexes: afferent fibers, central pathways, dis-
charge characteristics. Physiol Rev 53: 916-947
96 K.A. Yamada et al.

128. Schiimann HJ (1980) Are there alpha-adrenoceptors in the mammalian heart? Trends Phar-
macol Sci 1: 195-197
129. Schiimann HJ, Endoh M, Brodde OE (1975) The time course of the effects of the beta- and
alpha-adrenoceptor stimulation by isoprenaline and methoxamine on the contractile force and
cAMP level of the isolated rabbit papillary muscle. Naunyn Schmiedebergs Arch Pharmacol
289:291-302
130. Schwartz PJ, LaRovere MT, Zaza A, Pala M, Mazzoleni C, Specchia G (1985) Baroreceptor
reflexes and cardiac electrical instability in patients with a myocardial infarction. New Trends
Arrhythmias 1: 289-292
131. Schwartz PJ, Stone HL, Brown AM (1976) Effects of unilateral stellate ganglion blockade on
the arrhythmias associated with coronary occlusion. Am Heart J 92: 589-599
132. Schwartz PJ, Vanoli E, Zaza A, Zuanetti G (1985) The effect of antiarrhythmic drugs on life-
threatening arrhythmias induced by the interaction between acute myocardial ischemia and
sympathetic hyperactivity. Am Heart J 109: 937-948
133. Shahab L, Wollenberger A, Hause M, Schiller U (1969) Noradrenalineabgabe aus dem Hunde-
herzen nach voriibergehender Okklusion einer Koronararterie. Acta BioI Med Germ 22: 135-
143
134. Sharma AD, Saffitz JE, Lee BI, Sobel BE, Corr PB (1983) Alpha adrenergic-mediated ac-
cumulation of calcium in reperlused myocardium. J Clin Invest 72 : 802-818
135. Sheridan DJ, Penkoske PA, Sobel BE, Corr PB (1980) Alpha-adrenergic contributions to dys-
rhythmia during myocardial ischemia and reperlusion in cats. J Clin Invest 65: 161-171
136. Skinner JE, Lie IT, Entman ML (1975) Modification of ventricular fibrillation latency follow-
ing coronary artery occlusion in the conscious pig: The effects of psychological stress and beta-
adrenergic blockade. Circulation 51: 656-667
137. Skinner JE, Reed JC (1981) Blockade of frontocortical-brainstem pathway prevents ventricular
fibrillation of ischemic heart. Am J Physiol 240: H156-H163
138. Smith OA Jr, Nathan MA (1966) Inhibition of the carotid sinus reflex by stimulation of the
inferior olive. Science 154: 674-675
139. Staszewska-Barczak J, Ceremuzynski L (1968) The continuous estimation of catecholamine re-
lease in the early stages of myocardial infarction in the dog. Clin Sci 34:531-539
140. Stewart JR, Burmeister WE, Burmeister J, Lucchesi BR (1980) Electrophysiologic and anti-
arrhythmic effects of phentolamine in experimental coronary artery occlusion and reperlusion
in the dog. J Cardiovasc Pharmacol2: 77-91
141. Swanson LW (1982) Forebrain neural mechanisms involved in cardiovascular regulation. In:
Smith OA, Galosy RA, Weiss SM (eds) Circulation, neurobiology and behavior. Elsevier Bio-
medical, New York, pp 13-22
142. Tashiro N, Tanaka T, Fukumoto T, Hirata K, Nakao H (1985) Emotional behavior and ar-
rhythmias induced in cats by hypothalamic stimulation. Life Sci 36: 1087-1094
143. Thandroyen FT, Worthington MG, Higginson LM, Opie LH (1983) The effect of alpha- and
beta-adrenoceptor antagonist agents on reperlusion ventricular fibrillation and metabolic status
in the isolated perfused rat heart. J Am ColI Cardiol14: 1056-1066
144. Toda N, Shimamoto K (1968) The influence of sympathetic stimulation on transmembrane
potentials in the SA node. J Pharmacol Exp Ther 159: 298-305
145. Tsien RW, Carpenter DO (1978) Ionic mechanisms of pacemaker activity in cardiac Purkinje
fibers. Fed Proc 37:2127-2131
146. Uchida Y, Murao S (1973) Excitation of afferent cardiac sympathetic nerve fibers during coro-
nary occlusion. Am J Physiol 226: 1094-1099
147. Uchida Y, Murao S (1974) Afferent sympathetic nerve fibers originating in left atrial wall. Am
J Physiol 227: 753-758
148. Uchida Y, Murao S (1974) Bradykinin-induced excitation of afferent cardiac sympathetic nerve
fibers. Jpn Heart J 15: 84-91
149. Uchida Y, Murao S (1975) Acid-induced excitation of afferent cardiac sympathetic nerve
fibers. Am J PhysioI228:27-33
150. Ueda H, Uchida Y, Kamisaka K (1969) Distribution and responses of the cardiac sympathetic
receptors to mechanically induced circulatory changes. Jpn Heart J 10: 70-80
151. Vanoli E, Zaza A, Zuanetti G, Facchini M, Pippalettera M, Schwartz PJ (1982) Pharmacologic
prevention of malignant arrhythmias due to acute myocardial ischemia and sympathetic hyper-
activity. Circulation [SuppI2] 66:27
Sympathetic Influences on Arrhythmogenesis 97

152. Verrier RL, Lown B (1984) Behavioral stress and cardiac arrhythmias. Ann Rev Physiol46:
155-176
153. Verrier RL, Lown B (1982) Experimental studies of psychophysiological factors in sudden
cardiac death. Acta Med Scand [Suppl] 660:57-68
154. Verrier RL, Thompson PL, Lown B (1974) Ventricular vulnerability during sympathetic stimu-
lation: role of heart rate and blood pressure. Cardiovasc Res 8: 602-610
155. Vetter NJ, Strange RC, Adams W, Oliver MF (1974) Initial metabolic and hormonal response
to acute myocardial infarction. Lancet 1: 284-289
156. Videbaek J, Christensen NJ, Stemdorff B (1972) Serial determination of plasma catechol-
amines in myocardial infarction. Circulation 46: 846-855
157. Wallace AG, Sarnoff SJ (1964) Effects of sympathetic nerve stimulation on conduction in the
heart. Circ Res 14: 86-92
158. Webb SW, Adgey AA, Pantridge JF (1972) Autonomic disturbance at onset of acute myo-
cardial infarction. Br Med J 3 : 89-92
159. Wilber DJ, Lynch JJ, Pitt B, Lucchesi BR (1984) Protective effect of alpha! blockade in a con-
scious canine model of sudden death. Circulation [Suppl 2] 70: 179
160. Williams LT, Guerrero JL, Leinbach RC (1982) Prevention of reperfusion dysrhythmia by
selective coronary alpha-adrenergic blockade. Am J Cardiol 49: 1046
161. Wit AL, Rosen MR (1986) Afterdepolarizations and triggered activity. In: Fozzard HA,
Jennings RB, Haber E, Katz AM, Morgan HE (eds) The heart and cardiovascular system.
Raven, New York, pp 1449-1490
162. Wit AL, Cranefield PF (1977) Triggered and automatic activity in the canine coronary sinus.
Circ Res 41: 435-445
163. Witkowski FX, Corr PB (1984) An automated simultaneous transmural cardiac mapping sys-
tem. Am J Physiol247:H661-H668
164. Wollenberger A, Shahab L (1965) Anoxia-induced release of noradrenaline from the isolated
perfused heart. Nature 207: 88-89
165. Yamada KA, Lange LG, Sobel BE, Corr PB (1984) a-Adrenergic blockade ameliorates mito-
chondrial dysfunction associated with reperfusion of ischemic myocardium. Fed Proc 43: 812
166. Yamada KA, McAllen RM, Loewy AD (1984) GABA antagonists applied to the ventral sur-
face of the medulla oblongata block the baroreceptor reflex. Brain Res 297: 175-180
Chapter 7

Sympathetic Nervous System and Malignant Arrhythmias:

Evidence for Further Links
S.G. PRIORI and P. J. SCHWARTZ

Introduction

The importance of the autonomic nervous system in the genesis of cardiac arrhythmias
has become progressively more evident [4, 32]. Clinical and experimental observa-
tions have brought new insights into the relationship between cardiac arrhythmias
and the autonomic nervous system and have suggested new strategies for the preven-
tion of sudden cardiac death. Holter recordings of the minutes preceding the onset
of lethal arrhythmias, together with the results of clinical trials with antiadrenergic
interventions in postmyocardial infarction patients, have further established the rela-
tion between sympathetic activity and life-threatening ventricular arrhythmias.
Pantridge et al. [22] monitored the incidence of autonomic disturbances present
in the 1st h following the onset of acute myocardial infarction (MI) and demonstrated
signs of excessive sympathetic activity especially in an anterior wall myocardial in-
farction. Several investigators [9] demonstrated increases in the plasma norepineph-
rine (NE) level, persisting for at least several days after acute MI.
Adgey [1] observed significant increases in heart rate during the minutes preced-
ing ventricular fibrillation (VF) in patients with acute MI. Since the early 1970s, sev-
eral studies have evaluated the effect of P.adrenergic blocking drugs in the preven-
tion of sudden cardiac death among post-MI patients. Pooled data [7] indicated a
28% reduction in mortality with a 33% reduction in sudden cardiac death. In a recent
study, [34] a surgically selective antiadrenergic intervention (by high thoracic left
sympathectomy) dramatically reduced the incidence of sudden cardiac death in a
subgroup of high-risk post-MI patients.
These observations suggest that an augmented sympathetic activity is closely re-
lated to the development of ventricular tachyarrhythmias during an acute ischemic
episode. In this chapter, we will discuss the relationship between adrenergic activity
and life-threatening arrhythmias in the setting of acute myocardial ischemia.

Pathophysiology

Myocardial ischemia results in activation of cardiac vagal [37] and sympathetic affer-
ent fibers [20]. Several factors can contribute to this increased afferent neural activ-
ity. It has been suggested that local dyskinesia can activate mechanoreceptors located
in the ischemic area [2,20,38]; also chemical substances such as bradykinin [19] can
Sympathetic Nervous System and Malignant Arrhythmias 99

enhance their activity. The roles of serotonin [15], acidosis [40], prostaglandins [36]
and hyperkalemia [39] have also been suggested as responsible for afferent activation
mediated through chemoreceptors. This activation elicits an excitatory cardiocardiac
reflex [21] that takes place in the very early phase of ischemia and plays an important
role in the genesis of arrhythmias. Indeed, the interruption of the afferent limb of
this reflex greatly reduces the incidence of arrhythmias caused by a brief coronary
artery occlusion in cats [28, 35] . An abrupt increase in sympathetic activity acting on
an ischemic substrate can cause important changes in the coronary circulation as well
as in the electrophysiological properties of the myocardium, increasing the likeli-
hood of lethal arrhythmias.

Experimental Evidence for Adrenergically Mediated Arrhythmias

The considerations made above strongly support the evidence of a causal relation-
ship between neural sympathetic activation and arrhythmias ensuing in the very early
phase of an ischemic event. However, specifically designed experimental models are
necessary to support this concept and to enable an in-depth investigation of the
mechanisms involved in these arrhythmias.
In our laboratories, [29] an experimental model was developed in which ventricu-
lar arrhythmias, dependent on adrenergic hyperactivity superimposed on an episode
of transient ischemia, can be consistently reproduced (Fig. 1).
In a-chloralose-anesthetized cats, a 2-min occlusion of the anterior descending
branch of the left coronary artery is performed. Starting at the end of the 1st min of
ischemia, the left stellate ganglion is stimulated for 30 s. The interaction between

a b

-- .
c

Fig.!. Experimental model with examples of ECG changes. Letters indicate different phases of the
protocol during which ECG was recorded. Ventricular arrhythmias appear during left stellate gan-
glion stimulation, persist during the last part of occlusion, and ECG returns to control a few seconds
after release of coronary artery occlusion. LSG stirn, left stellate ganglion stimulation. (Schwartz
and Vanoli [29])
100 S. G. Priori and P. J. Schwartz

ischemia and sympathetic activation is arrhythmogenic in a high percentage of ani-

mals, while in most experiments neither ischemia per se nor left stellate stimulation
induce arrhythmias if performed independently.
In this experimental protocol, arrhythmias can be consistently induced in the
same animal for several trials. Indeed, when a specific type of arrhythmia is induced
in three consecutive trials, it will continue to develop in the next four or five trials.
This characteristic of the model confirms the close relation between ischemia and
sympathetic hyperactivity in the induction of malignant arrhythmias; furthermore, it
offers an ideal tool for studying pharmacological interventions in the prevention of
adrenergic arrhythmias with an internal control analysis.

Pharmacological Studies as a Tool to Speculate on Arrhythmogenesis

Several compounds from each of the four classes of antiarrhythmic agents, as de-
scribed by Vaughan Williams, [41] have been evaluated in the previously described
experimental preparation [24, 30, 31]. The drugs used were: lidocaine (2mg/kg +
4.3 mg/kg/h) , mexiletine (3.6mg/kg), and propafenone (2mg/kg + 2 mg/kg/h) , the
jJ-blocker propranolol (0.2mg/kg), the class III agent amiodarone (1.5 or 3 mg/kg),
and the calcium channel blockers nifedipine (0.02mg/kg), verapamil (0.2mg/kg),
and diltiazem (O.lmg/kg + 0.2mg/kg/h). In addition, two a-adrenergic blocking
agents were used: prazosin (0.1 mg/kg) and phenoxybenzamine (1 mg/kg).
The overall results showed that class I agents, namely lidocaine, mexiletine, and
propafenone, did not reduce the incidence of ventricular fibrillation nor that of ven-
tricular tachycardia. On the other hand, the antiadrenergic compounds propranolol
(Fig. 2), prazosin (Fig. 3), and phenoxybenzamine significantly reduced both the oc-
currence of ventricular fibrillation and ventricular tachycardia. However, the most
striking antiarrhythmic effects were observed after the administration of amiodarone,
verapamil, and diltiazem: all of these compounds completely prevented induction of
ventricular fibrillation and ventricular tachycardia. The other calcium antagonist
tested, nifedipine, had a protective effect in 10 of 13 cases (77%), comparable to that
of the antiadrenergic compounds.
These data have several implications. The results concur with the reduction of
mortality rate in post-MI patients obtained with jJ-adrenergic blocking agents. They
may contribute to explain the failure of the clinical trials with class I antiarrhythmic
drugs [8]. Therefore, while lidocaine-like compounds may be useful in controlling ar-
rhythmias in a later stage after MI, these drugs may lack a protective effect at the
onset of ischemia when the substrate for arrhythmias is different. These data suggest
that other groups of drugs such as calcium antagonists and amiodarone may be con-
sidered for the prevention of sudden death in coronary artery disease patients. From
these observations we may also derive an insight into the mechanisms for the genesis
of ventricular arrhythmias. We reached the unexpected conclusion that a "pure"
vasodilating agent, such as nifedipine, might be of help in the prevention of some of
the arrhythmic episodes. It is worth noting that nifedipine [24] was able to blunt the
reduction in ventricular fibrillation threshold that occurs when left stellate stimula-
tion is performed during a coronary artery occlusion. Such an effect was not present
during ischemia alone nor during sympathetic stimulation alone. It is therefore likely
Sympathetic Nervous System and Malignant Arrhythmias 101

":19
CONTROL PROPRANOLOL

Fig. 2. Effect of propranolol in 19 animals.

Each square represents the average result of
at least three trials in each cat. Each arrow
represents one experiment. Left squares
show results observed during three coronary
occlusions plus sympathetic stimulation per-
formed in control condition, i.e., before
drug administration. Right squares show re-
sults observed when three additional trials
were performed after drug administration.
It is evident that the injection of saline solu-
tion does not modify the response to coro-
nary occlusion plus sympathetic stimulation.
VF, ventricular fibrillation; VT, ventricular
tachycardia. Numbers in the squares indi-
cate number of premature ventricular con-
tractions. (Schwartz et al. [30])

CONTROL n =10 PRAZOSIN

B
G
11 50 I
1 -

B
Fig. 3. Effect of prazosin in ten animals. VF, ven-

8
tricular fibrillation; VT, ventricular tachycardia.
Numbers in the squares indicate number of prema-
ture ventricular contractions. (Schwartz et al. [30])
102 S. G. Priori and P. J. Schwartz

EFFECT OF COMBINED c:{- and B- BLOCKADE

"=5
<l or 6 <l + 6
Control
Blockade Blockade

~v;
B --VT--

111 - 50 1 111 - 50 1

Fig. 4. Effect of combined a- and fJ-

blockade in five animals. VF, ventricular
fibrillation; VT, ventricular tachycardia.
Numbers in the squares indicate number
of premature ventricular contractions.
(Schwartz et al. [30))

that the protective effect is largely due to the prevention of the further worsening of
ischemia induced by adrenergic activation.
The antiadrenergic compounds showed a fairly good protective effect in our model
even if, to afford complete protection, we had to administer both a- and ,B-blockers
(Fig.4). The previously demonstrated concept [5] that both a- and ,B-adrenergic
mechanisms contribute to arrhythmias associated with high sympathetic activity was
thus supported.
It seems indeed important to note that those compounds that provided the highest
degree of protection share both vasodilating and direct electrophysiological proper-
ties, suggesting that catecholamines are likely to affect both the electrophysiology of
the ischemic myocardium and the coronary blood flow.

Sympathetic Activation and Coronary Circulation

Coronary circulation has been considered for a long time to be a vascular bed mainly
regulated through local metabolic requirements. However, an important contribu-
tion of the adrenergic control has been suggested by several studies.
Feigl [6] demonstrated that an a-mediated sympathetic vasoconstriction can be
observed even when the metabolic demand of the myocardium is enhanced. Re-
cently, Heusch et al. [14] reported that the predominant vascular response to adren-
ergic activation in the vascular bed distal to a coronary stenosis results from the
interaction of the metabolic dilation and az-mediated adrenergic constriction.
Sympathetic Nervous System and Malignant Arrhythmias 103

700

650

LSG x

600

::r::
~
550
..,..
~
E Q Block
to
0- 500
Il
II::

C Fig. 5. Reactive hyperemia in conscious dogs

II
u
450 before and after left stellectomy (LSGx),
Ii phentolamine (a-block), and propranolol
A.
(f3-block). The data points represent the
means of 130 trials in 16 dogs and indicate
400
6 Block that left stellectomy and a-adrenergic block-
ade increase reactive hyperemia, a variable
that relates to the capability of the coronary
350 bed to dilate, whereas f3-adrenergic block-
ade has the opposite effect. The same kind
of responses were found when heart rate
was kept constant by pacing. (Schwartz and
300
control experimental Stone [33])

In conscious animals, Schwartz and Stone [33] demonstrated that a-adrenergic

activation exerts a tonic limitation on the vascular response to metabolic stimuli as
shown by an increase in the reactive hyperemia after left stellectomy or a-blockade
(Fig. 5).
Recent findings from Verrier [42] showed that sympathetic activation super-
imposed on a coronary stenosis induces a pronounced vasoconstriction at the end of
the stimulation of the cardiac nerves. This vasoconstriction is abolished by the a-
antagonist prazosin and by nifedipine [13]. His hypothesis [Verrier, personal com-
munication] is that the positive inotropic effect of the adrenergic activation and the
attendant increase in blood pressure counteracts the vasoconstriction during stimula-
tion of the left stellate ganglion, but when stimulation is terminated the effect on the
vascular bed is left unmasked. This important observation may lead to the concept
that on a functionally impaired myocardium, in which a high adrenergic activity is
unable to improve cardiac performance, the sudden onset of sympathetic discharge
may lead to a highly arrhythmogenic environment.

Sympathetic Activation and Myocardial Electrophysiology

Catecholamines elicit changes in electrophysiological parameters through the effect

on both a- and f3-adrenergic receptors.
104 S. G . Priori and P. J. Schwartz

•
C
• 0
• •
E

o
TO potential map 2' control occl. 4Omv[ E-+_ +
'---'
200ms

:t
3mm o -4 -8 -12 - 16 mV

TO potential map 2' occl. L. stellectomy

Fig. 6. Selected DC electrograms recorded 2 min after coronary occlusion from sites A to E before
and after left stellectomy. On the left, the area from which 60 electrograms were simultaneously
recorded is indicated, and iso-potentials during the TO segment are shown. The upper panel depicts
the situation during the control occlusion (second occlusion), and the lower panel shows the situation
2 min after a subsequent occlusion when the left stellate ganglion had been removed. Note that the
degree of ischemic changes is less marked after stellectomy. (Janse et al. [16])

In Purkinje cells fJ-stimulation induces a shortening of the action potential dura-

tion and an increase in spontaneous diastolic depolarization [11, 26]. It has been sug-
gested that a-adrenergic stimulation produces an inhibitory effect on the fJ-response,
as indicated by the augmentation of the fJ-mediated changes in the presence of a-
blockers [26]. The direct effect of a-agonists includes action potential lengthening
and a decrease in spontaneous diastolic depolarization and would therefore be con-
sidered nonarrhythmogenic. However, adrenergic stimulation during ischemia pro-
duces effects not present in normal conditions. It has been suggested that a differen-
tial activation of a- or fJ-receptors may be induced in the ischemic vs the normal
zone; this mechanism might induce an arrhythmogenic condition of electrical dis-
homogeneity in the myocardium. Han and Moe [12] showed that activation of the
cardiac sympathetic nerves increases the dispersion of ventricular refractory periods
favoring the development of reentrant circuits.
In a recent study by lanse et a1. , [16] DC extracellular electrograms were recorded
from 60 left ventricular epicardial sites during a brief coronary artery occlusion.
Changes in TO segment potential were considered markers of the degree of mem-
Sympathetic Nervous System and Malignant Arrhythmias 105

brane depolarization of the myocardium under the electrode. Left stellate stimula-
tion increased the degree of TQ segment depression, enlarged the area in which TQ
segment changes were observed, and resulted in ventricular tachyarrhythmias indi-
cating a sympathetically mediated worsening of the membrane depolarization (Fig. 6).
A rather unexpected finding of this study was the marked reduction of conduction
delay in the ischemic zone, despite the greater TQ segment depression. These data
can be explained by the previous observation that action potentials can be conducted
faster in this situation because of the reduced gap between resting membrane poten-
tial and excitation threshold [23]. However, this finding would make the genesis of
the reentrant circuits less likely to occur and would suggest automatic activity as a
more likely mechanism for adrenergic arrhythmias.
During myocardial ischemia, depolarization of the tissue may lead to inactivation
of the Na+ channels, increasing the role of the slow inward current in the maintenance
of depolarization and conduction. Catecholamines can increase the magnitude of the
slow inward current through ,B-adrenergic stimulation which augments the intracel-
lular cyclic adenosine monophosphate (cAMP) concentrations [18]. Recent work
also suggests that a-adrenergic mechanisms may playa role in increasing intracellular
calcium during reperfusion [27].
It has been suggested [4] that this accumulation of calcium might activate an in-
ward current that is operated by increased cytosolic calcium [3], thus favoring de-
polarization or development of delayed afterdepolarizations.
Kimura et al. [17] recently demonstrated that Purkinje fibers overlying infarcted
tissue present enhanced automaticity in response to a-adrenergic interventions and
showed triggered activity mediated both by a- and ,B-adrenergic effects.
Recently, using monophasic action potential recordings from the endocardium of
the cat, we showed [25] that left stellate ganglion stimulation induced delayed after-
depolarizations in vivo. This is the first demonstration of an existing link between
sympathetic activation and triggered activity, supporting the observations made dur-
ing in vitro experiments.

Conclnsions

The evidence for a close relationship between augmented sympathetic activity and
malignant arrhythmias is growing continuously.
Accordingly, experimental data indicate that sympathetic hyperactivity can favor
the onset of ventricular tachyarrhythmias by reentry, by enhanced automaticity, and
also by triggered activity.
Logical implications include that both prevention of ischemic episodes and car-
diac effects of sympathetic hyperactivity (by pharmacological blockade or by surgical
interruption of the neural pathways) can be expected to significantly reduce the oc-
currence of lethal arrhythmias.
106 S.G.Priori and P. J. Schwartz

References
1. Adgey AAJ, Webb SW (1979) The treatment of ventricular arrhythmias in acute myocardial
infarction. Br J Hosp Med 21 : 356
2. Casati R, Lombardi F, Malliani A (1979) Afferent sympathetic unmyelinated fibers with left
ventricular endings in cats. J Physiol (Lond) 292: 135-148
3. Colquhoun D, Neher E, Reuter H, Stevens CF (1981) Inward current channels activated by
intracellular Ca in cultured cardiac cells. Nature 294: 752-754
4. Corr PB, Yamada KA, Witkowski FX (1986) Mechanisms controlling cardiac autonomic func-
tion and their relation to arrhythmogenesis. In: Fozzard HA, Haber E, Jennings RB, Katz AM,
Morgan HE (eds) The heart and the cardiovascular system. Raven, New York, pp 1343-1404
5. Corr PB, Shayman JA, Kramer JB, Kipnis RJ (1981) Increased alpha-adrenergic receptors in
ischemic cat myocardium: a potential mediator of electrophysiological derangements. J Clin
Invest 67: 1232-1236
6. Feig! EO (1983) Coronary physiology. Physiol Rev 63: 1-205
7. Frishman WH, Furberg CD, Friedewald WT (1984) Beta-adrenergic blockade for survivors of
acute myocardial infarction. N Eng! J Med 310: 830-837
8. Furberg CD (1983) Effect of antiarrhythmic drugs on mortality after myocardial infarction. Am
J Cardiol 52: 32C-36C
9. Gazes PC, Richardson JA, Woods EF (1959) Plasma catecholamine concentrations in myo-
cardial infarction and angina pectoris. Circulation 19: 675-661
10. Goldstein DS (1981) Plasma norepinephrine as an indicator of sympathetic neural activity in
clinical cardiology. Am J Cardiol 48: 1147-1154
11. Giotti A, Ledda F, Mannaioni PF (1973) Effects of noradrenaline and isoprenaline in combina-
tion with alpha- and beta-receptor blocking substances, on the action potential of cardiac Pur-
kinje fibers. J Physiol (Lond) 229: 99-113
12. Han J, Moe GK (1964) Nonuniform recovery of excitability in ventricular muscle. Circ Res 14:
44-60
13. Hagestad EL, Verrier RL, Lown B (1985) Delayed myocardial ischemia following the cessation
of sympathetic stimulation. Fed Proc 45: 749
14. Heusch G, Deussen A, SchipkeJ, VogelsangH, Hoffmann V, ThamerV (1985) Role of cardiac
sympathetic nerves in the genesis of myocardial ischemia distal to coronary stenoses. J Cardio-
vasc Pharmacol [Suppl 5] 7: S13-18
15. James TN, Isobe JH, Urthaler F (1975) Analysis of components in a hypertensive cardiogenic
chemoreflex. Circulation 52: 179-192
16. Janse MJ, Schwartz PJ, Wilrns-Skopman F, Peters RJG, Durrer D (1985) Effects of unilateral
stellate ganglion stimulation and ablation on electrophysiologic changes induced by acute myo-
cardial ischemia in dogs. Circulation 72: 585-595
17. Kimura S, Basset AL, Kohya T, Kozloukis PL, Myerburg PJ (1987) Automaticity, triggered
activity and responses to adrenergic stimulation in cat subendocardiac Purkinje fibers after heal-
ing of myocardial infarction. Circulation 75: 651-660
18. Li T, Sperelakis N (1983) Stimulation of slow action potentials in guinea pig papillary muscle
cells by intracellular injection of cAMP, Gpp(NH)p, and cholera toxin. Circ Res 52: 111-117
19. Lombardi F, Patton CP, Della Bella P, Pagani M, Malliani A (1982) Cardiovascular and sym-
pathetic responses reflexly elicited through the excitation with bradykinin of sympathetic and
vagal cardiac sensory endings in the cat. Cardiovasc Res 16:57-65
20. Malliani A, Recordati G, Schwartz PJ (1973) Nervous activity of afferent cardiac sympathetic
fibers with atrial and ventricular endings. J PhysioI229:457-469
21. Malliani A, Schwartz PJ, Zanchetti A (1969) A sympathetic reflex elicited by experimental coro-
nary occlusion. Am J Physiol 217: 703-709
22. Pantridge JF (1978) Autonomic disturbances at the onset of acute myocardial infarction. In:
Schwartz PJ, Brown AM, Malliani A, Zanchetti A (eds) Neural mechanisms in cardiac ar-
rhythmias. Raven, New York, pp 7-17
23. Peon J, Ferrier GR, Moe GK (1978) The relationship of excitability to conduction velocity in
canine Purkinje tissue. Circ Res 43: 125-131
24. Priori SG, Zuanetti G, Schwartz PJ (1988) Ventricular fibrillation induced by the interaction be-
tween acute myocardial ischemia and sympathetic hyperactivity: effect of nifedipine. Am Heart
J 116:37-43
Sympathetic Nervous System and Malignant Arrhythmias 107

25. Priori SG, Mantica M, Schwartz PJ (1988) Delayed after depolarizations elicited in vivo by left
stellate ganglion stimulation. Circulation 78: 178-185
26. Rosen MR, Hordof AJ, Ilvento JP, Danilo P Jr (1977) Effects of adrenergic amines on electro-
physiological properties and automaticity of neonatal and adult canine Purkinje fibers. Evidence
for alpha- and beta-adrenergic actions. Circ Res 40: 390-400
27. Sharma AD, Saffitz JE, Lee BI, Corr PB (1983) Alpha-adrenergic mediated accumulation of
calcium in reperfused myocardium. J Clin Invest 72: 802-818
28. Schwartz PJ, Foreman RD, Stone HL, Brown AM (1976a) Effect of dorsal root section on the
arrhythmias associated with coronary occlusion. Am J Physiol231: 923-928
29. Schwartz PJ, Vanoli E (1981) Cardiac arrhythmias elicited by interaction between acute myo-
cardial ischemia and sympathetic hyperactivity: a new experimental model for the study of anti-
arrhythmic drugs. J Cardiovasc Pharmacol3: 1251-1259
30. Schwartz PJ, Vanoli E, Zaza A, Zuanetti G (1985) The effect of antiarrhythmic drugs on life-
threatening arrhythmias induced by the interaction between acute myocardial ischemia and sym-
pathetic hyperactivity. Am Heart J 109: 937-948
31. Schwartz PJ, Priori SG, Vanoli E, Zaza A, Zuanetti G (1986) Efficacy of diltiazem in two feline
experimental models of sudden cardiac death. J Am Coll Cardiol 8: 661-668
32. Schwartz PJ, Priori SG (1988) Adrenergic arrhythmogenesis and long QT syndrome. In: Vaughan
Williams EM, Campbell TJ (eds) Antiarrhythmic drugs. Handbook of experimental pharmacol-
ogy, vol 89. Springer, Berlin Heidelberg New York Tokyo, pp 519-543
33. Schwartz PJ, Stone HL (1977) Tonic influence of the sympathetic nervous system on myocardial
reactive hyperemia and on coronary blood flow distribution. Circ Res 41: 51-58
34. Schwartz PJ, Motolese M, Pollavin G, Malliani A, Bartorelli C, Zanchetti A and the Sudden
Death Italian Prevention Group (1985) Surgical and pharmacological antiadrenergic interventions
in the prevention of sudden death after a first myocardial infarction. Circulation 72 [Suppl 3] : 358
35. Schwartz PJ, Foreman RD, Stone HL, Brown AM (1976) Effect of dorsal root section on the
arrhythmias associated with coronary occlusion. Am J Physiol231: 923-928
36. Staszewska-Barczak J, Ferreira SH, Vane JK (1976) An excitatory nociceptive cardiac reflex eli-
cited by bradykinin and potentiated by prostanglandins and myocardial ischemia. Cardiovasc
Res 10: 314-327
37. Thoren PN (1976) Activation of left ventricular receptors with non-medullated vagal afferent
fibers during occlusion of a coronary artery in the cat. Am J Cardiol 37: 1046-1051
38. Uchida Y, Murao S (1973) Excitation of afferent cardiac sympathetic nerve fibers during coro-
nary occlusion. Am J Physiol 226 : 1094-1099
39. Uchida Y, Murao S (1973) Potassium-induced excitation of afferent cardiac sympathetic nerve
fibers. Am J Physiol 226: 603-607
40. Ueda H, Uchida Y, Kamisaka K (1969) Distribution and responses of the cardiac sympathetic
receptors to mechanically induced circulatory changes. Jpn Heart J 10: 70-80
41. Vaughan Williams EM (1970) Classification of antiarrhythmic drugs. In: Sand5e E, Flensted-
Jensen E, Olesen KH (eds) Symposium on cardiac arrhythmias. Astra, S5dertlilje, pp 449-472
42. Verrier RL, Hagestad EL, Lown B (1987) Delayed myocardial ischemia induced by anger.
Circulation 75: 249-254
Chapter 8
Modulation of Cardiac Arrhythmias
by the Autonomic Nervous System
R.F. GILMOUR JR., J.J.SALATA, and D.P. ZIPES

Introduction

Previous studies have shown that automaticity and conduction in various cardiac tis-
sues can be accelerated, decelerated, or suppressed completely (annihilated) by
properly timed short bursts of vagal discharges [1, 12-15, 18, 23, 26, 28-30, 33].
Phase-resetting and annihilation of automaticity and conduction also can occur fol-
lowing the delivery of brief hyperpolarizing or depolarizing current pulses [3, 7-11].
These observations suggest that cardiac arrhythmias caused by automaticity or by
conduction abnormalities may be modulated by phasic discharges of the autonomic
nervous system.
In recent experiments, we have investigated whether phase-resetting of automa-
ticity and conduction contributes to certain types of cardiac arrhythmias. These ar-
rhythmias include: sinus arrhythmia and A V block in intact dogs produced by phasic
stimulation of the vagus nerves during continuous background stimulation of the
stellate ganglia [23]; parasystole caused by abnormal automaticity in isolated prepa-
rations of human ventricular myocardium resected from patients having intractable
ventricular arrhythmias [7]; and ventricular tachycardia caused by reentry across iso-
lated canine Purkinje muscle junctions exposed to conditions similar to those present
during acute myocardial ischemia. In the following discussion, these studies are sum-
marized briefly and their possible contribution to our understanding of the relation-
ship between the autonomic nervous system and cardiac arrhythmias is explored.

Effects of Sympathetic Tone on the Phase-Dependent Chronotropic

and Dromotropic Responses to Brief Vagal Stimulation

Both the sympathetic and parasympathetic branches of the autonomic nervous sys-
tem importantly influence sinoatrial (SA) and atrioventricular (AV) node function,
thereby regulating heart rate and A V conduction. The rapid onset and offset of car-
diac responses to vagal stimulation allows for dynamic vagal modulation of heart rate
and A V conduction, whereas the slow temporal response to stellate stimulation pre-
cludes beat-to-beat regulation by sympathetic activity [28, 30]. Accordingly, the
chronotropic and dromotropic responses to vagal stimulation are markedly phase de-
pendent [1,12,13], while the responses to stellate stimulation are phase independent
[14, 28). Recently, Stuesse et al. [29] demonstrated that small changes in sympathetic
Modulation of Cardiac Arrhythmias 109

tone may alter the phasic effects of vagal stimulation on heart rate. The purpose of
our study was to determine whether a similar relationship between tonic stellate
stimulation and phasic vagal stimulation exists in the A V node.
Open-chest, anesthetized mongrel dogs were used for these studies. The afferent
connections of both stellate ganglia and both cervical vagi were isolated, doubly
ligated, and cut. Bipolar platinum or iridium wires, insulated except at the tip, were
used to stimulate the nerves. Stellate stimulation consisted of rectangular pulses 4 ms
in duration and 2mA in amplitude delivered at frequencies of I-4Hz. Vagal stimuli
were rectangular pulses 2 ms in duration with a current strength equal to 75% of the
current strength needed to produce complete A V block at a stimulus frequency of
10Hz. Brief bursts of stimuli (1-2 pulses) were delivered at a frequency of 100Hz.
Bipolar electrogram recordings were obtained from the high and low right atrium,
bundle of His, and right ventricle. For studies of A V conduction, SA node automa-
ticity was suppressed by infusion of verapamil (0.5-2.0 mg) into the SA node artery
[34], and the atria were paced at a constant cycle length of 300-600ms.
Figure lA shows a typical phase-response curve (PRC) for heart rate (A-A inter-
val) during vagal stimulation alone and in combination with tonic sympathetic stimu-
lation. The A-A interval is plotted on the ordinate as a function of the position, or
phase (0), of a single vagal stimulus delivered at different phases of the cardiac cycle.
During vagal stimulation alone, the A-A interval prolonged as a function of the vagal
stimulus phase until it reached a maximum (0 max), then shortened rather abruptly
to a minimum (0 min). The effects of vagal stimuli delivered during the later phases
of the cycle (the latent period) were postponed until the next cycle (not shown).
Tonic stellate stimulation at 1 and 2 Hz shortened the free-running cardiac cycle
length from a control value of 570ms to 465 and 400ms, respectively (compare val-
ues during latent period in Fig. lA). Stellate stimulation also shifted the PRC to the
left, i.e., to earlier phases, decreased the maximum prolongation of A-A interval by
vagal stimulation, and shortened the latent period. Similar, but less pronounced, ef-
fects of stellate stimulation were observed during repeated delivery of phase-coupled
vagal stimuli (Fig. lB).
Figure 2 illustrates the effects of stellate stimulation on the phase-response curve
for A V conduction (A-H interval) following the delivery of single vagal stimuli
(Fig.2A) and following repeated delivery of phase-coupled stimuli (Fig. 2B). Data
were obtained during atrial pacing at a constant cycle length of 500 ms. During vagal
stimulation alone, the A-H interval prolonged as a function of the cardiac phase until
A V block occurred (indicated by arrows). At the latermost phases, the effects of
vagal stimuli were postponed to the next cardiac cycle (not shown).
Tonic stellate stimulation at frequencies of 1,2, and 4Hz reduced the A-H inter-
val from a control value of 92ms to 82, 60, and 52ms, respectively (compare values
during latent period in Fig. 2A). Stellate stimulation shifted the PRC slightly to the
left and reduced the maximum prolongation of A-H interval by vagal stimulation.
Changes in the A-H interval were less marked following delivery of phase-coupled
vagal stimuli alone and in the presence of tonic sympathetic stimulation (Fig. 2B).
Figure 3 illustrates that the PRC for heart rate and A V conduction were phase-
shifted with one another, as expected from previous studies [26]. For a given cycle
length, the maximum prolongation of the A-H interval following a single vagal stim-
ulus (point D, Fig. 3) occurred at later phases than the maximum vagally induced
110 R. F. Gilmour Jr. et al.

A
,•
¢MAX

.
1000

..
.\ • V5 2pp Alone

•
.
A + I H, 55

.
900 ~
• + 2H, 5S

....... .."."
"\
800 " ""
700 ,," ~,,~ \'"

••
. cpMIN
600
• e"•••••
500

..•
'U 400 •••••
.§.
c B
I
•
•••
t •

..
C 1000
•• " A
•
" t
950
•• "
•
"
•A A
•
900 " A " •
••• • • ••
850 ••• • • • t
•
•
800
•
750
••• •• • ••
700
i
o 100 200 300 400 500 600
o (msec)
Fig. 1 A, B. Effects of stellate stimulation on the vagally induced PRC for cardiac cycle length (A-A
interval, ordinate). Vagal stimuli (VS) consisted of2 pulses (pp) each. A A-A intervals following the
delivery of single vagal stimuli at various phases of the cardiac cycle (0, abscissa) alone and in the
presence of tonic stellate stimulation at 1 or 2Hz (see upper right). Phases at which the maximum
and minimum effects occurred are indicated by @max and @min, respectively. B Steady-state A-A
intervals produced by vagal bursts delivered repetitively, coupled to each previous A wave. Arrows
indicate position of corresponding values of omax for single PRC from panel A. See text for dis-
cussion. (Salata et al. [23])

prolongation of A-A interval (point A). As shown in Figs. 1 and 2, both PRe were
shifted downward and leftward by tonic stellate stimulaton (+SS, Fig. 3). Thus, in
the absence of tonic stellate stimulation, a burst of vagal activity at a phase of 350 ms
would prolong cardiac cycle length (point A) which, in turn, would shorten the A-H
interval, despite a direct negative dromotropic effect of the vagus on A-H interval.
With the addition of tonic sympathetic tone, a vagal burst delivered at a phase of
Modulation of Cardiac Arrhythmias 111

A tt
225 • VS 2pp Alone • ti
6.,HzSS
•
•• . .. \ \
• • 2Hz SS
200
.
C·4HzSS

• i !;
175
\
.
150 •• ..
\
• \
•
125
• •
• .. i
!

.... \:

. • .... ..
i
100
.... •• ~
• • \ •
I -.• • • • • •• • .i.
75 .. .
:I: ••
.< 50 c c c c c c c c c
c c c c c c c !II •
c

150 B
Fig. 2 A, B. Effects of stellate stimulation on the
•••• vagally induced PRC for AV conduction (A-H
•
.. .. .. •
••
interval, ordinate). Format and abbreviations

• • .. .... .
125
are the same as for Fig. I. A A-H intervals fol-

.. •.. • •
lowing the delivery of single vagal stimuli at
100 • • .. ..
• • •...... various phases of the cardiac cycle (0, abscissa)
.... • • • • • • • • .. ..
alone in the presence of tonic stellate stimula-

75 •••••• ••••
tion at frequencies of 1, 2, or 4Hz (see upper
left). B Steady-state A-H intervals produced by
c c c c c c c c c c c c c c c c c c vagal bursts delivered repetitively, coupled to
each previous A wave. See text for discussion.
50
i i i i i
(Salata et al. [23])
0 100 200 300 400 500
o(msec)

350ms would prolong the A-H interval (point B), but have little effect on heart rate.
Thus, in the extreme case, it is possible that a small increase in sympathetic tone
could produce marked AV prolongation or block. Conversely, a vagal stimulus de-
livered at a phase of 400 ms during tonic stellate stimulation would have little effect
on the A-H interval (point C), yet the same stimulus would prolong the A-H interval
markedly following a decrease in sympathetic activity (near point D).
The examples of possible interactions between the phase-dependent chronotropic
and dromotropic effects of vagal stimulation and the phase-independent effects of
sympathetic stimulation given in Fig. 3 are only a sampling of the full range of poten-
tial interactions. It is possible that a more complete understanding of these inter-
actions could help to unravel some of the apparent complexities of atrial arrhythmias
and A V conduction disturbances, particularly those that elicit baroreceptor-
mediated responses from the autonomic nervous system.
112 R. F. Gilmour If. et al.

1100 225

o
1000
/\ 200

~(/ \
...
.t.
900
./11 175
i
! l>
I
150~
i
\\
.
125
600 !
!
500 i 100

400
\ _...............
...
75
i • •
400 500 600

Fig. 3. Postulated changes in cardiac cycle length (A-A interval, left ordinate, solid curves) and A V
node conduction (A-H interval, right ordinate, dotted curves) produced by small changes in sym-
pathetic tone coincident with brief vagal bursts delivered at vairous phases of the cardiac cycle (0,
abscissa). All values are in ms. Upper and lower curves represent PRe in the absence (-SS) or pres-
ence (+ SS) of stellate stimulation, respectively. The lower A V conduction PRe is a hypothetical
curve derived by combining the direct and indirect effects of stellate stimulation on the A-H interval.
The direct effect is a decrease in the maximum prolongation of the A-H interval (cf. Fig. 2). The
indirect effects, caused by a decrease in cardiac cycle length, are an increase in the maximum pro-
longation of A-H interval and a shift of the PRe to shorter cycle lengths. See text for discussion.
(Salata et al. [23])

Phase-Resetting of Automaticity in Diseased Human Ventricular Tissue

The results of several studies suggest that the phase-dependent chronotropic and
dromotropic effects of vagal stimulation described in the previous section derive
mainly from acetylcholine-induced hyperpolarization of cells within the SA and A V
nodes [8, 10, 12, 16,26]. Hyperpolarization of SA nodal cells increases the time re-
quired to depolarize from the maximum diastolic membrane potential to the thresh-
old potential for a spontaneous discharge. Consequently, the spontaneous discharge
rate is slowed as a function of the phase at which the vagal burst is introduced (Fig. 4)
[8,10,12,26]. In the A V node, moving the membrane potential further from thresh-
old may decrease excitability and conduction velocity [16]. The effects of acetyl-
choline-induced hyperpolarization on automaticity and conduction can be mimicked
by the delivery of brief hyperpolarizing current pulses [8, 10, 12]. In addition to slow-
ing the spontaneous discharge rate, a hyperpolarizing pulse delivered at a precise
time in diastole, known as the singular point [31], can annihilate automaticity [8].
The spontaneous discharge rate of a cardiac pacemarker also can be reset or an-
nihilated by brief subthreshold depolarizing current pulses [3, 7-10]. Pulses of "low"
intensity delay the next expected spontaneous discharge as a function of the phase at
Modulation of Cardiac Arrhythmias 113

..
.
.. ..
:: ."0
::41

o
...
---r~~~------------­

I ·'"
-20
c , ~O--~2~'5----~~O~~7~~--~100
Bel)
:' 50
; mV

Fig. 4A-D. Phase-dependent effects of brief pulses of acetylcholine (Ach) on a simulated sinus node
pacemaker. The basic cycle length (BeL) of the pacemaker was 727ms. Ach pulses were applied at
various phases (0) of the pacemaker cycle and produced phase shifts (Ll0). A An Ach pulse applied
early in the cycle (indicated by the arrow) accelerated the next spontaneous discharge by 12% (com-
pare solid line of actual discharge with dotted line of expected discharge in the absence of Ach).
B, eAch pulses applied progressively later in the cycle (arrows) delayed the next spontaneous dis-
charges by 13% and 62%, respectively. D Phase-response curve obtained by plotting Ll0 as a func-
tion of 0. See text for discussion. (Jalife and Michaels [9])

which they are introduced. The later the phase, the greater the delay. Pulses of
"high" intensity delivered during the first half of diastole delay the next expected dis-
charge, whereas pulses delivered during the latter half of diastole accelerate the next
discharge. Pacemaker delay probably is caused by depolarization-induced activation
of repolarizing potassium current and resetting of pacemaker current [3, 10]. Pace-
maker acceleration results from summation of the depolarizing current pulse and
spontaneous diastolic depolarization so that threshold potential is reached sooner [3,
10]. Depolarizing current pulses of the appropriate intensity delivered at the transi-
tion between pacemaker delay and pacemaker acceleration may annihilate spon-
taneous activity [7, 8].
Recently, we investigated the possibility that spontaneous activity in depolarized
human ventricular tissue [25, 27] may be reset or annihilated by subthreshold stimuli
[7]. Pieces of human ventricle were obtained following left ventricular aneurysmec-
tomy and endocardial resection in patients having intractable ventricular tachycar-
dia. The tissue was placed initially in cold oxygenated Tyrode solution and transfer-
red to a tissue bath where it was superfused with normal Tyrode solution at 37°C.
The preparations were allowed to beat spontaneously, and action potentials were re-
corded using standard microelectrode techniques. To reset the spontaneous dis-
charge rate, depolarizing current pulses of 1O-50ms duration and 50-400nA were
injected via the recording rnicroelectrode, using a rapid switching system.
114 R.F. Gilmour Jr. et al.

A
0-

0-
Fig. 5 A, B. Modulation of pacemaker activity in dis-
eased human ventricle by subthreshold current pulses.
A Two recording sites along the same trabeculum in a
spontaneously active preparation. Current pulses of
B 1740
30ms duration were injected through the lower micro-
electrode (filled circles) and produced a subthresh-
0-
old depolarization, as recorded by the upper micro-
electrode. The interval between spontaneous action
0- potentials is given in milliseconds above each interval.
B Same cells and format as panel A. Vertical calibra-
tion, 50mV; horizontal calibration, 2s. See text for
discussion. (Gilmour Jr. et al. [7])

Figure 5 illustrates phase-resetting in one of the preparations. A depolarizing cur-

rent pulse delivered early in diastole (680 ms after the upstroke of the preceding
spontaneous discharge) delayed the next spontaneous discharge by 400ms (Fig. 5A).
A current pulse of the same intensity and duration delivered later in diastole (950 ms
after the preceding discharge) accelerated the next spontaneous discharge by 210 ms,
relative to the previous spontaneous cycle length (Fig. 5A). A current pulse delivered
930ms after the preceding spontaneous discharge annihilated automaticity (Fig. 5B).
Experiments such as those illustrated in Fig. 4 have provided a cellular electro-
physiological basis for the phasic modulation of SA node automaticity by the vagus.
In addition, they have led to speculation that the spontaneous discharge rates of sub-
sidiary cardiac pacemakers and ectopic foci may be modulated by phasic discharges
of the vagus and by electrotonic interactions between pacemakers [9, 11, 19]. Such
interactions may pertain to the behavior of certain cardiac arrhythmias, particularly
parasystole [2, 9, 11, 19,21].

Phase-Resetting of Circus Movement Reentry

in Isolated Canine Purkinje-Muscie Preparations

It has been suggested that sustained reentry represents a type of biological oscillator
and, as such, may be subject to phase-resetting [32]. However, the functional basis
for this supposition has not been explored in detail. One impediment to testing this
hypothesis has been the lack of a suitable model of reentry in which to examine the
effects of low-intensity perturbations. Therefore, the purpose of our study was to in-
corporate previous knowledge concerning the Purkinje-muscle junction into a work-
ing model of circus movement reentry and to attempt to phase-reset the reentry
using subthreshold electrical stimuli delivered to the Purkinje-muscle junction.
Figure 6 illustrates the basic premise of the model. A premature impulse blocks
at a Purkinje-muscle junction whose refractory period has been extended by expo-
sure to some of the metabolic alterations known to occur during acute myocardial
 .
Modulation of Cardiac Arrhythmias 115

)(

Fig. 6. Schematic diagram of postulated reentry

circuit. The upper branches of the figure represent
Purkinje tissue and the lower portions represent
papillary muscle. Purkinje-muscle junctions are
shown between points 1 and 2 and between points
3 and 4. Stimuli are delivered at point X. The Pur-
kinje-muscle junction between points 1 and 2 has
been depressed by exposure to a combination of
hyperkalemia, hypoxia, and acidosis . See text for
discussion

ischemia (hypoxia, hyperkalemia, and acidosis). The impulse conducts across a nor-
mal Purkinje-muscle junction located in a different part of the preparation and ac-
tivates ventricular muscle. Conduction continues through muscle and retrogradely
across the previously blocked Purkinje-muscle junction. Despite the presence of
anterograde block at the "depressed" Purkinje-muscle junction, retrograde conduc-
tion across the junction is expected because of the natural anisotropy present at Pur-
kinje-muscle junctions [17, 20, 24], particularly under these experimental conditions
[4-6,22]. The electrical isolation between Purkinje tissue and ventricular muscle, ex-
cept at Purkinje-muscle junctions [17,20], prevents the impulse from short-circuiting.
The reentry circuit presented in Fig. 6 is redrawn in the inset of Fig. 7 as a sym-
metrical circuit, with different symbols corresponding to different regions of the cir-
cuit. In the remainder of the figure, the distance traveled around the circuit during
reentry is plotted as a function of time, using data from preliminary experiments.
Initially, the reentrant wavefront conducts retrogradely across the depressed Pur-
kinje-muscle junction. The distance traveled is small (1 mm) and the time required is
long (120ms). Conduction through adjacent partially refractory Purkinje cells also is
slow (0.035 m/s). Once the wavefront reaches normal Purkinje tissue, conduction ac-
celerates to 2.0m/s. On the return route, the wavefront traverses the normal Pur-
kinje-muscle junction and ventricular muscle at normal speed (0.4 m/s) , but slows (to
0.04 m/s) as it reaches partially refractory muscle cells near the "depressed" Purkinje-
muscle junction.
As represented in Fig. 7, the distance/time plot of the reentry circuit resembles
the voltage/time plot characteristic of a cardiac pacemaking cell (cf. Figs. 4 and 5).
This similarity suggested to us that subthreshold depolarizing or hyperpolarizing
stimuli delivered to the depressed Purkinje-muscle junction or to surrounding Pur-
kinje cells might reset the reentry cycle, just as subthreshold stimuli delivered during
116 R. F. Gilmour Jr. et al.

14
A
12 "PMJ n

10
E
E
- 8
UJ
U
Z
<l 6
I-
r/I
C

o _
..............-.............
.........- - - - - '
i I I I I I I I I
20 60 80 100 120 1 ~0 16 0 180
TI ME (msec )

Fig. 7. Relationship between time (abscissa) and distance (ordinate) traveled in a postulated reentry
circuit (inset). Different symbols on the reentry circuit diagram correspond to symbols on the time-
distance plot. The reentry circuit is similar to that shown in Fig. 6. Reentry occurs when a premature
impulse blocks at the depressed Purkinje-muscle junction (PMJd), conducts from normal Purkinje
tissue (PFn) across the normal Purkinje-muscle junction (PMJn), into normal papillary muscle
(PMn) , retrogradely into partially refractory papillary muscle (PMr) proximal to PMJd, across
PMJd to partially refractory Purkinje tissue (PFr) distal to the junction. See text for discussion

diastolic depolarization reset the spontaneous discharges of a pacemaker cell. Obvi-

ously, the mechanism by which the stimuli might reset reentry would be different
from the mechanism by which the stimuli reset automaticity. Nevertheless, the out-
come, i.e., acceleration, deceleration, or annihilation of oscillatory activity, would
be similar.
To test this hypothesis, canine Purkinje muscle preparations containing two Pur-
kinje-muscle junctions were mounted in a partitioned tissue bath. One of the Pur-
kinje-muscle junctions was superfused with normal Tyrode solution and the other
Purkinje-muscle junction was superfused with hyperkalemic Tyrode solution (KCL
= 8 mM) . The locations of the Purkinje-muscle junctions were identified by mapping
the activation sequence of the preparation using microelectrodes. After mapping had
been performed, continuous simultaneous recordings were obtained from normal
Purkinje tissue near the stimulating electrode, Purkinje and muscle cells near the
normal Purkinje-muscle junction, and Purkinje and muscle cells near the depressed
Purkinje-muscle junction (Fig. 8).
Under these conditions, a premature stimulus delivered to the normal Purkinje
fiber blocked at the depressed Purkinje-muscle junction, conducted across the nor-
mal Purkinje-muscle junction, through papillary muscle, and retrogradely across the
depressed Purkinje-muscle junction. The impulse continued to conduct through nor-
mal Purkinje tissue before blocking finally at the normal Purkinje-muscle junction.
Stable single reentrant cycles of this type were obtained for 2-3 h.
In the preliminary experiments performed thus far , injection of brief (10-30 ms
duration) hyperpolarizing current pulses (50-200 nA) into Purkinje cells near the de-
pressed Purkinje-muscle junction during the first 30% of the reentry cycle produced
Modulation of Cardiac Arrhythmias 117

:,-~
~,- ~
P F n 2\ . ~

100 mSeC

Fig. 8. Annihilation of circus movement reentry in an isolated canine Purkinje-muscle preparation.

Two cycles of reentry, one before and one after delivery of a hyperpolarizing current pulse, are
superimposed. Action potentials were recorded from regions similar to those given in Fig. 7, i.e.,
from the Purkinje fiber (PF), from Purkinje and papillary muscle cells near the normal Purkinje-
muscle junction (PFn and PMn, respectively), and from Purkinje and papillary muscle cells near the
depressed Purkinje-muscle junction (PFr and PMr, respectively). Following the last stimulus of a
train of ten stimuli delivered at a basic cycle length of 300ms (action potentials to the left), a single
premature stimulus (S2) was delivered to PF at a coupling interval of 166 ms. The order of activation
following delivery of S2 is given by numbers 1-8. The premature impulse conducted from PF (1) to
Pfn (2) and PFr (3), conducted across the normal Purkinje-muscle junction (from 2 to 4), blocked at
the depressed Purkinje-muscle junction (between 3 and 5), conducted retrogradely through papillary
muscle (from 4 to 5), retrogradely across the depressed Purkinje-muscle junction (from 5 to 6) and
into PF (7), before blocking in Purkinje tissue proximal to the normal Purkinje-muscle junction (8).
Delivery of a hyperpolarizing current pulse (10 ms, 90 nA) to PFr 94 ms after S2 (indicated by the
asterisk) resulted in more rapid activation of PFr (indicated by the arrow) and failure of the impulse
to propagate to PF or to PFn. See text for further discussion

a post-pulse increase in Purkinje fiber action potential duration and refractory

period, thereby delaying retrograde activation by 7-22 ms. Current injection during
the latter 30% of the reentry cycle shortened Purkinje fiber action potential duration
and accelerated retrograde activation by 5-20ms. Current injection at 40%-60% of
the reentry cycle suppressed reentry either by increasing Purkinje fiber action poten-
tial duration to such an extent that retrograde conduction block occurred or by short-
ening Purkinje fiber action potential duration and decreasing retrograde conduction
delay such that the retrograde impulse blocked distal to the junction in Purkinje tis-
sue having a longer refractory period. An example of the later phenomenon is shown
in Fig. 8.
These studies suggest that circus movement reentry may be accelerated, deceler-
ated, or annihilated by properly timed subthreshold electrical stimuli. They also raise
the possibility that reentry may be reset or suppressed by brief hyperpolarizations
secondary to phasic discharges of parasympathetic nerves, provided these nerves are
intact and functioning near a region of slow conduction that is required for mainte-
nance of the reentry cycle.
118 R. F. Gilmour Jr. et al.

Acknowledgments. We thank Ruth Gerbig for technical assistance and Sue Hennigar for preparing
the manuscript.
Supported in part by the Herman C. Krannert Fund, Indianapolis, IN, by Grants HL-06367,
HL-06308, and HL-07182 from the National Heart, Lung and Blood Institute of the National Insti-
tutes of Health, Bethesda, MD, by the Attorney General of Indiana Public Health Trust, and by the
Roudebush Veterans Administration Medical Center, Indianapolis, IN.

References

1. Brown G, Eccles J (1934) The action of a single vagal volley on the rhythm of the heartbeat.
J Physiol (Lond) 82:211-241
2. Castellanos A, Luceri RM, Moleiro F, Kayden DS, Trohman RG, Zaman L, Myerburg RJ
(1984) Annihilation, entrainment and modulation of ventricular parasystolic rhythms. Am J
CardioI54:317-322
3. Clay JR, Guevara MR, Shrier A (1984) Phase-resetting of the rhythmic activity of embryonic
heart cell aggregates. Experiment and theory. Biophys J 45: 699-714
4. Evans JJ, Gilmour RF Jr, Zipes DP (1984) The effects of lidocaine and quinidine on impulse
propagation across the canine Purkinje-muscle junction during combined hyperkalemia, hypoxia
and acidosis. Circ Res 55: 185-196
5. Gilmour RF Jr, Evans JJ, Zipes DP (1984) Purkinje-muscle coupling and endocardial response
to hyperkalemia, hypoxia and acidosis. Am J PhysioI247:H303-H311
6. Gilmour RF Jr, Evans JJ, Zipes DP (1985) Preferential interruption of impulse transmission
across Purkinje-muscle junctions by interventions that depress conduction. In: Zipes DP, Jalife
J (eds) Cardiac electrophysiology and arrhythmias. Grone and Stratton, New York, pp 287-300
7. Gilmour RF Jr, Heger JJ, Prystowsky EN, Zipes DP (1983) Cellular electrophysiologic abnor-
malities of diseased human ventricular myocardium. Am J Cardiol51: 137-144
8. Jalife J, Antzelevitch C (1979) Phase resetting and annihilation of pacemaker activity in cardiac
tissue. Science 206: 695-697
9. Jalife J, Michaels DC (1985) Phase-dependent interactions of cardiac pacemakers as mecha-
nisms of control and synchronization in the heart. In: Zipes DP, Jalife J (eds) Cardiac electro-
physiology and arrhythmias. Grone and Stratton, New York, pp 109-119
10. Jalife J, Moe GK (1976) Effect of electrotonic potentials on pacemaker activity of canine Pur-
kinje fibers in relation to parasystole. Circ Res 39: 801-808
11. Jalife J, Moe GK (1979) A biologic model of parasystole. Am J CardioI43:761-772
12. Jalife J, Moe GK (1979) Phasic effects of vagal stimulation on pacemaker activity of the isolated
sinus node of the young cat. Circ Res 45: 595-607
13. Levy MN, Martin PJ, lano T, Zieske H (1970) Effects of single vagal stimuli on heart rate and
atrioventricular conduction. Am J Physiol 218: 1256-1262
14. Levy MN, Zeiske H (1969) Autonomic control of cardiac pacemaker activity and atrioventricular
transmission. J Appl Physiol 27: 465-470
15. Martin PJ (1977) Dynamic control of atrioventricular conduction: Theoretical and experimental
studies. Ann Biomed Eng 3: 275-295
16. Mendez C (1980) The slow inward current and A V nodal propagation. In: Zipes DP, Bailey JC,
Elharrar V (eds) The slow inward current and cardiac arrhythmias. Martinus Nijhoff, The
Hague, pp 285-294
17. Mendez C, Mueller WJ, Urguiaga X (1970) Propagation of impulses across the Purkinje fiber-
muscle junctions in the dog heart. Circ Res 25: 135-150
18. Michaels DC, Slenter VAJ, Salata JJ, Jalife J (1983) A model of vagus-sinoatrial node inter-
actions. Am J Physiol 245 : H1043-H1053
19. Moe GK, Jalife J, Mueller WJ, Moe B (1977) A mathematical model of parasystole and its
application to clinical arrhythmias. Circulation 56: 968-979
20. Myerburg RJ, Nilsson K, Gelband H (1972) Physiology of canine intraventricular conduction
and endocardial excitation. Circ Res 30:217-243
21. Nau GT, Aldariz AE, Acunzo RS, Halpern MS, Davidenko JM, Elizari MV, Rosenbaum MB
(1982) Modulation of parasystolic activity by non-parasystolic beats. Circulation 66: 462-469
Modulation of Cardiac Arrhythmias 119

22. Overholt ED, Joyner RW, Veenstra RD, Rawling D, Weidmann R (1984) Unidirectional block
between Purkinje and ventricular layers of papillary muscles. Am J Physiol 247: H584-H595
23. Salata JJ, Gill RM, Gilmour RF Jr, Zipes DP (1986) Effects of sympathetic tone on vagally in-
duced phasic changes in heart rate and atrioventricular node conduction in the anesthetized dog.
Circ Res 58: 584-594
24. Sasyniuk BI, Mendez C (1971) A mechanism for reentry in canine ventricular tissue. Circ Res
28:3-15
25. Singer DH, Baumgarten CM, Ten Eick RE (1981) Cellular electrophysiology of ventricular and
other dysrhythmias: studies on diseased and ischemic heart. Prog Cardiovasc Dis 24:97-156
26. Slenter VAJ, Salata JJ, Jalife J (1984) Vagal control of pacemaker periodicity and intranodal
conduction in the rabbit sinoatrial node. Circ Res 54: 436-446
27. Spear JR, Horowitz LN, Hodess AB, MacVaugh H III, Moore EN (1979) Cellular electro-
physiology of human myocardial infarction. I. Abnormalities of cellular activation. Circulation
59:247-256
28. Spear JF, Moore EN (1973) Influence of brief vagal and stellate nerve stimulation on pacemaker
activity and conduction within the atrioventricular conduction system of the dog. Circ Res 32:
27-41
29. Stuesse SL, Wallick DW, Zieske H, Levy MN (1981) Changes in vagal phasic chronotropic
responses with sympathetic stimulation in the dog. Am J Physiol 241 : H850-H856
30. Warner HR, Russell RO Jr (1969) Effect of combined vagal and sympathetic stimulation on
heart rate in the dog. Circ Res 24:567-573
31. Winfree AT (1970) Integrated view of resetting a circadian clock. J Theor Bioi 28:327-374
32. Winfree AT (1983) Sudden cardiac death: a problem in topology? Sci Am 248: 144-161
33. Yang T, Levy MN (1984) The phase-dependency ofthe cardiac chronotropic responses to vagal
stimulation as a factor in sympathetic-vagal interactions. Circ Res 54: 703-710
34. Zipes DP, Fischer JC (1974) Effects of agents which inhibit the slow channel on sinus node auto-
maticity and atrioventricular conduction in the dog. Circ Res 34: 184-192
Chapter 9

Supraventricular Tachycardia
and the Autonomic Nervous System
D.L.Ross

Introduction

Clinical and experimental observations of termination of supraventricular tachy-

cardia by various methods of increasing vagal tone [11, 12, 26], and facilitation of
tachycardia at times of increased sympathetic tone such as exercise and emotion,
have pointed to the major role played by the autonomic nervous system in the induc-
tion and maintenance of supraventricular tachycardia. Further observations of the
onset and first few minutes of tachycardia induced in an electrophysiological study
have shown significant changes in hemodynamics [6, 19, 20], atrioventricular (A V)
nodal conduction [3, 4, 15, 28], and propensity to spontaneous termination of the
tachycardia due to block in one of the components of the reentrant circuit [3, 4, 15,
28]. These observations are consistent with several phases in the induction and main-
tenance of tachycardia:

Induction of the First Cycle of Reentry. The three conditions for reentry must be
satisfied for this to occur, namely, a potential reentrant circuit, one-way block in one
arm of the circuit, and slow conduction through the circuit to allow previously de-
polarized tissues to recover excitability.

The Rate-Dependent Phase. This lasts approximately the first 5 s of tachycardia. The
abrupt increase in rate causes rate-dependent increases in A V nodal refractoriness
and conduction time and rate-dependent decreases in atrial, ventricular, and acces-
sory pathway refractory periods. These rate-dependent changes in refractoriness
take several beats to plateau out and occur to the greatest degree in the most
specialized tissues (the A V node and the His-Purkinje system).
During this phase, the tachycardia is prone to spontaneous termination whenever
the reentrant impulse encounters an area still refractory from the last cycle. Varia-
tions in A V nodal, His-Purkinje system, and accessory pathway conduction times
often occur in this phase causing variable and often alternating short and long cycle
lengths [15]. Brief burst of direct vagal stimulation has been shown to cause effects
on A V nodal conduction in dogs with onset in 200-500 ms, peak effect in 400 ms, and
return to baseline in 1-2s [13, 23]. In man, direct 15- to 90-s stimulation of the
carotid sinus nerves caused similar time responses with onset in 0.8-1.6s and peak
effect in 5 s [1]. The time for the full reflex arc from baroreceptor stimulation to
onset of vagally induced change in A V nodal conduction will therefore take longer.
This interval is approximately 1.5-3 s based on responses to carotid sinus massage
Autonomic Nervous System in SVT 121

[26]. Thus, withdrawal of vagal tone during the first seconds of tachycardia due to
the precipitous fall in arterial pressure (when the rate is over 150 beats/min) takes a
few seconds to begin and about 5 s to peak effect. Therefore, significant changes in
A V nodal conduction time are not usually observed until after 5 s because of the
delayed onset of decreased vagal activity and the counterbalancing effect of rate-
dependent changes in the AV node. Curry [4] noted that onset of changes in AV
nodal conduction time took approximately 5-15 beats of tachycardia (representing
the first 2-5 s of tachycardia).

The Autonomic Phase. As explained above, withdrawal of vagal tone becomes signif-
icant after approximately 5 s of tachycardia. Brief direct stimulation of the stellate
ganglion in dogs causes changes in AV nodal conduction with onset in 1 s, peak in
7.5 s, and disappearance in 20-25 s [23]. Borst et a1. [1] showed that stimulation of
the carotid sinus nerve in man after atropine caused reflex withdrawal of sympathetic
tone with onset in 2-3 s and maximum in 20 s. This is an estimate ofthe time taken
for the full sympathetic reflex arc from afferent stimulation to effect. Arterial pres-
sure falls to a nadir in the first 3-5 s of tachycardia and then rises and sometimes ex-
ceeds the arterial pressure present before onset of tachycardia. The arterial pressure
in tachycardia then generally stabilizes after 15-45 s. The phase of increasing blood
pressure and overshoot prior to steady state causes increased vagal tone. Thus, the
first 15 s of the autonomic phase are associated with withdrawal of vagal tone and
augmentation of sympathetic tone. The next 15-30 s are associated with an increase
in vagal tone prior to achievement of steady-state conditions.

Stable Sustained Tachycardia. Steady state in autonomic balance is present in this

phase and the tachycardia will continue indefinitely until interventions alter auto-
nomic tone or the circuit is perturbed by ectopic beats or sudden failure in conduc-
tion (which may occur in both the A V node and accessory pathway).

Observations During Spontaneous Termination

of Reentrant Supraventricular Tachycardia

The A V node is a component of the reentrant circuit in both circus movement tachy-
cardias using an accessory pathway and AV junctional (AV nodal) reentry. These
two types of supraventricular tachycardia account for 90% of cases. Thus, both
should be sensitive to changes in autonomic tone. Study of the time pattern of spon-
taneous termination of tachycardia after induction shows a characteristic pattern in
tachycardia using an accessory pathway (Fig. 1) [15]. Termination due to block in
any of the three major components of the circuit (AV node, His-Purkinje system,
and accessory pathway) may occur in the rate-dependent phase, depending on which
has the longest refractory period. In the first half of the autonomic phase, the A V
node and His-Purkinje system disappear as causes for spontaneous termination due
to the facilitation of A V nodal conduction by withdrawn vagal tone and increased
sympathetic tone, and the ability of the His-Purkinje system to markedly shorten its
refractory period in response to increase in rate. The AV node returns as a site of
122 D.L.Ross

No. of spontaneous
terminations

20
HPS
block 10

30

AVN 20
block
10

Fig. 1. Time course prior to spontaneous

20 termination of circus movement tachy-
cardia (CMT) by block in the
AP
block 10 His-Purkinje system (HPS), A V node
(AVN), or accessory pathway (AP) .
(Ross et al. [15])
5 10 15 20 25 30 >30 sec.
Duration of CMT

Fig. 2. Spontaneous termination of circus movement tachycardia using a left free wall accessory path-
way by block in the A V node following a short cycle during cycle length alternans. Variation in A V
nodal conduction times (AH intervals) caused the alternating cycle lengths. Surface ECG leads,
I, II, III, Vb V6 ; HRA, high right atrial electrogram; CS, coronary sinus proximal (p) and distal (d)
electrograms; His, His bundle electrogram; A, atrial deflection; H, His bundle deflection; paper
speed, l00mm/s. (Ross et al. [15])
Autonomic Nervous System in SVT 123

block in 30-60 s due to the later phase of increased vagal tone secondary to blood
pressure overshoot. Waxman et al. [28] found similar results with a mean ± SEM of
28 ± 5 s until spontaneous termination in the A V node during this phase of increased
vagal tone. The accessory pathway may block in any of the phases of tachycardia, but
does so predominantly in the rate-dependent and sympathetic autonomic phases. In-
creased sympathetic tone causes shortening of the A V nodal conduction time and
therefore accelerates the tachycardia. This may then lead to rate-dependent block in
the accessory pathway [4, 15].
Spontaneous termination of supraventricular tachycardia is common and occurs
in 40% -45% of patients under electrophysiological study [15, 28]. The reason for the
high incidence of this phenomenon is that the longest refractory period in the
tachycardia circuit is not much shorter than the cycle length. This margin of safety is
only 20 ms for those patients with spontaneous termination and 48 ms for those with-
out [15]. Thus, the margin of safety ranges from only 7% to 16% of tachycardia cycle
length. This delicate equilibrium makes supraventricular tachycardias prone to ter-
minate after perturbation. Frequently, oscillations in tachycardia cycle length pre-
cede spontaneous termination (Fig. 2). This observation led on to analysis of the fac-
tors responsible for the stability of sustained tachycardia [16, 22].

Determinants of Stability of Reentrant Supraventricular Tachycardia

The most labile component of the tachycardia circuit is the A V node. As the simplest
approximation, conduction time around the remainder of the circuit can be regarded
as a constant in both accessory pathway and A V junctional (AV nodal) reentrant
tachycardias. A V nodal function can be characterized by measurement of the A V
nodal conduction times of increasingly premature atrial beats, each delivered after a
drive chain of eight atrial paced beats (e.g., Fig. 3). Using these data a computer

• Basic cycle length = 600 msec

•
200 •
••
t
AVN ERP
• •••
•••
•••• • • •
• •
100

200 300 400 500

Fig. 3. Representative example of a normal human A V nodal function curve in an electrophysiological

study. A 2H 2, AV nodal conduction time of induced premature atrial beats; A 1A 2, coupling interval
of the induced atrial premature beat; A VN ERP, A V nodal effective refractory period. Note the
gradual increase in A V nodal conduction time with prematurity and that the maximum slope of
curve in this example is greater than 45°
124 D.L.Ross

A B

A
t
40
I
AH1 AH1
24
AH3
.... ......
30~""""''''' 30 0

....
..............

• 40·
AA1 AA AA1 AA2 AA

C 0

~~~----------~
....
30 0

12
...................

AA AAtAA4 AA

AA
E

Beats of
tachycardia
Fig.4A-E. Effects of perturbation of a computer model circus movement tachycardia if the AV
nodal function curve is a straight line with slope of 30° (Le., flat). AIHb AV nodal conduction time
of a tachycardia of cycle length AAJ prior to perturbation. If the A V nodal conduction time is sud-
denly increased by 40 ms A, then the tachycardia cycle length is lengthened by 40 ms to AA2 ms. This
produces a decreased A V nodal conduction time of 24 ms compared with AHI B, which then causes
tachycardia cycle length to shorten to AA3 ms C, which then causes the A V nodal conduction time
to lengthen to AH4 ms D etc. The net effect on successive cycles following the perturbation is shown
in E. Thus, the perturbation of tachycardia is rapidly damped in just a few cycles. (Ross et al. [16])

model of supraventricular tachycardia can be constructed [16]. Use of straight line

A V nodal function curves in this model illustrates the basic points. If the line is flat
(gradient greater than -1 or slope less than 45°), the tachycardia is stable and any
perturbation is rapidly damped after a few cycles (see Fig. 4).
If the line is steep with a gradient less than -1 (slope greater than 45°), any per-
turbation is amplified in subsequent cycles with increasingly wilder oscillations lead-
ing to block in the A V node after a cycle shorter than the A V node refractory period
Autonomic Nervous System in SVT 125

A B C
AH AH AH

tAVN ERP tAVN ERP

AH2 0(------
I
40
I
AHl AHl AHl
i
70

\ AH3
t \
\
\
\
\ '.
60°\ 40- \ ~70~ 60°\\
AAl AA AAl AA2 AA3 AAl A

AA
o Fig. SA-D. Effects of perturbation of a computer
model circus movement tachycardia if the AV nodal
function curve is a straight line of slope 60° (i.e.,
steep). The abbreviations are the same as in Fig. 4.
A sudden prolongation of A V nodal conduction
time perturbs tachycardia cycle length, and the
perturbation is amplified in subsequent cycles lead-
ing to block in the A VN when a short cycle falls
short of the A V nodal effective refractory period
(AVN ERP)
Beats of
tachycardia

(Fig. 5). If the line has a gradient of exactly -1 (slope equal to 45°), a perturbation
leads to stable sustained cycle length alternation (Fig. 6). The effects of perturbation
of a stable sustained tachycardia in a patient are shown in Fig. 7 and parallel those
predicted in the flat A V node function curve model. This suggests that the A V node
in stable sustained tachycardias has decreased refractoriness. This is secondary to the
increased sympathetic tone present in the steady-state phase of tachycardia. This is
the major reason why sustained supraventricular tachycardias are inherently stable
after the first few minutes. In contrast, when drugs such as verapamil are given, A V
nodal refractoriness increases, tachycardia cycle length increases, and often cycle
length alternation supervenes due to alternation in A V nodal conduction time [25].
Termination of the tachycardia often occurs then following a short cycle (Fig. 8).
These effects are due to increasing the slope of the A V nodal function curve during
tachycardia (similar to the steep slope model), which tends to make the tachycardia
unstable and prone to perturbation and termination. Similar effects will occur with
any maneuver (including increased vagal tone) which increases A V nodal refractori-
ness. Thus, the autonomic control of A V nodal refractoriness is a fundamental deter-
minant of the stability of supraventricular tachycardias.
126 D.L.Ross

AH A AH B

AA AA
AH

AH3~-------'
." .
•

40
.
45° ", 40
AA AA
AA E
I Fig. 6 A-E. Effects of perturbation of a
40 computer model circus movement tachy-
I cardia if the A V nodal function curve
I is a straight line of slope 45°. Same ab-
40
I breviations as in Fig. 4. A sudden pro-
longation of A V nodal conduction time
initiates stable sustained cycle length
alternation of the tachycardia E.
(Ross et al. [16])
Beats of
tachycardia

Effect of Cardiac Innervation on A V Nodal Function at Rapid Heart Rates

The role of the autonomic nervous system in altering A V nodal function was further
examined at heart rates similar to supraventricular tachycardia in six cardiac trans-
plant recipients [17]. These chronically denervated hearts [14, 24] were compared
with hearts of ten normal patients. In the normal hearts there was significant shorten-
ing of the A V nodal refractory period and a flatter slope of the A V node function
curve at the fastest basic cycle length of atrial pacing. Increasing heart rates therefore
moved the A V node function curve to the left. In comparison, the denervated hearts
showed classic rate-dependent increases in A V nodal refractoriness producing a
steeper slope of the A V nodal function curve at the fastest basic cycle length of atrial
pacing. Thus, the A V nodal function curve was progressively moved to the right.
Autonomic NeIVous System in SVT 127

AH (msec)
160

150

140

130

120
~
110
mAA
313
AA
279
•
mAA
313
AA
268
• msec

Consecutive Beats of Tachycardia _

Fig.7. Example of the effects of perturbation of a stable sustained human circus movement tachy-
cardia by an induced atrial premature beat. AH, A V nodal conduction time; AA, tachycardia cycle
length; mAA, mean cycle length of stable tachycardia. The changes induced in AH inteIVal caused
identical changes in subsequent tachycardia cycle lengths. The effects of atrial premature beats intro-
duced 34 and 45 ms prematurely were rapidly damped in 2-3 cycles, similar to the effects seen in the
flat straight line A V nodal function CUIVe in the computer model tachycardia. (Ross et al. [16])

CL (msec)
560

520

:;: ____ n~W

42 0 ffI JJll
360

Fig.8. Effects of intravenous infusion of verapamil on tachycardia cycle length (eL) in a patient.
Note the induction of sustained cycle length alternation. Mean tachycardia cycle length gradually in-
creases until the tachycardia terminates by block in the AV node following a short cycle. These
changes are identical to those predicted from analysis of the computer model tachycardia if the A V
nodal function CUIVe is moved progressively to the right. (Ross et al. [16])

These studies therefore show that cardiac innervation is necessary for the facilitatory
changes observed during supraventricular tachycardia. Without them tachycardias
would become unstable and terminate spontaneously. It also appears that circulating
catecholamines play a relatively minor role compared with direct neural effects in
modifying A V nodal function during tachycardia.
128 D.L.Ross

Vagal Maneuvers for Termination of Supraventricular Tachycardia

Waxman [26-28] and colleagues have performed an interesting series of studies ex-
amining termination of supraventricular tachycardia by modification of vagal tone.
They showed that the effectiveness of carotid sinus massage can be increased by en-
hancing vagal tone with edrophonium (an anti-cholinesterase inhibitor) or use of
phenylephrine to elevate arterial pressure. The latter was the most effective method.
Conversely, use of atropine abolished spontaneous termination at the level of the
AV node. Use of J3-adrenergic blockade with propranolol also abolished spontane-
ous termination of tachycardia at the level of the A V node during the increased vagal
tone period presumably by blunting the rise in blood pressure and therefore reducing
the vagal response. In addition, propranolol slows the rate of tachycardia which
makes it less likely to terminate spontaneously. In another study, the same group
showed that a deep inspiration and a dependent position facilitated spontaneous ter-
mination of tachycardia due to increased vagal tone.

Effects of the Autonomic Nervous System on Specific Types

of Supraventricular Tachycardia

The effects of the autonomic system on circus movement tachycardias using acces-
sory pathways have been discussed above. Most of these observations also apply to
A V junctional (A V nodal) reentry. The exception is that induction of this arrhythmia
is critically dependent on background autonomic tone, and atropine is often required
to facilitate induction during clinical electrophysiological studies [31]. The retro-
grade, fast pathway used in the classic form of this tachycardia has many characteris-
tics similar to an accessory pathway rather than A V nodal tissue [7, 8]. However, in
some rare cases, control of conduction over the retrograde pathway is exercised by
A V nodal-like tissue which can be facilitated by isoprenaline [9] or minor changes in
background autonomic tone [18]. Retrograde (VA) conduction in the normal heart
is very dependent on autonomic tone and is absent in 30% -50% of human and
canine hearts at rest [10, 21]. However , VA conduction returns in 90% -95% of cases
after modification of autonomic tone [21]. Isoprenaline is the most potent stimulant
in this regard [21] and is the most potent agent for aiding induction in patients with
A V junctional reentry.
Isoprenaline also facilitates ante grade conduction over atrioventricular accessory
pathways, shortens refractory period [29] and increases the maximum rate of trans-
mission of impulses during atrial fibrillation [5]. This correlates with the clinical im-
pression that the severity of episodes of atrial fibrillation in Wolff-Parkinson syn-
drome is increased with enhanced autonomic tone.
True atrial tachycardias are relatively rare (7% of supraventricular tachycardias).
They also show dependence on the autonomic nervous system for induction and
maintenance of the tachycardia even though the A V node is not part of the reentrant
circuit. These tachycardias may also be terminated by vagal maneuvers [30].
Atrial fibrillation and flutter are common clinical arrhythmias. Some types are
precipitated by increased vagal tone and do not respond well to conventional treat-
Autonomic Nervous System in SVT 129

ment [2]. Vagal shortening of atrial refractoriness is thought to be important in this

clinical analogue to acetylcholine and aconitine induced atrial fibrillation and flutter
in experimental preparations. The other major role for the autonomic nervous sys-
tem in atrial fibrillation and flutter is modulation of A V nodal transmission. As ex-
pected, decreased vagal tone and increased sympathetic tone increase the rate of the
ventricular response.

Summary aud Conclusions

The stability and rate of the most common types of supraventricular tachycardia are
determined by the characteristics of the A V nodal function curve during tachycardia,
which is a measure of the degree of refractoriness of that structure. Onset of
tachycardia is associated with a series of phases in autonomic tone, conduction times,
and refractoriness prior to achievement of steady-state conditions. Alterations in
several factors predispose to spontaneous termination of tachycardia during these
phases. The autonomic system, predominantly via its cardiac nerves, is a major deter-
minant of survival of tachycardia through both these initial phases and the later
steady-state sustained tachycardia.

References

1. Borst C, Karemaker JM, Dunning AI (1982) Prolongation of atrioventricular conduction time

by electrical stimulation of the carotid sinus nerves in man. Circulation 65: 432-434
2. Coumel P, Leclercq J, Attuel P, Lavallee J, Lammang D (1979) Autonomic influences in the
genesis of atrial arrhythmias: Atrial flutter and fibrillation of vagal origin. In: Narula OS (ed)
Cardiac arrhythmias: electrophysiology, diagnosis and management. Williams and Wilkins,
Baltimore, pp 243-255
3. Curry PVL, Rowland E, Fox KM, Krikler DM (1978) The relationship between posture, blood
pressure and electrophysiological properties in patients with paroxysmal supraventricular
tachycardia. Arch Mal Coeur 71: 293-299
4. Curry PVL (1979) The hemodynamic and electrophysiological effects of paroxysmal tachycardia.
In: Narula OS (ed) Cardiac arrhythmias: electrophysiology, diagnosis, and management. Wil-
liams and Wilkins, Baltimore, pp 364-381
5. German LD, Gallagher JJ, Broughton A, Guarneri T, Trantham JL (1983) Effects of exercise
and isoproterenol during atrial fibrillation in patients with Wolff-Parkinson-White syndrome.
Am J Cardiol51: 1203-1206
6. Goldreyer BN, Kastor JA, Kershbaum KL (1976) The haemodynamic effects of induced supra-
ventricular tachycardia in man. Circulation 54: 783-789
7. Gomes JAC, Dhatt MS, Rubenson DS, Damato AN (1979) Electrophysiologic evidence for
selective retrograde utilisation of a specialised conducting system in atrioventricular nodal re-
entrant tachycardia. Am J Cardiol 43: 687-698
8. Gomes JAC, Dhatt MS, Damato AN, Akhtar M, Holder CA (1979) Incidence, determinants
and significance of fixed retrograde conduction in the region of the atrioventricular node. Evi-
dence for retrograde atrioventricular nodal bypass tracts. Am J Cardiol 44: 1089-1098
9. Hariman RI, Gomes JAC, El Sherif N (1983) Catecholamine dependent atrioventricular nodal
reentrant tachycardia. Circulation 67: 681-686
10. Hunt GB, Ross DL (1988) Reassessment of AV and VA conduction and AV junctional reentry
in the normal dog heart: The role of altered autonomic tone. PACE 11: 550-561
130 D.L.Ross: Autonomic Nervous System in SVT

11. Josephson ME, Seides SE, Batsford WD, Caracta AR, Damato AN, Kastor JA (1974) The
effects of carotid sinus pressure in reentry paroxysmal supraventricular tachycardia. Am Heart
J 88: 694-697
12. Klein HO, Hoffman BF (1974) Cessation of paroxysmal supraventricular tachycardia by para-
sympathomimetic interventions. Am Intern Med 81: 48-50
13. Levy MN, Martin PJ, Iano T, Zieske H (1970) Effects of single vagal stimuli on heart rate and
atrioventricular conduction. Am J Physiol 218: 1256-1262
14. Mason JW (1980) Overdrive suppression in the transplanted heart: Effect of the autonomic
nervous system on human sinus node recovery. Circulation 62: 688-696
15. Ross DL, Farre J, Bar FWHM, Vanagt EJ, Brugada P, Wiener I, Wellens HJJ (1981) Spon-
taneous termination of circus movement tachycardia using an atrioventricular accessory path-
way: incidence, site of block and mechanisms. Circulation 63: 1129-1139
16. Ross DL, Dassen WRM, Vanagt EJ, Brugada P, Bar FWHM, Wellens HJJ (1982) Cycle length
alternation in circus movement tachycardia using an atrioventricular accessory pathway: a study
of the role of the atrioventricular node using a computer model of tachycardia. Circulation 65:
862-868
17. Ross DL, Anderson KP, Mitchell LB, Rasmussen K, Hunt S, Stinson EB, Winkle RA, Mason
JW (1983) AV nodal function at heart rates similar to supraventricular tachycardia in the inner-
vated and denervated human heart. Aust NZ J Med 13: 404-405
18. Ross DL, Denniss AR, Uther JB (1985) Electrophysiologic study in supraventricular arrhythmias.
In: Schroeder JS, Brest AN (eds) Invasive cardiology, cardiovascular clinics. Davis, Philadel-
phia, pp 187-213
19. Saunders DE, Ord JW (1962) The haemodynamic effects of paroxysmal supraventricular tachy-
cardia in patients with the Wolff-Parkinson-White syndrome. Am J Cardiol 9: 223-236
20. Schlepper M, Weppner HG, Merle H (1978) Haemodynamic effects of supraventricular tachy-
cardias and their alterations by electrically and verapamil induced termination. Cardiovasc Res
12:28-33
21. Schuilenburg RM (1976) Patterns of VA conduction in the human heart in the presence of nor-
mal and abnormal AV conduction. In: Wellens HJJ, Lie KI, Janse MJ (eds) The conduction sys-
tem of the heart. Lea and Febiger, Philadelphia, pp 485-503
22. Simson MB, Spear JF, Moore EN, Kastor JA (1980) Determinants of stabiltiy of tachycardia in
an experimental model. Am J Cardiol 45 : 492
23. Spear JF, Moore EN (1973) Influence of brief vagal and stellate nerve stimulation on pacemaker
activity and conduction within the atrio-ventricular conduction system of the dog. Circ Res 32:
27-41
24. Stinson EB, Griepp RB, Schroeder JS, Dong E Jr, Shumway NE (1972) Hemodynamic observa-
tions one and two years after cardiac transplantation in man. Circulation 45: 1183-1194
25. Vohra J, Hunt D, Stuckey J, Sloman G (1974) Cycle length alternation in supraventricular
tachycardia after administration of verapamil. Br Heart J 36: 570-576
26. Waxman MB, Wald RW, Sharma AD, Huerta F, Cameron DA (1980) Vagal techniques for
termination of paroxysmal supraventricular tachycardia. Am J Cardiol 46: 655-664
27. Waxman MB, Bonet JF, Finley JP, Wald RW (1980) Effects of respiration and posture on
paroxysmal supraventricular tachycardia. Circulation 62: 1011-1020
28. Waxman MB, Sharma AD, Cameron DA, Huerta F, Wald RW (1982) Reflex mechanisms re-
sponsible for early spontaneous termination of paroxysmal supraventricular tachycardia. Am J
Cardiol 49: 259-272
29. Wellens HJJ, Brugada P, Roy D, Weiss J, Bar FW (1982) Effect of isoproterenol on the antegrade
effective refractory period of the accessory pathway in patients with Wolff-Parkinson-White syn-
drome. Am J Cardiol50: 180-184
30. Wu D, Amat-Y-Leon F, Denes P, Dhingra RC, Pietras RJ, Rosen KM (1975) Demonstration
of sustained sinus and atrial re-entry as a mechanism of paroxysmal supraventricular tachycardia.
Circulation 51: 234-243
31. Wu D, Denes P, Bauernfeind R, Dhingra RC, Wyndham C, Rosen KM (1979) Effects of
atropine on induction and maintenance of atrioventricular nodal re-entrant tachycardia. Circula-
tion 59: 779-788
Chapter 10

Heart Rate Changes and ECG Rhythm Disturbances

in the Cluster Headache Syndrome
D.RussELL

Cluster Headache

The cluster headache syndrome includes two headache conditions, cluster headache
and chronic paroxysmal hemicrania.
Cluster headache is characterized by recurrent attacks of extremely severe uni-
lateral headache, which are often accompanied by symptoms and signs which suggest
a disturbance in the function of the autonomic nervous system. The latter present
ipsilateral to the headache and include increased tearing, injection of the conjunc-
tiva, ptosis, miosis, and nasal stuffiness or rhinorrhea.
Changes in heart rate and rhythm may also accompany attacks of cluster head-
ache. In a detailed study [10], electrocardiograms (ECG) were recorded for periods
of at least 24 h in 27 cluster headache patients. Figure 1 shows an overall view of the
heart rate changes which accompanied 81 spontaneous attacks, with histograms of
the heart rates recorded before, at the onset (initial), during, at the end, and follow-
ing attacks. In addition, histograms of the standard deviations of the heart rate val-
ues recorded before, during, and after attacks are shown on the right hand side of
Fig. 1. The changes in mean heart rate for the group as a whole are shown in Table 1.
Statistical calculations were carried out using the Biomedical Computer Program
(BMCP) developed at the University of California, Berkeley, Los Angeles [3]. (For
details, see reference 10.)
At the onset of attacks there was a significant increase in heart rate (Table 1): a
clear peak in the heart rate generally occurring immediately following the patient's
signal that the attack had begun. This increase is also clearly seen in Fig. 1, where the
histogram of the initial heart rate values is displaced to the right when compared with
the heart rate values before attacks.
The increase in heart rate at the onset of attacks was followed by a significant
relative decrease in heart rate during attacks when the latter were compared with
initial heart rate values (Table 1). This decrease is shown in Fig. 1, where the histo-
gram of the heart rate values during attacks is displaced to the left when compared
with the histogram of initial heart rate values. During attacks there were also fre-
quent, rapid changes in heart rate. The standard deviation of heart rate values calcu-
lated during attacks were therefore significantly higher than those recorded before
and after attacks. This is shown in Fig. 1, where the histogram ofthe standard devia-
tions of heart rate values during attacks is displaced to the right compared with the
histograms of standard deviations before and after attacks.
At the end of attacks there was a significant relative increase in heart rate when
compared with values during attacks (Table 1). The histogram of the heart rate val-
132 D.Russell

HEART RATE STANDARD DEVIATION

20
18 Before
11

6
4
2
o
40 50 60 10 60 90 100 110 120 130 1 2 3 4 5 6 7 8 9 10 >10
Initial

12
10
8
6
4
2
o
40 50 60 70 80 90 100 IlO 120 130

28
26
24
22
20
18
16 Du ring
14
12
10
8
6
4
2
o
40 50 60 70 60 90 100 110 120 130 1 2 3 4 5 6 7 8 9 10 >10
End

14
12
10
8
6
4
2
o
40 50 60 70 80 90 100 110 120 130
A fier

16
14
12
10
8
6
4
2
o
40 50 60 70 80 90 100 110 120 130 1 2 3 4 5 6 7 8 9 10 >10
The Cluster Headache Syndrome 133

Table 1. Mean heart rate changes (81 attacks) [10]

Mean heart rate (beats/min) ± 1 SD Significance

Before (B) 71.1± 17.7
B-1 P~O.OOO1
Initial (I) 85.4± 17.7
I-D P~O.OOO1
During (D) 73.1±21.7
D-E P=O.013
End (E) 78.7±19.7
E-A P~O . OOO1
After (A) 73.9±17.6

A OB.35 : HR122 Before

,;ovJ'.\.("I ~p..J.I"-oI.A-..i"""".A-oJ.J'-J'(
T " , I I '1
OB.45:HR96 During Gradel

~:\r-JrJ:v---J~~
OB.55: HR 95 After
----.0-Jr'-"~~k..-~
B 10.25:HR9B Before

-+---r-+--+·,-+Jlr +"-Y-1r
10.45 :HR100 During Grade3

-t"---r---r-+,J:.r-Y~r
C l2.25:HRllB Before

r-J~~~r
l2.35:HR100 During

+'--.Jr~-+'-+'-+/'-1I.-J~ Fig. 2. Patient No. 1. The ECG showed

l3.00:HR116 After an increased frequency of ventricular
premature beats (x) during attacks A- C.
~ (Russell and Storstein [10])

ues at the end of attacks is therefore displaced to the right when compared with the
histogram of heart rate values during attacks (Fig. 1).
After attacks the heart rate showed a relative decrease when compared with val-
ues at the end of attacks (Table 1), and the histogram of heart rate values is displaced
to the right when compared with that for the heart rate values at the end of attacks.
During the 24-h ECG recordings 5 (18.5%) of the 27 patients had ECG rhythm
disturbances.
The first patient was a 36-year-old male who had primary chronic right-sided clus-
ter headache attacks during the past 4 years. He had no cardiac complaints with the

..
Fig.t. Histograms of before, initial, during, end, and after heart rate values (left side) and histo-
grams of the standard deviations of heart rate values before, during, and after attacks (right side).
Vertical axes, number of attacks; horizontal axes, heart rate (beats/min) and standard deviations.
(Russell and Storstein [10])
134 D.Russell

Fig. 3. Patient No.2. Frequent

ventricular premature beats (x)
throughout the ECG recording,
with no apparent increase in
frequency during attacks.
(Russell and Storstein [10])

exception of occasional palpitations, and the physical findings were normal. Radio-
logical evaluation showed normal configuration and size of the heart. A 12-lead
ECG was normal with the exception of ventricular premature beats. Holter monitor-
ing revealed that these premature beats were frequent and as can be seen from Fig. 2
they were more frequent during headache attacks.
The second patient was a 47-year-old male who had periods with right-sided clus-
ter headache attacks during the past 2 years. Continuous tape-recording revealed
frequent ventricular premature beats with no apparent increase in frequency during
headache attacks (Fig. 3). This patient occasionally complained of dizziness and pal-
pitations; otherwise, he had no cardiac symptoms. The heart size on the X-ray was
normal, as were the resting and exercise electrocardiograms.
The third patient was a 56-year-old male who had periods with left-sided cluster
headache attacks during the past 20 years. He had no cardiac complaints and normal
clinical, heart X-ray, and 12-lead electrocardiographic findings. Continuous 24-h
ECG monitoring showed that the patient's headache attacks were associated with
atrial fibrillation which was otherwise not present. This patient experienced head-
ache attacks which started in atrial fibrillation and ended in sinus rhythm (Fig. 4) or
started in sinus rhythm with transition into atrial fibrillation during attacks (Fig. 5).
The Cluster Headache Syndrome 135

18.37,HR 88 AF Initial 18.41,HR 158 AF During Grade5

~ ~r('¥Jr-J('Jr¥~/'
During SR

~~ 19.05,HR 68 End SR

J)lJ)L AJ J\.-.J~JL I

Fig.4. Patient No.3. An attack associated with atrial fibrillation (AF), which started before the
onset of the attack, with conversion to sinus rhythm (SR) during the attack. x, Extrasystoles. (Russell
and Storstein [10])

The fourth patient was a 24-year-old male who had periods with right- or left-
sided cluster headache attacks during the past 6 years. He had no cardiac symptoms
and normal findings on clinical, radiological, and electrocardiographic examination.
During attacks of cluster headache he developed transient first-degree atrioventricu-
lar block (Fig. 6) .
The fifth patient was a 44-year-old male who had left-sided cluster headache for
9 years. During attacks this patient consistently developed bradycardia and sinoatrial
block with nodal or ventricular escape beats. The minimum recorded heart rate dur-
ing an attack was 20 beats/min and the longest observed pause between two consecu-
tive beats was 5.4s . He experienced no symptoms from this bradycardia. The ECG
findings during one attack are shown in Fig. 7. This patient was studied in detail first
during a bout (not during an attack) and then during a headache attack. The heart
rate in the resting supine position was 61 beats/min. A Valsalva maneuver did not
significantly change the heart rate while carotis massage reduced the heart rate by 5
beats/min. During a postural test, which lasted 6 min, heart rate, blood pressure, and
ECG were recorded at 2-min intervals. The heart rate increased from 55 to 65 beats/
min while blood pressure fell from 110175 to 100170mmHg. Both responses were
within the normal range. The heart rate at the beginning of the attack was 51 de-
creasing to 35 after 2 min, while SA block developed after 4 min. The heart rate re-
mained low while the pain gradually increased and the patient experienced coughing,
stuffiness of the nose on the left side, and increased sweating on the left side of the
forehead was observed. During the phase of relatively stable pain, 1 mg atropine was
injected slowly intravenously for 3 min. During the injection the patient's heart rate
increased from 35 to 73 beats/min and sinus rhythm reappeared , while the headache
completely subsided within the 5 min following the injection.
136 D.Russell

04.30,HA 63 Before SA

~~/Lv++ 1I

~~ 05.1~'HA 58 During I SA

~~~
II ~ I

Durll'l9 SFf

Grade1

Fig. 5. Patient No.3. Atrial fi-

brillation (AF) developed during
the attack and continued after
the attack had ended. x, Supra-
ventricular and ventricular extra-
systoles; SR, sinus rhythm;
Grade, headache intensity.
(Russell and Storstein [10])

Chronic Paroxysmal Hemicrania

Chronic paroxysmal hemicrania (CPH) resembles cluster headache with regard to

localization, intensity, and character of the pain. The main clinical differences be-
tween CPH and cluster headache are: (a) female predominance; (b) high maximum
frequency in CPH which exceeds 15 attacks per 24h and can be much higher; (c)
CPH attacks are usually shorter than in cluster headache lasting between 5 and
20min; (d) attacks of CPH are abolished completely by prophylactic indomethacin
treatment.
The autonomic symptoms and signs which accompany attacks of CPH and cluster
headache are similar. However, in contrast to cluster headache no typical heart rate
change was found in association with 105 spontaneous attacks in 5 patients [11].
A prominent finding in all CPH patients was that there were often large and rapid
The Cluster Headache Syndrome 137

06.145:HR76

07.10:HR110 During

-' plit\p[flAf
J ""_,J L.,( ~ 1

·w
07.3oJHR6~ During

' i~ ~
~tl

Fig.6. Patient No.4. The development of first-degree atrioventricular block during an attack. p,
p wave. (Russell and Storstein [1OJ)

07.15 : HR 69 Before 07.30 :HR76 Initial 07.35:HR49 Start SR

~ ~-- j- ~
07.45tHR20 SA Block During
r 1 A
~--~__~~ ·~------~v~--~~~~~--~~~y~-
U £8 ~ U
07.47:HR 28 SA Block
~
(8 (8
~------~----~~----~~----y~
(8 EI f8
07.52:HR2& SA Block
p

£8
SA Block
4,"

f8
08.10:HR78 End SR 08.35:HR72 After SR

Fig. 7. Patient No.5. During attacks the ECG showed bradycardia and sinoatrial block (SA block)
with nodal or ventricular escape beats (EB) . SR, Sinus rhythm ; p, low-voltage p wave; I I, pause
between two consecutive beats. (Russell and Storstein [1OJ)

variations in heart rate in association with attacks, which were not systematic but
which could occur "before," "during," or "after" attacks. Figure 8 demonstrates this
feature in one of the patients, showing the large variations in heart rate which accom-
panied six typical attacks. Attack No.2 is an example with three peaks of sinus tachy-
cardia (i.e., heart rate> 100 beats/min) , whereas attack No.6 showed two peaks
138 D.Russell

1311 HEART RATE

1211
,\' ...
2 I,

1\
,
I ,
110 I ,

.1,,,\ ...... .. ..
I ,
I ,
I ,

,,
100 I '
3 I

__.-.of

6
eo

50
-
'\..-/.

81 B2 113 5 01 112 03 D4 E AI A2 All

TIME
Fig. 8. CPR patient. The heart rate values (vertical axis) before (B l - 3), at the start (S), during (D l -4),
at the end (E), and after (A l - 3) six attacks. The attack numbers are shown on the left side of the figure.
(Russell and Storstein [11])

with sinus tachycardia, the highest heart rate being 126 beats/min and the lowest 60
beats per min, i.e., a range of 66 beats/min.
Two of the five CPH patients developed arrhythmias in association with attacks.
The first patient was a 30-year-old woman who had a normal resting ECG and no
cardiac complaints. She first took part in the study immediately after the diagnosis
of CPH was made, prior to initiation of indomethacin treatment. During this study
period, an ambulatory EeG recording was carried out for 17 h, and she developed
ten extremely severe headache attacks. Sinoatrial block with bradycardia developed
during eight of them (Fig. 9). These findings were most pronounced at the height of
the attacks. She was studied again 56 weeks later, in a period when indomethacin
was discontinued. During this second study period, ambulatory ECG recording was
carried out for 24 h, and she had five mild and one moderate attack. The ECG
showed marked sinus arrhythmia during all of these attacks but with no episodes of
sinoatrial block. An ambulatory ECG recording was carried out for the third time
when she was symptom-free on indomethacin treatment. This 24-h recording showed
regular sinus rhythm without either sinus arrhythmia or sinoatrial block.
The second patient was a 65-year-old woman who had a normal resting ECG
when headache-free and no symptoms suggestive of cardiovascular disease. During
the 24-h ambulatory ECG recordings, she had 38 extremely severe headache attacks.
Her ECG during this recording showed intermittent bundle-branch block, probably
a right bundle-branch block as the R wave in V1 was much higher than the R waves
seen in the narrow QRS complexes. She also experienced episodes with atrial fibril-
lation in association with 28 of the 38 attacks which occurred "before," "during," or
"after" the attacks (Fig. 10). A 24-h ambulatory ECG recording in this patient taken
when she was symptom-free on indomethacin treatment showed regular sinus rhythm,
with no signs of bundle-branch block or atrial fibrillation.
The Cluster Headache Syndrome 139

81 HA 15 H .5. ~

83 HA 143

Dl HA 31

DZ HA4Z

HAll

AZ H" 11

Fig.9. A 30-year-old CPH patient who had normal sinus rhythm (B I ), which changed into sinus
tachycardia with a heart rate of 143 beats/min and showed ST-T changes (B3)' During this attack
(D], D 2), she developed bradycardia and sinoatrial block with nodal escape beats (EB). At the end
(E) and after the attack (A 2), she again had normal sinus rhythm. HR, Heart rate; p, low·voltage p
wave. (Russell and Storstein [11])

Discussion

The pattern of heart rate change in association with attacks of cluster headache sug-
gests that there may be an imbalance in the central control of the function of the
autonomic nervous system. At the onset of attacks there is a rapid increase in heart
rate followed by a relative decrease during attacks. In addition, variations of heart
rate are much more pronounced during attacks. This concept of an imbalance in the
central control of autonomic function in cluster headache is to some extent supported
by observations made in patients where autonomic impulses via the sympathetic and
parasympathetic nervous system do not reach the heart. In this situation, variations
in heart rate are greatly reduced. In cardiac transplant recipients, the donor heart
rates at rest are faster than normal and similar to "intrinsic" heart rates which are ob-
tained by combined vagal and adrenergic blockage [6, 12]. Respiratory influences on
the donor heart rate are absent, and there is a reduced or absent heart rate response
to various pharmacological and physiological maneuvers [4, 12]. Similar findings
have been reported in a patient with total cardiac denervation due to diabetic auto-
nomic neuropathy [8].
140 D.Russell

81 HR8I AF £.u . ~

s HR 121 AF

02 HRI9 AF

HR 13 SR

At HRI2 SR

Fig. 10. CPH patient who showed atrial fibrillation (AF) before (B/), at the start (S), and during (D 2)
this attack, which changed into regular sinus rhythm (SR) with bundle-branch block at the end(E)
and after (A/) the attack. HR, Heart rate. (Russell and Storstein [11])

A central disturbance in autonomic function to the heart may also be present in

association with attacks of CPH . However, this disturbance may possibly be of a dif-
ferent nature and longer-lasting than in cluster headache. A typical pattern of heart
rate change could not be detected in association with CPH attacks. In addition, non-
systematic and rapid variations in heart rate were observed not only during attacks
but also "before" and "after" attacks of CPH.
An imbalance in central autonomic output to the heart may also explain the oc-
currence of specific cardiac arrhythmias in association with attacks of cluster head-
ache and CPH. These may possibly be due to parasympathetic overactivity since
stimulation of specific sites in the central nervous system in animals has resulted in
arrhythmias which suggest enhanced parasympathetic activity [5], i.e., sinus brady-
cardia, atrial flutter, and atrial fibrillation. Most centrally induced arrhythmias, how-
ever, probably represent the net effect of adrenergic and parasympathetic discharges
[13]. Together with simultaneous adrenergic stimulation, parasympathetic influences
promote instability of the predominant pacemaker site [1, 13]. Parasympathetic ac-
tivity may induce arrhythmias by depressing the activity of the sinus pacemaker cells
The Cluster Headache Syndrome 141

by blocking or slowing atrioventricular conduction and by promoting nonuniformity

of repolarization of adjacent regions of the myocardium [1, 13]. In this regard,
cholinergic stimulation has been known for many years to be associated with induc-
tion and prolongation of atrial fibrillation [1, 2, 7].

Conclusions

Attacks of cluster headache and chronic paroxysmal hemicrania may be accom-

panied by a disturbance in both heart rate and rhythm. This suggests that the auto-
nomic disturbance is not confined to the symptomatic area and that a central distur-
bance in autonomic function may at least partly be responsible for the autonomic
symptoms and signs in these two headache types [9].

References

1. Alessi R, Nusynowitz M, Abildskov JA, Moe GK (1958) Nonuniform distribution of vagal

effects on the atrial refractory period. Am J Physiol194: 406-410
2. Burn JH (1957) Acetylcholine and cardiac fibrillation. Br Med Bull 13: 181-184
3. Dixon WJ, Brown MB (eds) (1979) BMDP Biomedical Computer Programs, P-series. Univer-
sity of California Press, Berkeley, Los Angeles
4. Cannom DS, Graham AF, Harrison DC (1973) Electrophysiological studies in the denervated
transplanted human heart: responses to atrial pacing and atropine. Circ Res 32: 268-278
5. Hockman CH, Mauck HP Jr, Hoff EC (1967) Experimental neurogenic arrhythmias. Bull NY
Acad Med 43 : 1097-1105
6. Jose AD, Collison D (1970) The normal range and determinants of the intrinsic heart rate in
man. Cardiovasc Res 4: 160-167
7. Loomis TA, Captain MC, Knop S (1955) Auricular fibrillation induced and maintained in ani-
mals by acetylcholine or vagal stimulation. Circ Res 3: 390-396
8. Lloyd-Mostyn RH, Watkins PJ (1976) Total cardiac denervation in diabetic autonomic neu-
ropathy. Diabetes 25 : 748-751
9. Russell D (1985) Studies of autonomic functions in the cluster headache syndrome. Thesis, Uni-
versity of Oslo
10. Russell D, Storstein L (1983) Cluster headache: a computerized analysis of 24h Holter ECG
recordings and description of ECG rhythm disturbances. Cephalalgia 3 : 83-107
11. Russell D, Storstein L (1984) Chronic paroxysmal hemicrania: Heart rate changes and ECG
rhythm disturbances. A computerized analysis of 24 h ambulatory ECG recordings. Cephalalgia
4:135-144
12. Stinson EB, Griepp RB, Schroeder JS, Dong E Jr, Shumway NE (1972) Haemodynamic obser-
vations one and two years after cardiac transplantation in man. Circulation 45 : 1183-1194
13. Wallace AG, Schaal SF, Sugimoto T, Rozear M, Alexander JA (1967) The electrophysiologic
effects of beta-adrenergic blockade and cardiac denervation. Bull NY Acad Med 43: 1119-1137
Chapter 11
Blood Pressure Assessment
in a Broad Chronobiologic Perspective
F.HALBERG, E.BAKKEN, G.CORNELISSEN, I. HALBERG, E.HALBERG, and P. DELMORE

Chronobiology and Medicine

Chronobiology deals with the rhythmic patterns that occur in all forms of life [1-5,
7-32,34-37,39-43]. Virtually all organisms exhibit approximately daily (circadian)
cycles. In human beings, prominent rhythms are found in blood pressure, circulatory
pulse, body core, and surface temperature and in chemical variables of blood, urine,
and tissues. The medical section of this new science explores the relationships of
rhythms to prediction, prevention, diagnosis and treatment of diseases. Several de-
cades of research worldwide have established that medical diagnoses can be subject
to a much higher proportion of false positives (e.g., "office hypertension") and false
negatives (e.g., "odd-hour" hypertension) when only single samples are taken at
arbitrary times of the day instead of taking rhythms into account. Radiation, chemo-
therapy and other treatments have been shown to have markedly different efficacy
and safety depending upon the pattern of administration within the day. Without en-
gineering tools, chronobiology had to rely on manual-measurement series that were
cumbersome to collect and required methods of analysis that were not generally
available. Modern bioengineering with microcomputers can bring the findings of
chronobiology on circadian and other rhythms into the mainstream of medicine.
Chronobiology can thus provide new dimensions to health care.
Medicine stands on the threshold of a new era in prevention as well as diagno-
sis and treatment, based on the combination of several emerging engineering
technologies with an understanding and application of the health effect of life's
rhythms. Through chronobiologic programs, modern engineering technologies can
lead to:

The large-scale use of portable, personal long-term ambulatory monitors of bio-

logic variables such as blood pressure, pulse, the ECG or EEG, variables that
undergo large, spontaneously and reactively recurrent changes
The availability of data base systems to acquire and analyze the considerable vol-
umes of data obtained from personal monitors
The availability of statistical procedures to analyze and model the biologic
rhythms and from them to devise optimal-dosage time patterns for specific indi-
viduals
The availability of portable, programmable devices to administer therapy, e.g.,
by physiologic rate-adjusted cardiac pacemakers or drug pumps.
Blood Pressure and Chronobiology 143

The routine employment of chronobiologic methods in medical screening, diag-

nosis, prognosis, treatment and, most important, disease prevention, remains the
challenge to be met.
The importance of chronobiology in medicine is herein illustrated in the context
of blood pressure, even if recent debates revolving around measuring cholesterol
[38] may also benefit from a chronobiologic approach [28, 29, 42]. More than half of
the serious disease and death occurring in middle-aged people is linked to behaviors
that can be changed or health conditions that can be prevented or treated. Three
very important examples are smoking, a diet high in saturated fat, and high blood
pressure.

Blood Pressure Measurements that miss the Chronobiologic Point

As an example for the application of chronobiology, health assessment of an indi-

vidual need not rest only on a visit to the clinic. Visits are currently scheduled at a
time convenient to both patient and physician, without taking into account the great
variations that occur in all of our physiologic variables over a 24-h period and over a
weekly period and even a monthly and yearly period. Since quite large variations
take place in human blood pressure over 24-h periods, chronobiology as the study of
the parameters of biologic variation in relation to time is particularly pertinent to our
way of handling blood pressure [3, 13, 19-22, 24, 25]. Indeed, at present (non-
chronobiologic) blood pressure screening is usually offered according to the recom-
mendation of the World Health Organization (WHO) and on only some occasions by
the use of 24-h ambulatory cardiovascular monitoring.
Most health professionals and most of the broad public ask only: What is your
blood pressure? Official convention focuses upon "the" blood pressure, relying on
casual measurements. Such measurements are recommended, among others, in the
WHO Guidelines [6, 44, 47]. For a test of a mild blood pressure elevation (90-104
mmHg diastolic), without organ damage and "to provide a safe classification of pa-
tients," the International Society of Hypertension joins WHO in recommending the
criterion of the average of 2 (casual) diastolic measurements. If the average of these
is 90mmHg or above, further measurement pairs on at least two further days over
four weeks are recommended. If the average values of these are above 100mmHg,
drug treatment is to start. If in turn these means are below 100mmHg, further mea-
surements over 3 months and behavioral intervention, but not drug treatment, are
recommended. If the overall average is now above 95 mmHg, drug treatment is re-
commended; if the average is below 95 mmHg, behavioral intervention is prescribed
with further observations over 3 months. It is left to the individual to pick the time
of measurements.
Implicit in the foregoing convention are the assumptions that (a) variations in
blood pressure and heart rate are the effect of changes in posture, responses to
meals, exercise or other stimuli, emotional or physical, and that (b) a short span of
sitting or lying down of, say, five minutes or so, in the setting of a medical office, is
a better standardization than no standardization whatsoever, in order to somehow
"control" the stimuli of daily life and thus to approximate the (true) blood pressure.
144 F.Halberg et ai.

The recommendation of 24-h ambulatory blood pressure and heart rate monitor-
ing as such exploits the availability of new monitors only and does not take full ad-
vantage of the data in the light of chronobiologic facts. Assumption (a) above is the
excuse for refraining from a consideration of the dynamics of blood pressure. These
continue to be regarded as trivial effects of changes in posture, notably exercise and
other stimuli. With most ambulatory monitoring instruments, it remains necessary to
stop what one does physically during the actual measurement, but with this qualifica-
tion, assumption (b) is changed into endeavor (c), namely into the aim of obtaining
a reflection of average pressure in "real life" by a 24-h and perhaps also by a daytime
and nighttime average pressure. Since all variations are treated as responses to exter-
nal stimuli, one does not do much if anything about them, except perhaps for asking
that changes in posture, meals, etc., be recorded. In debates about nonchronobio-
logic ambulatory 24-h records versus casual measurements, it should be realized that
both approaches present serious shortcomings. Built-in rhythms and their character-
istics such as a measure of the extent of a regular 24-h change (circadian amplitude)
and of the timing of high values (acrophase) are ignored. Definitions of some rhythm
characteristics are provided in Fig. 1. That three of these characteristics of heart rate
are partly inherited is shown by studies on twins reared apart [26]. Indirect evidence
for the endogenicity of the human heart rate and blood pressure rhythms is available
from their free running with natural periods differing from environmental ones in
social isolation [2] and under certain other conditions [16]. The fact is also demon-
strated that responses to external stimuli (Le., the blood pressure change in response
to the immersion of the hand in cold water [23] or simply reading aloud) depend
upon the stage of partly endogenous circadian rhythms.
The cost effectiveness of nonchronobiologic 24-h ambulatory monitoring has
been scrutinized thus far without considering what can be learned from the spon-
taneous and reactive blood-pressure and heart-rate dynamics. The investment into
new machines has not been accepted as justified on the basis of an improved 24-h
average or of averages for fractions of the day, such as the waking, sleeping or other
spans. Thus far, the health profession's response has been guarded. Whatever their
data source, whether they rely upon conventional, manual, or ambulatory automatic
measurements, most health care personnel continue to act according to what "the (or
the average)" blood pressure of a given person is. This inquiry only into the or the
average blood pressure is insufficient when a more reliably defined mean and added
dynamic information can be extracted from the same record. There is indeed another
alternative to the WHO guidelines [6, 44, 47], one that adds a new dimension to
blood pressure assessment: the chronobiologic approach, as summarized in Fig. 2.
The message of this note is that there is a need for chronobiologic monitoring in
order to improve the sensitivity of screening, the reliability of diagnosis, the accuracy
of prognosis and the timing and efficacy of treatment. But first and foremost, there
is a need to emphasize approaches for the prevention of high blood pressure.
Blood pressure chronobiology focuses upon both the predictable and the global
dynamics of the variation in blood pressure that is likely to be missed if one measures
at only one or another convenient time of day. As in the case of the 24-h profile,
some treated or untreated cases of high blood pressure show elevations at times
when casual measurements are not taken as a rule, as for instance during odd hours
or sleep (Fig.3) [25]. To detect such cases, systematic manual or preferably auto-
Blood Pressure and Chronobiology 145

rhythmic
/
function

Fig.t. Definition of Rhythm Characteristics

MESOR, rhythm-adjusted mean (Midline-
Estimating Statistic Of Rhythm), defined as
average value of rhythmic function (e.g.,
cosine curve) fitted to data; expressed in

L----------------
same units as original data. Note that
MESOR will differ from arithmetic mean if
data are unequidistant (e.g., concentrated
total predicted change near crest of rhythm) and/or cover non-
(double ompntude) integral number of cycles; PERIOD, dura-
tion of one complete cycle in rhythmic func-
reference
lime
tion; expressed in time units, such as sec-
(ACROPHASE onds, hours, days or years, or in physiologic
units such as complete cardiac, respiratory
or mentrual cycle; equated to 360° for angu-
lar expression of acrophase; AMPLITUDE,
half of total predictable change in rhythm,
lIme defined by rhythmic function fitted to data;
expressed in original or "relative" units,
e.g., as percentage of series mean or
MESOR; ACROPHASE, lag from ref-
erence time of rhythm's crest-time, defined
by rhythmic function fitted to data; usually
ACROMETRON
expressed in (negative) degrees, with 360° ==
period, 0° = reference time; customary time
units (e.g., clock-hours and minutes, days,
weeks, months or years) or physiologic units
(e.g., number of heart beats, respiratory
or menstrual cycles) also appropriate for
rhythm synchronized with corresponding
period; ACROMETRON, highest predicted
value taken by rhythmic variable, i.e., the
( sum of the amplitude and MESOR
146 F.Halberg et aI.

1) Conventional 2) Chronobiologic

~
Up to 3 casual measurements
made while seated
+
Automatic readings
(24-48 hours or longer)
on up to 3 occasions

E
~ Descriptive statistics: ......f - - - - - -......
Hourly
Waking
Means - - - - - - - - - - - 1 Sleeping
Standard deviation 24-hour
Range
Graph of pressures and events
Cumulative frequency distribution
Inferential statistics: +-_____----l
Reliable mean (MESOR)

t
Chronoclesms
Sex-,age-,clock-hour-,season- and cardiovascular risk-specified
individualized and/or peer-grouD standards ()
For a better interpretation of subsequent OJ
single samples and of time series for o
15
ii
~
Rhythm dynamics
oc
e
~
()
Circadian amplitude,A
Circadian acrophase,0
Waveform:{A,0) of harmonics
Period(s) when record is of suitable length
Measures of extent and timing of excess and/or deficit,expressed
in mm Hg x h,including a projection over 10 years .....---'
Fig. 2. Scheme indicating that the chronobiologic approach includes descriptive and inferential statis-
tics. Chronobiologic around-the-clock monitoring yields several kinds of chronodesms. Some of
these are time-specified reference limits for the interpretation of strategically placed (rather than
casual) measurements. Other reference standards, paradesms, serve for interpreting the MESOR;
amplitude, A; acrophase, 0; and periods, r, separately or jointly as features of the quantified dy-
namics of time series

matic around-the-clock measurement is required. Chronobiologic evaluation also re-

veals that a 24-h profile does not suffice for a first screen. Thus, in a majority of
cases, statistically significant differences are found in the 24-h average of two 24-h
profiles repeated w\thin a few weeks, even when each profile is based on relatively
dense (7.5-min) measurements. A 48-h average, more reliable than the 24-h average,
is not the sole and not necessarily the major merit of the- chronobiologic approach.
An important point in the chronobiologic approach is the need to start any screening
with around-the-clock measurements rather than carrying out such a monitoring pro-
file only in the case of a suspicion of deviant blood pressure. The contribution by
chronobiology, however, is the ability to account for a usually large circadian varia-
tion, encountered over 24 h under ambulatory conditions in assessing any cumulative
Blood Pressure and Chronobiology 147

w
a:: 180~~----------r---------~~--~------'--------'
~
til
til
W
a::
D. :',:': ..:...: ....• .0:,.: .. ': _
CO; 130 .
o:!: ','
.:.:...... '.' ',' ,.,:-0:":':' .
·'0·
.'" : "', ~ '.-":".

9~
m-
(.) 80
::::i
~

01' ~~ _ _ _ _ _ _ _ _ _ _L -_ _ _ _ _ _ _ _ _ _ ~ __________ ~ ______ ~

E
80r--,---------,-----------,---------.-------.

65

50
0< I
00:00 2:00 AM 4:00 AM 6:00 AM 7:30 AM

TIME (Clock Hours on February 25, 1985)

Fig. 3. An unsolved problem: hypertensive episode during sleep without concomitant tachycardia. A
marked rise in systolic blood pressure (BP) during the night (boldface), lasting for about 50 min, is
seen (top row), in the absence of corresponding changes in heart rate (bottom row). This fact sug-
gests that the changes in blood pressure are not due to changes in movement. In the absence of a
concomitant electroencephalographic tracing, one can only speculate as to whether the subject had
a panic episode in a dream. If indeed he dreamt, one would also have to account for the mechanisms
that under certain conditions, in a dream, bring about an increase in BP and under others a decrease
in BP, as observed during another sleep episode [25] on the same subject (not shown). Whether the
changes noted, the hypertensive episode in particular, are related to the medication (nardil) which
the subject took, must be explored. The changes in BP may reflect states of mind of interest to those
concerned about brain-heart interactions. The 2325 values per variable monitored on this man (not
all shown) allow the isolation in his BP spectrum of 1.6- and 1.8-h components. Certainly, the data
during the episode exceed the range of change usually encountered for BP ultradians. The 50-min
duration of the episode is in keeping with the possibility that it is an exacerbation lasting for half the
duration of an underlying ultradian (-1.7-h) rhythmicity. Such episodes would not be detected by
sole reliance upon daytime measurements. The episode may pose a threat if it is maintained for
longer than half of an ultradian cycle (about 50min), a period actually isolated for this particular
subject, and/or if it represents more than an isolated episode in the particular BP series examined

excess or deficit on the one hand while resolving part of the overall variation as pre-
dictably rhythmic and assessing any changes in rhythm characteristics on the other
hand.
By the fit of cosine functions with periods of 24 h and submultiples thereof, com-
puter software resolves the characteristics of variation, as shown in Fig. 1. Chrono-
biologic methods can be applied to measurements taken manually, with a sampling
148 F. Halberg et al.

schedule shown to be useful for a given purpose and adopted with a discipline that
has to be taught in the context of early formal as well as adult education. The use of
presently available modern automatic, fully ambulatory recording systems, however,
greatly facilitates data collection and, in combination with proper software, renders
the chronobiologic approach to blood pressure variation much more practical than
reliance only upon a few casual measurements.

Methods

If one measures any of the body's physical variables continuously or at frequent

intervals around the clock (preferably for at least two days), the plotted data invari-
ably show rises and falls that have an obvious 24-h periodicity. By submitting these
data to mathematical analysis, one can pull out the fundamental frequency and plot
it as a sinusoidal function. The cosine curve can be used for convenience. From this
plotted cosine, one can see the amplitude of the fundamental variability and the tim-
ing of the peak(s) and valley(s) and can also plot the midline for the variability.
These mathematically calculated features are often sensitive indicators of present or
potential disease that cannot be found in a few casual measurements, e.g., of blood
pressure and heart rate, taken at an arbitrary time in a physician's office. In many
cases, this chronobiologic approach yields meaningful information when casual read-
ings do not.

Hardware

Until recently, the taking of blood pressure readings every 10 min or so around the
clock has not been clinically practical. Now, however, automatic blood pressure re-
cording equipment has become available. Smaller and smaller wearable equipment
for fully ambulatory recording is being developed [25]. Apart from the nearly noise-
less and hence less sleep-disturbing miniaturized ambulatory monitor, manufactured
by Colin Medical Instruments (Japan), now extensively used by us and others,
perhaps the ambulatory monitors most widely used today are the Pressurometer III
(Del Mar Avionics, USA); the Ambulatory Blood Pressure Monitor (Instruments
for Cardiac Research, subsidiary of Squibb, USA); and the Acutracker Model103C
(Oxford Medilog, USA). In their present form, the large-scale, long-term use of
these monitors is difficult and costly. The Pressurometer demands the placement of
electrodes that are a nuisance for the patient and, after long-term use, may cause
skin rashes; the Squibb instrument requires frequent battery changes; and the Colin
Medical Instrument requires CO 2 cartridges. There are also limitations in terms of
data storage in solid-state memory, although the Colin Medical Instrument holds
over 600 measurements that can easily accomodate more than one week of data with
a sampling interval of 30 min. This kind of equipment begins to give us data that
represent real life and take into account what happens to the blood pressure during
most of the range of life experiences from different stages of sleep to various kinds
of experience in wakefulness on weekends as well as weekdays.
Blood Pressure and Chronobiology 149

Recordings over at least 48 h and preferably a week show the effects of emo-
tional, physical and time impactors that cannot be assessed in single readings, as, for
instance, the episodes of relatively high or low pressure occurring during sleep (Fig. 3)
[2S]. During one night, the blood pressure rose for -SO min, without a concomitant
change in heart rate. In another night, in the same subject [2S], an episode of a rela-
tively low blood pressure occurred.
It is important to know the duration of episodes of such hyper- or hypotension
and what their extent might be, whenever they occur. A chronobiologic approach is
particularly rewarding when it uncovers events that are not seen in casual measure-
ments.

Analyses

A chronobiologic blood pressure assessment relies on both parametric and non-

parametric approaches. For either approach, the reference standard is derived from
data of living age- and sex-matched peers rather than solely from the tombstones of
a previous generation. When time-varying reference limits are unavailable as yet, ex-
cess can refer to the sum of the circadian amplitude and the rhythm-adjusted mean
or MESOR. This sum, the acrometron, is based upon around-the-clock sampling
on healthy peers and is thus preferred to an arbitrary limit such as 140 and 90mmHg
for systolic and diastolic blood pressure, respectively. The reference limits used here
are derived from only 39 clinically healthy ambulatory North American men, each
monitored at 7.S-min intervals for 24h. This sample size of a reference standard will
be replaced by one based on several hundred already available series obtained by
noninvasive 24-h monitoring at 7.S- or lO-min intervals. The recommended monitor-
ing span has now been extended to a minimum of 48 h. Hence, reference standards
based on 48-h samples are needed, as are reference standards for "skeleton" sampling
[2S], namely for uninterrupted sampling every lOmin during sleep with interruptions
during wakefulness, implemented with an automatic oscillometric room-restricted
Nippon Colin machine.

Sphygmochrons: Monitoring ProIDes over Time for Comparison with Peers

Special forms have been designed to report the blood pressure status of patients
chronobiologically. These forms focus on circadian features of blood pressure varia-
bility, such as the MESOR, amplitude and acrophase (Fig. 1). In addition, indices of
blood pressure excess (or deficit) are evaluated by comparing the patient's own re-
cord with the characteristics of healthy peers. These forms assist in the proper detec-
tion and treatment of patients and can be used as a convenient tool in clinical prac-
tice. This data summary chart, developed by us, has been dubbed a sphygmochron,
a blood pressure and heart rate monitoring profile over time [2S]. It represents a
computer comparison of a person's data with clinically healthy peer group limits.
These sphygmochrons describe a number of chronobiologic characteristics that result
from an approach summarized in Table 1.
150 F. Halberg et al.

Table 1. Chronobiologic endpoints, e.g., of human blood pressure

1. MESOR,M Midline-estimating statistic of rhythm,

based on systematic sampling, taking dynamics into account;
replaces mean of casual measurements at arbitrary times
2. Range,R total
partial, e.g., 90%, 50%, circadian range
(e.g., of hourly mean values)
3. Standard deviation, SD
4. Trends Intercept and slope of linear trend;
coefficients of higher-order (e.g., quadratic or cubic) polynomials
5. Periodicities Rhythm characteristics, e.g.
circadian amplitude, A
circadian acrophase, 0
circadian waveform, (A, 0) of harmonics
6. Non-periodic, Parameters describing shape of spectrum,
non-trend components e.g., high frequency dominant over low frequency or vice versa
All of the above endpoints mayor may not be deviant. Other endpoints, related to cumulative deviation,
are:
7. Excess short-term: extent: hyperbaric index
duration & timing of elevation
long-term projection; e.g., lO-year excess
8. Deficit short-term: extent: hypobaric index
duration & timing of deficit
long-term projection; e.g., lO-year deficit

Results

When one records blood pressure around the clock, one can observe even larger
changes than those that occur from moment to moment both in systolic and diastolic
readings. Figure 4 shows the distribution of blood pressures of a clinically healthy
woman, ranging systolically from 73 to 153 mmHg and diastolically from 51 to 91
mmHg. The overlap between the two distributions is shown as black dots. It may be
foolhardy to disregard circadian changes which are as large as the differences Fig. 3
documents. Healthy individuals can have an average dispersion of 60mmHg in sys-
tolic and 50 mmHg in diastolic pressure. By submitting these data to a computer pro-
gram, one can plot not only the time course of the blood pressure as such, but one
can also derive characteristics of an approximating curve (consisting of a fundamental
such as a 24-h cosine plus a few harmonics). This curve for the individual can then be

Fig. 4. Fully ambulatory automatic monitoring of a clinically apparently healthy woman (EH, age 64
years) reveals a wide range of cardiovascular variability. Questions about the blood pressure (BP) of
a patient are rhetorical. The cost-effective recording and interpretation of variation are mandatory.
Note bimodal rather than bell-shaped distribution of systolic and diastolic BP of the subject and the
large overlap between the two distributions. Hourly averages of data mostly at -10 min intervals, with
interruptions, for 26 days; M, MESOR; A, circadian amplitude; SBP, systolic BP (double dashed
line); DBP, diastolic BP (single dashed line); PP, average pulse pressure (PP = SBP.M - DBP.M)
Blood Pressure and Chronobiology 151

SYSTOLIC MESOR
35

30

25

20

15
II
•
"
""..
1111
IIIIH
·• •"
II
II II U
II II II
1111" II
II

II

" "
111111111111 II II II II II II U
10 II It II 111111 II II If II II II II II n
II II II II II II H II II II I II 11 II II tI " II II II
H 11111111111111 II 1111 II II II \I II II II II II
1111111111111111 1111111111 It II II II II II H II II II
II II U II II II II U II tI It II " II II II II II II II II II ,. II
II II II II II
~I::I:::I:: II II II II II I c::n:::~n:::II::
II II U II If II II II M II II U II II I IIIIIIIINIIIIUI U 111111
" n II II II II U 1\ II 1111111111111 II II II II II n II II II " II \I II II
II II II II
II II 1111 II II II II '* II II II II II II I II II II II II II II II II II II ... II
111111111111 IInIlIIlUIIIIII.IIUIIIIIIH"
II " II • II D II If II

DIASTOLIC MESOR
35

30

I
25
I:
til I
Q)
::J
co> 20 I
I
I
I I
I II
I II
I I I
'0 15
I
I
I I
I I
I I I
Qj I
I
I I
I I
.0 I I I
E 10 I
I
I
I
I I
I I
::J I I I I
Z I
I I
I
I
I I
I I
I
I
I I I I I I
I I I I I I
I I I I I I
I I I I I I

,,
I I I I I I I
I I I I I I I

:(
I
pp . ,, ~
I

I' . ~ 2A If I "I
35
OSP I~ "-./ '--./' SSP 11

30

I
I I
25 I I I
I I
I I
I I I I
I I I
I II I
20 I I I I

··
•"
·
I I I I
I I I II

15
I
I
I I I
II I
I I
I I
, •
•• ""•
.
"" • •
I I I II I I I
I I I I III I I
,
•n ," " "" ," • ••nn
I I I I I I I I I
I I I 11 I I I I II .... II
II II H
•
I I 111111
10 I I I I I
,
I I I 111111 "

" '""
I I I I I I I I I I II II 11 .. 1\ U \I II U II II

n"
I I I I I I I I H II H II II II II U II II II II
I I I

•g :: c : : ••
I I II II II II II H II .. II II II I • 111111 II II II If
I I I I I I I I I
II
I I
I I
I
I
I
I
I
I
I I I I I
I I II I
I I I I I
I I
1IIIHInOHII
H II II tI II II
I I I 111111111 H II
I I I J II II It II II 11
II II II II
II II II • II
II II II H ..
•• ,n.1I1
I
•• 111111

II H H II II
II II II II
1111111
H It II II II II II
II 11 • II II II II
I I I I I I I I I I I
I I I I I I I I I I I I 1 I I II II II II II II It II • It II II II II 11 II II " " • • IIUI"
I I I I I ' • • • • DUU " H If II H II I II II II II II II II II 0 II II • n
I I I I I I I 11111 • • • 1111111111
I I I I I I I I •• 1.1 • • • • • 1111 II II II II • • U • II • II

BLOOD PRESSURE ( ... Hg)

152 F. Halberg et al.

compared with upper and lower limits of a standard profile done for a large peer
group [20]. If the individual plot strays outside of the limits of the peer group at a
given point in 24 h, this may be termed an excess as defined for that person with re-
spect to the peer group. If, for example, the patient's plot is above the upper limit of
the peer group for, say, 2h by lOmmHg, then the area between the curve and the
limit represents the pressure x time excess. To get the area accurately, it has to be
integrated by computer. Therefore, even if the person is normotensive for 22 h of the
day, there will be a 2-h x lO mmHg excess day after day. Over time, this excess could
explain possible increases in morbidity and mortality. Replotting this curve after
taking corrective action (e.g., drugs, sodium restriction or addition, etc.) will ade-
quately show the results of the therapy, as compared to sole reliance upon measure-
ments at office visits.
Based on the studies done using the chronobiologic approach to blood pressure
recording and analysis, it has become clear that many people are being treated for
high blood pressure who need not be [25] and that there are probably many others
who should receive treatment but whose blood pressure elevation has not been
detected.

False-positive Diagnoses

That the traditional approach is unsatisfactory can be suggested on the basis of a

scholarly analysis of data published by Wilcox, Mitchell and Hampton [46]. These
authors raise the suggestion whether clinical practice should be based on the results
of clinical trials. On the basis of the results from these same clinical trials, Potschke-
Langer and Apfelbach instead advocate the use of a "smorgasbord" of drugs [33].
The chronobiologically unacceptable design of the reviewed studies may be con-
sidered.
The outcome of large-scale clinical trials such as the Australian Therapeutic
Trial in Mild Hypertension reflects the situation best. As shown in Fig.5, out of
nearly 2000 individuals in this study who received a placebo, almost half responded
to this treatment, most of them within less than a year [25]. A high percentage of
responders to placebo renders it likely that "office hypertension" leads to the entry
into large-scale studies of many false positive cases. This point is also apparent from
Fig. 6.
By the same token, the exact numbers of those treated with placebo and "con-
verting" from "hypertension" to "normotension" must not be taken seriously. A sub-
stantial error in diagnosis may well be made in clinical trials and otherwise, not only
at the outset of a study in terms of many false positives entering the study, but also
at the final evaluation in which the ratio of "cured normotensives" is likely to be
complicated by a substantial number of false-negative diagnoses. The outcome is
somewhat less ambiguous in those studies relying on endpoints such as mortality and
incidence of heart attack, stroke, etc., that are directly related to actual cardiovascu-
lar disease, irrespective of any blood pressure measurement; but in these cases as
well, the uncertainty associated with the diagnosis at the entry into the study may
greatly complicate or even invalidate the outcome.
Blood Pressure and Chronobiology 153

AUSTRALIAN THERAPEUTIC TRIAL

IN MILD HYPERTENSION CONTROL GROUP: N=I943

START
(Screenlngl
RANGE OF OSP ImmHgI 95-109 I Mean y =1021
AFTER 3 YEARS OF PLACEBO TRE'ATMENT

END
(3 Years Laterl
Is%1112%li32%i r---48%
1) 2) 3) 4)
Fig. 5. 1) Ischemic coronary heart disease or stroke; 2) > 109DBP (mmHg); 3) 95-109DBP (mmHg);
4)<95 DBP (mmHg) y = 9l.
The large proportion in clinical trials of "placebo responders" may be explained in part by entering
individuals with "white coat" hypertension, i.e., by false positive diagnoses, relying on one or a few
casual blood-pressure measurements. In a multi (60) million dollar clinical trial, 48% "cures" from
placebo may be: in part placebo responders; in part false responders to start with, and; in part false
negatives at the end

False-negative Diagnoses

In Fig. 7, a patient's systolic blood pressures (SBPs), shown as dots, are mostly with-
in the reference interval of peers during midday (when he is usually examined by his
physician). At other times, however, notably at night, many of his SBPs are above
the upper chronobiologic reference limit. The treatment of this patient was assumed
to be satisfactory on a casual midday measurement basis. It is not, however, ade-
quate in a chronobiologic perspective. Fig. 7 shows why the same misdiagnosis
applies to the same patient's diastolic blood pressures.

Therapy

Both the timing of treatment and the assessment of treatment effects are a
chronobiologic matter. Figure 8 exemplifies a successful treatment insofar as it is
judged by the reduction to zero of both the circadian percent time elevation of blood
pressure and the circadian hyperbaric index. The upper graph in Fig. 8 (circadian
percent-time elevation) shows that when the patient was off treatment, even though
his means were below conventional limits, he was actually above a chronobiologic
limit 49% of the time for systolic and 10% for diastolic blood pressure. This percent-
age time was reduced by treatment to small, negligible values and to zero thereafter.
 154 F.Halberg et al.

180 (SBP} 140 (DBP)

165

110

90 50

75 35

00:00 04:00 08:00 12:00 16:00 20:00 00:00 00:00 04:00 08:00 12:00 16:00 20:00 00:00

TIME (CLOCK HOUR)

Fig. 6. False-positive diagnosis. The figure shows single measurements (as dots) of ambulatory sys-
tolic (S) blood pressure (BP) (left) and of diastolic (D) BP (right) from five subjects who, on three
different office visits, had a standing DBP higher than 95 mmHg and were labeled "borderline hyper-
tensive". Hourly mean values (dashed lines) drawn for each subject reveal that these individuals' BP
are mostly below the solid line representing the upper time-varying 95% prediction limit computed
from the data of 39 clinically healthy individuals, used as a peer group reference range. In the ab-
sence of around-the-clock monitoring, such individuals could have been (mis)treated for long spans,
perhaps for a lifetime. In these cases, failure to comply with treatment is desirable for the person in-
volved. It should be noted that some individual values are around 170 mmHg, whereas others are
around 90mmHg. Note different vertical scales

In this case, then, a snapshot in mmHg was replaced by the estimate of an index
of excess expressed in mmHg x h. Future work will have to provide for respite and
a nonlinear accumulation of harm without thresholds by postulating the existence of
an increasing harm function of blood pressure. Such a harm function should estimate
the pressure-related cardiovascular disease risk by a pressure-related health index
(PRHI), time-qualified as discussed elsewhere [13, 25]. The step from mmHg to
mmHg x h, from the use of a fixed to that of variable threshold, and the use of pa-
rameter tests to assess the statistical significance of a change in indices may represent
a step toward reducing harm associated with diseases related to a high blood pressure
[12, 17].

Prevention

Chronobiology emphasizes prevention and seeks the tools to recognize risk as early
as birth. Sensitive indices of blood pressure dynamics can actually be used to probe
Blood Pressure and Chronobiology 155

230 SSP

1 SE of S II~ Hou1y Means, S

210 I Individual RaadlnCjJs (n:922)

190

170
;(
150 " .
.~-'-'--'".,.. M••n =138
130

90

01.......m£~=c~~==c=.~=c=r1~.
00:00 06:00 12:00 18:00 00:00

122 DSP
1SE of O(~ Hourly means. 0
_/.Individual ReadlnQs (0=925)
112 r) .
102

92

82
Fig. 7. High blood pressures (BP) at
odd times contrast with mostly ac-
72 ceptable BPs around noon in a pa-
tient believed to be well-treated on
the basis of a casual examination at
62
midday. Systolic BPs shown on top
(SBP); diastolic BPs on the bottom
52 (DBP). Based on data at 7.5 min
intervals for 24h from 39 clinically
healthy men, 23-59 years of age.
O~"~gc~~~~~~~ Note different scales
00:00 06:00 12:00 18:00 00:00
Time (Clock Hour)

the genetic aspects of high blood pressure. More specifically, without any observed
side effect, the blood pressure and heart rate of 144 newborns were automatically
monitored at 30-min intervals for 48 h, starting within the first few hours or days of
life. On the basis of questionnaires given to the parents, the newborns were assigned
to either a negative or positive family history group according to the absence or pres-
ence of high blood pressure and/or related cardiovascular disease in the family. The
data series from 94 babies at term were grouped by family history (negative or posi-
156 F. Halberg et al.

Percent Ti me Elevation Hyperbar i c Index

Systolic
24 .c

~
x
01
:x:
E
...
o E u
x
-'" UJ

u .____~____~o~__~o ~o
..
~
::>

Dia stolic ~
~
~~ Diast olic lo2~_~o :2 o
~

~ ::Lt__ L-_O.;......_~O ~
1. 3 0

l.0 ______ ____..:O__...J.

Rx-: No Yes Yes Yes Yes No Yes Yes Yes Yes

Mon itoring Span ldays) :
8 6 2 5 10 8 6 2 5 10
Calendar Month and Year:
Dec.81-Jan.82 JuneS3 June& Dec.86-Jan.S7 Dec.81-Jan.82 JuneS3 June& Dec.86-Jan.S7

Fig. 8. Nonparametric blood pressure assessment. The question may be debated as to whether one
should treat, by drugs, a systolic and a diastolic blood pressure (BP) of 128 and 82 mmHg, respec-
tively. In this case , it was treated and the BPs were reduced by treatment at an appropriate time. Off
treatment , the BPs were above peer-group limits 49% (SBP) or 10% (DBP) of the time, after treat-
ment excess was eventually eliminated (left). The corresponding reduction and eventual disappear-
ance of blood pressure excess, measured in mmHg x h, denoted as "hyperbaric index," is also
shown (right). * Chlorothiazide 25 mg by mouth, mornings

tive) for comparison. The chronobiologic endpoint, the circadian amplitude, was in-
variably the most sensitive parameter for blood pressure . For heart rate, the MESOR
appears superior. A circadian population rhythm characterizes systolic blood pres-
sure (P = 0.009), mean arterial pressure (P = 0.014) and diastolic blood pressure
(P = 0.011) of the newborns with a positive family history of high blood pressure but
not of those with a negative family history (P> 0.2) . For heart rate, a population
rhythm is apparent in both risk groups. Individually, the rhythm is demonstrable in
only about 35% of the newborns.
The effect of covariates such as sex, body mass index (BMI = weight/height 2) at
birth and age at start of monitoring was further examined. Linear regression of these
variables as well as maternal and gestational age on dynamic indices of blood pres-
sure and heart rate variability was also performed.
Dispersion indices of heart rate increase as a function of the newborn's age at
start of monitoring (P < 0.05); there is little if any effect of BMI on heart rate; new-
borns in the positive family-history group have a lower heart rate MESOR than
those in the negative family history group; girls have a larger heart rate MESOR as
compared to boys (P<0.05). A gender effect is not seen for blood pressure. Loca-
tion and dispersion indices for newborns with a positive family history are larger than
those of newborns with a negative family history, and contrary to expectation from
studies in older children, location and dispersion indices are smaller in babies with a
larger BMI (P < 0.05) .
Blood Pressure and Chronobiology 157

Table 2. Circadian amplitude (A) of diastolic blood pressure (BP) (automatically monitored every
15 min for 48 h on 15 boys and 15 girls with differing family history of high BP), but not mean value,
correlates with echocardiographically determined thickness of the interventricular septum

Kind of Index All children Children with PFH a Does the

BPindex ofhighBP index
r Pvalue work?
r Pvalue

Chronobiologic Circadian A 0.385 0.045 0.672 0.016 Yes

Conventional Mean value -0.039 0.840 0.059 0.848 No

a PFH = Positive Family History [from 4].

A breakdown of family history in terms of paternal or maternal family history of

high blood pressure reveals a stronger influence of the maternal family history of
high blood pressure on the circadian blood-pressure !lmplitude of the newborn. The
consideration of a family history of diabetes and of pregnancies complicated by an
elevated blood pressure and/or the taking of beta-adrenergic drugs and/or corticoids
shows alteration of the circadian amplitude. These results suggest the possibility of
intervention in utero, for instance by dietary means in pregnancy.
Serial blood pressure measurements provide sensitive indices that focus on blood
pressure variability. These features of the time structure evolve rapidly following
birth. On the basis of around-the-clock rhythmometry, our work presents several in-
dications of the sensitivity of the circadian blood pressure amplitude not only as a
harbinger of cardiovascular disease risk but also as a correlate of target organ in-
volvement. In healthy 14-year-olds, we find that the circadian amplitude, but not the
24-h mean of diastolic blood pressure, correlates with the interventricular septum
thickness (Table 2).

Discussion

The uncertainty of an elevated blood pressure is usually discussed on the assumption

that blood-pressure variability is mainly due to measurement error, to the patient's
posture and to responses to external stimuli. This view disregards the fact that part
of the variation in blood pressure is spontaneous and accounted for by rhythms of
several frequencies, among which the 24-h (or circadian) rhythms are most promi-
nent. In order to assess the variation in blood pressure, repeated measurements
should be taken, preferably at strategically selected times so as to quantify and
exploit more reliable mean values and novel dynamic endpoints. Although presently
available automatic recorders greatly facilitate data collection, the analytical
methodology available as computer software for resolving the characteristics of
rhythm can also be applied to measurements taken manually, once a proper sampling
schedule is adopted. The methodology offered by chronobiology allows the assess-
ment of the extent and timing of rhythmic change and the establishment of time-
varying norms for healthy peer groups and/or individuals. By comparing these
dynamic indices of blood-pressure variation of a given individual with those of peers,
158 F. Halberg et al.

one seeks to recognize the risk of deviant behavior (e.g., of blood pressure), both
early and in a time-dependent fashion to prompt preventive action. When such
methods are applied to MESOR- or amplitude-hypertensive patients, they serve to
assess their response to treatment.
The sphygmochron presents, illustrates and interprets the circadian rhythm char-
acteristics of blood pressure and heart rate. When many measurements are taken
around the clock, one likely finds values above or below reference limits, not only
the fixed limits recommended by WHO, but also outside chronobiologic limits, the
so-called chronodesms. In addition to providing an estimate of such deviant single
values, the sphygmochron also addresses new and more pertinent questions such as
"How long does blood pressure remain above a given time-varying threshold?"; "To
what extent does the blood pressure exceed the threshold at a given time?"; "What
is the total excess and its projection over a lifetime?" and "When does most of the
excess occur?" On this rational basis, treatment can be timed more appropriately
and its effects can be objectively monitored and tested statistically, not only on a
popUlation basis, but also for the given individual. In summary, characteristics of
rhythms and of blood pressure excess (or deficit) serve several purposes:
1. To initiate treatment if needed
2. To time treatment when it is most effective and least harmful in terms of un-
desired effects
3. To help in prognosis and diagnosis with a better assessment of health status
4. To screen for early recognition of risk before disease sets in, so as to intervene by
preventive measures such as exercise, diet or other nondrug treatment.
We are at the threshold of a constructive evolution in medicine, in particular with
respect to blood pressure and heart rate. We recognize variability, not as a foe to
medical diagnosis and treatment, but as the source of information to be extracted by
special hardware and software for use in a refined screening and prevention as well
as diagnosis and cure. Automatic recorders are available at a cost that will decrease
as their use increases. Algorithms analyze dense and long records for dynamic char-
acteristics. Accumulating evidence from the use of modern hardware and software
reveals the health effects of rhythmic changes and some of their mechanisms.
Rhythm characteristics include a better mean, the MESOR, than that of one or a
few casual measurements. Additional rhythm and trend characteristics complement
traditional means, ranges and standard deviations, notably to detect at birth risk of
developing civilization diseases later in life.
We believe that the chronobiologic method of blood pressure analysis will lead to
more accurate and successful diagnosis and treatment. If large numbers of today's
false positives can be removed from therapy and if a substantial percentage of today's
false negatives can receive the appropriate treatment, the reduction in morbidity and
mortality will be considerable. Getting people into the right treatment category will
clearly cut health costs while improving health care for millions.
The chronobiologic component in blood pressure reflects but one end result of
the body's coordinating systems that link all the "feedsideward" mechanisms through-
out the body into one harmonious symphony. The intermodulating rhythms occur at
all levels from the single cell to the whole organism. The synchronization of these
changes is done via the central and autonomous nervous system and is influenced
Blood Pressure and Chronobiology 159

by the social regimen. Our getting up and going to bed at the times set by society
each day reset our free-running internal rhythm setters into step with worldtime.
The brain (or mind) is therefore a vital link in the time region's coordination of the
heart, the rest of the cardiovascular system and the whole person. Looking at all bio-
physical variables with a chronobiologic viewpoint will greatly enhance diagnosis,
therapy and prevention in nearly all facets of medicine, but the application to blood
pressure can be used already now with great service to the quality of life for many pa-
tients.
The chronobiologic view is that circadian and other rhythms are endogenous
components that are synchronized by physicochemical as well as socioecologic factors.
The existence of an underlying intrinsic rhythmic mechanism is supported by the ob-
servation that blood pressure rises before awakening and cannot thus result only
from a change in posture or activity [7, 25, 43]. It was also shown that the circadian
amplitude in blood pressure decreases only slightly under conditions of bedrest [25,
35]. Halberg in 1953 [7a] documented the phenomenon of an activation of both the
adrenal cortex and medulla, codetermining a rise in blood pressure prior to awakening.
The adrenal cortex is critical in the interdigitating networks coordinating meta-
bolic and adjusting networks as well as complementary networks for growth, repair
and reproduction. The adrenal cortical cycle has been described as the pacemaker of
circadian rhythms in circulating blood eosinophil cells and to act at the cellular level
upon rhythms in phospholipid labelling and mitosis [34]. The adrenal cycle was
shown to persist in the histologically validated absence of the suprachiasmatic nuclei
[40].
The persistence of adrenal cortical bioperiodicity in rodents was documented
directly by corticosterone determinations after stepwise brain ablation [5] and in
vitro [1]. Indirectly, on human beings, the persistence of circadian adrenal periodicity
has also been documented by counts of circulating blood eosinophils after hemi-
spherectomy [15], during regressive electroshock [8] and coma [30].
Robel and Baulieu have shown that, apart from the long-known contribution of
major steroidogenic glands, dehydroepiandrosterone and its ester sulfate are also made
in the brain [37]. These authors also showed that dehydroepiandrosterone undergoes
a circadian cycle in the brain, and that following adrenalectomy and orchidectomy in
rats, it not only continues to be found, but continues to exhibit a circadian rhythm
in the brain [36]. These results support the hypothesis for the existence of an auto-
nomous cerebral mechanism coordinating the production of ,15-3fJ-hydroxysteroids.
It is also pertinent to note that the acrophase of pregnenolone in brain precedes, with
statistical significance, the acrophase of corticosterone in rats [45], as does the acro-
phase of dehydroepiandrosterone in plasma. Moreover, in the rat, the circadian
rhythm of corticosterone, pregnenolone and dehydroepiandrosterone are in phase
between their variation in plasma and brain. This is not the case for the adrenal
phase relations in humans, the rise of cortisol preceding that of dehydroepiandros-
teron-sulfate by about 6h [17]. The differences in phase relations between the two
kinds of steroids observed between the two species (rodents and human beings) de-
scribe internal phase relations and are therefore independent of the difference in the
rest/activity cycle between the two species. These differences may result in part from
a modulatory effect exerted by the eNS. The pineal gland is a logical candidate to
intervene in such a coordinating mechanism by means of its feedsideward action [34].
160 F. Halberg et al.

Summary and Conclusions

The diagnosis and treatment of deviant blood pressure can be undertaken with greater
accuracy and higher success if one uses the "dynamic" characteristics of the pressure
rather than just the present "static" viewpoint based primarily on an average pres-
sure. In lieu of sole reliance upon repeated casual measurements, as recommended
by WHO, and also in lieu of the inspection only of records from automatic 24-h
monitoring, we advocate a third alternative, namely a chronobiologic exploitation of
data collected systematically. These can be obtained by traditional sphygmomano-
meters or by means of automatic monitoring devices. Modern ambulatory monitors
become cost-effective once the merit of chronobiologic endpoints, such as a measure
of excess or deficit, each assessed in relation to time-varying limits derived from peer
groups, is recognized. The information obtained by additional endpoints describing
the extent and timing of change, all presented in a computer-prepared profile over
time, the sphygmochron, is documented by illustrative examples and by reference
to a more detailed presentation and bibliography. Indications for the use of the
chronobiologic approach to blood pressure evaluation are also succinctly presented.
By these examples from the field of blood pressure evaluation, we have tried to
document the need for awareness of the broader scope of the science of physiologic
variation in relation to time: chronobiology, for medical practitioners as well as re-
searchers.

Acknowledgements. The writing of this paper and the work summarized therein were supported by
the U.S. National Institutes of Health, in particular GM-13981; the Fondazione Hoechst, Milan,
Italy; Medtronic Inc., Minneapolis, Minnesota, USA; Colin Medical Instruments, Komaki, Japan;
the Minnesota Medical Foundation and, with special thanks due, the Dr. Betty Sullivan Fund.
Michael Moore (Science Writer, University of Minnesota) kindly edited part of this manuscript.

References

1. Andrews RV (1968) Temporal secretory responses of cultured hamster adrenals. Comp Bio-
chern Physiol 26: 179-193
2. Carandente F, Halberg F (eds) (1984) Chronobiology of blood pressure in 1985. Chronobiologia
11(3)
3. Cornelissen G (1987) Instrumentation and data analysis methods needed for blood pressure
monitoring in chronobiology. In: Scheving LE, Halberg F, Ehret CF (eds) Chronobiotechnol-
ogy and chronobiological engineering. Nijhoff, Dordrecht, pp 241-261
4. Croppi E, Livi R, Scarpelli L, de Leonardis V, de Scalzi M, Halberg F, Romano S, Cagnoni M,
Scarpelli PT (1987) Chronobiologically assessed blood pressure and left ventricular wall thick-
ness in children with and without a family history of a high blood pressure. In: Tarquini B, Ver-
gassola R (eds) Social diseases and chronobiology. Proc III int symp social diseases and chrono-
biology, Florence, Nov 29, 1986. Societa Editrice Esculapio, Bologna, pp 135-144
5. Galicich JH, Halberg F, French LA, Ungar F (1965) Effect of cerebral ablation on a circadian
pituitary adrenocorticotropic rhythm in C mice. Endocrinology 76: 895-901
6. Gross F, Strasser T (eds) (1987) Mild hypertension: recent advances. Raven, New York
7. Halberg E, Halberg F, Shankaraiah K (1981) Plexo-serial linear-nonlinear rhythmometry of
blood pressure, pulse and motor activity by a couple in their sixties. Chronobiologia 8: 351-366
7a. Halberg F (1953) Some physiological and clinical aspects of 24-hour periodicity. Lancet 73:
20-32
Blood Pressure and Chronobiology 161

8. Halberg F (1960) Temporal coordination of physiologic function. Cold Spring Harbor Symp
Quant Bioi 25: 289-310
9. Halberg F (1969) Chronobiology. Annu Rev Physiol3l :675-725
10. Halberg F (1981) Biologic rhythms, hormones and aging. In: Vernadakis A, Timiras PS (eds)
Hormones in development and aging. Spectrum, New York, pp 451-476
11. Halberg F (1983) Quo vadis basic and clinical chronobiology: promise for health maintenance.
Am J Anat 168: 543-594
12. Halberg F (1987) Perspectives of chronobiologic engineering. In: Scheving LE, Halberg F,
Ehret CF (eds) Chronobiotechnology and chronobiological engineering. Nijhoff, Dordrecht,
pp 1-46
13. Halberg F, Bingham C (1987) The scope and promise of chronobiology and biostatistics: inter-
penetrating, inseparable disciplines. Proc Biopharmaceutical Section, American Statistical As-
sociation, Chicago, Aug 15-18, 1986, pp 11-32
14. Halberg F, Montalbetti N (1983) Chronobiologic specification of reference values. Bull Molec
Bioi Med 8:75-103
15. Halberg F, Barnum CP, Silber RH, Bittner JJ (1958) 24-hour rhythms at several levels of integra-
tion in mice on different lighting regimens. Proc Soc Exp Bioi Med 97: 897-900
16. Halberg F, Good RA, Levine H (1966) Some aspects of the cardiovascular and renal circadian
system. Circulation 34: 715-717
17. Halberg F, Cornelissen G, Sothern RB, Wallach LA, Halberg E, Ahlgren A, Kuzel M, Radke
A, Barbosa J, Goetz F, Buckley J, Mandel J, Schuman L, Haus E, Lakatua D, Sackett L, Berg
H, Wendt HW, Kawasaki T, Ueno M, Uezono K, Matsuoka M, Omae T, Tarquini B, Cagnoni
M, Garcia Sainz M, Perez Vega E, Wilson D, Griffiths K, Donati L, Tatti P, Vasta M, Locatelli
I, Camagna A, Lauro R, Tritsch G, Wetterberg L (1981) International geographic studies of
oncological interest on chronobiological variables. In: Kaiser H (ed) Neoplasms - comparative
pathology of growth in animals, plants and man. Williams and Wilkins, Baltimore, pp 553-596
18. Halberg F, Sanchez de la Peiia S, Fernandes G (1983) Immunochronopharmacology. In: Hadden
J, Chedid L, Dukor P, Spreafico F, Willoughby D (eds) Advances in immunopharmacology.
Pergamon, Oxford, pp 463-478
19. Halberg F, Ahlgren A, Haus E (1984) Circadian systolic and diastolic hyperbaric indices of high
school and college students. Chronobiologia 11: 299-309
20. Halberg F, Drayer JIM, Cornelissen G, Weber MA (1984) Cardiovascular reference data base
for recognizing circadian MESOR- and amplitude-hypertension in apparently healthy men.
Chronobiologia 11: 275-298
21. Halberg F, Carandente F, Montalbetti N (1986) The sphygmochron, a cardiovascular summary,
illustrates rhythmometry for advanced technology. Biochim Clin 10: 939-942
22. Halberg F, Cornelissen G, Bingham C, Tarquini B, Mainardi G, Cagnoni M, Panero C, Scar-
pelli P, Romano S, Miirz W, Hellbrtigge T, Shinoda M, Kawabata Y (1986) Neonatal monitor-
ing to assess risk for hypertension. Postgrad Med 79: 44-46
23. Halberg F, Kausz E, Winter Y, Wu J, Miirz W, Cornelissen G (1986) Circadian rhythmic re-
sponse in cold pressor test. J Minn Acad Sci 51: 14
24. Halberg F, Reale L, Tarquini B (eds) (1986) Proc 2nd Int Conf Medico-Social Aspects of Chrono-
biology. Florence, Oct 2,1984. Istituto Italiano di Medicina Sociale, Rome
25. Halberg F, Cornelissen G, Halberg E, Halberg J, Delmore P, Bakken E (1987) Chronobiology
of human blood pressure. Medtronic Continuing Medical Education Seminars
26. Hanson BR, Halberg F, Tuna N, Bouchard TJ Jr, Lykken DT, Cornelissen G, Heston LL
(1984) Rhythmometry reveals heritability of circadian characteristics of heart rate of human
twins reared apart. Cardiologia 29:267-282
27. Hermida RC, Halberg F, Halberg E (1986) Closer to a psychoneuroendocrine hemopsy? Bio-
chim Clin 10: 1053-1066
28. Kanabrocki EL, Scheving LE, Halberg F, Brewer RL, Bird TJ (1973) Circadian variations in
presumably healthy men under conditions of peace-time army reserve unit training. Space Life
Sci 4: 258-270
29. Kanabrocki EL, Scheving LE, Halberg F, Brewer RL, Bird TJ (1974) Circadian variation in pre-
sumably healthy young soldiers. Department of the Army, Document # PB 228427, 56
30. Levine H, Ramshaw WA, Halberg F (1967) Least-squares spectral analyses on core temperature
and blood pressure of a comatose girl. Physiologist 10: 230
162 F. Halberg et al.: Blood Pressure and Chronobiology

31. Mainardi G, Maggioni C, Halberg F, Hurley P, Panero C, Tarquini B, Cariddi A, Sorice V, Cor-
nelissen G, Bingham C, Scarpelli P, Scarpelli L, Livi R, Romano S, Cagnoni M (1986) Family
history of high blood pressure (BP) and diabetes and newborn's circadian BP amplitude. In:
Tarquini B, Vergassola R (eds) Proc III Int Symp Social Diseases and Chronobiology. Florence,
Nov 29,1986, pp 64-65
32. Panero C, Maggioni C, Halberg F, Hurley P, Mainardi G, Tarquini B, Cariddi A, Sorice V,
Cornelissen G, Bingham C, Scarpelli P, Scarpelli L, Livi R, Romano S, Cagnoni M (1986) Beta-
adrenergics in pregnancy and maternal high blood pressure (BP) increase newborn's circadian
BP amplitude. In: Tarquini B, Vergassola R (eds) Proc III Int Symp Social Diseases and Chrono-
biology. Florence, Nov 29, 1986, pp 67-68
33. Potschke-Langer M, Apfelbach J (1986) Milde Hypertonie - auf dem Wege zu einem inter-
nationalen und nationalen Konsens. Fortschr Med 44: 852-856
34. Redfern PH, Campbell I, Xavier JA, Martin KF (eds) (1985) Circadian rhythms in the central
nervous system. Macmillan, London
35. Reinberg A, Ghata J, Halberg F, Gervais P, Abulker Ch, Dupont J, Gaudeau C (1970) Rythmes
circadiens du pouls, de la pression arterielle, des excretions urinaires en 17-hydroxycortico-
steroides, catecholamines et potassium chez I'homme adulte sain, actif et au repos. Ann Endo-
crinol (Paris) 31 :277-287
36. Robel P, Synguelakis M, Halberg F, Baulieu EE (1986) Neurophysiologie. - Persistance d'un
rythme circadien de la dehydroepiandrosterone dans Ie cerveau, mais non dans Ie plasma, de
rats castres et surrenalectomises. C R Acad Sci 303 [III] 6: 235-238
37. Robel P, Bourreau E, Corpechot C (1987) Neuro-steroids: 3.B-hydroxy-L15-derivatives in rat and
monkey brain. Proc 7th Int Cong Hormonal Steroids, Madrid. J Ster Biochem 27: 649-655
38. Roberts (1987) Measuring cholesterol is as tricky as lowering it. Science 238: 482-483
39. Scheving LE, Halberg F (eds) (1980) Chronobiology: principles and applications to shifts in
schedules. Sijthoff and Noordhoff, Alphen aan den Rijn
40. Scheving LE, Tsai TS, Powell EW, Pasley IN, Halberg F, Dunn J (1983) Bilateral lesions of
suprachiasmatic nuclei affect circadian rhythms in [3H]thymidine incorporation into deoxy-
ribonucleic acid in mouse intestinal tract, mitotic index of corneal epithelium, and serum cor-
ticosterone. Anat Rec 205 : 239-249
41. Schweiger H-G, Berger S, Kretschmer H, Morler H, Halberg E, Sothern RB, Halberg F (1986)
Evidence for a circaseptan and a circasemiseptan growth response to light/dark cycle shifts in
nucleated and enucleated Acetabularia cells, respectively. Proc Natl Acad Sci USA 83: 8619-
8623
42. Smolensky M, Halberg F, Sargent F II (1972) Chronobiology of the life sequence. In: Itoh S,
Ogata K, Yoshimura H (eds) Advances in climatic physiology. Igaku Shoin, Tokyo, pp 281-318
43. Stadick A, Bryans R, Halberg E, Halberg F (1987) Circadian cardiovascular rhythms during
recumbency. In: Tarquini B, Vergassola R (eds) Social diseases and chronobiology. Proc III Int
Symp Social Diseases and Chronobiology. Florence, Nov 29, 1986, Societa Editrice Esculapio,
Bologna, pp 191-200
44. Strasser T, Ganten D (eds) (1987) Mild hypertension: from drug trials to practice. Raven,
New York
45. Synguelakis M, Halberg F, Baulieu EE, Robel P (1985) Evolution circadienne de L15-3.B-hydroxy-
steroides et de glucocorticosteroides dans Ie plasma et Ie cerveau de Rat. C R Acad Sci 301 :
823-826
46. Wilcox RG, Mitchell JRA, Hampton JR (1986) Treatment of high blood pressure: should clini-
cal practice be based on results of clinical trials? Br Med J 296: 433-437
47. 1986 guidelines for the treatment of mild hypertension. memorandum from a WHO/ISH meet-
ing. J Hypertens 4:383-386
Part III
Syncope and Sudden Death
Chapter 12
Mechanisms of Syncope and of Sudden Death
Due to Ventricular Tachyarrhythmias
G. BREITHARDT, M. BORGGREFE, A. PODCZECK, and A. MARTINEZ-RuBIO

Introduction
Syncope (defined as a sudden and transient loss or impairment of consciousness due
to changes in cerebral function produced by circulatory, metabolic, or neuropsycho-
logical mechanisms) and sudden cardiac death are two alarming events that may
have similar underlying mechanisms such as short- or long-lasting episodes of
bradycardia or tachycardia. In addition, syncope may be due to innocent disturbances
of circulatory regulation (such as orthostatic dysregulation, hypervagotonia, etc.)
that may lead to transient reduction in cerebral blood flow and, thus, to loss of con-
sciousness. The various causes of syncope as well as of the underlying cardiac dis-
orders in which syncope is common are listed in Tables 1 and 2.
The evaluation of syncope is a common problem in patients admitted to the hos-
pital. Although noninvasive medical and neurological investigation may clarify the
cause of syncope in some patients, the symptoms still remain unexplained in a large
proportion of patients. In a recent prospective study [42] in 204 patients with syn-
cope, the cause remained unknown in 97 patients (47.5%). Furthermore, long-term
electrocardiographic monitoring provided a probable cause of syncope in only 30%-
60% of cases [70, 71].

Table 1. Arrhythmic causes of syncope

Bradycardias
1. Sinus node disease
- extreme sinus bradycardia
- sinus standstill
- prolonged sinus node recovery time
2. A V conduction disturbances
3. Intraventricular conduction disturbances
4. Carotis sinus syndrome (cardioinhibitory type)
Tachycardias
1. Supraventricular
- atrial fibrillation, flutter with rapid A V conduction
- atrial fibrillation in WPW syndrome
2. Ventricular
- ventricular tachycardia
- ventricular fibrillation
- QT syndrome
166 G. Breithardt et al.

Table 2. Underlying cardiac disorders in which syncope is

not uncommon

Aortic stenosis
HOCM
Mitral stenosis
Tetralogy of Fallot
Pulmonary hypertension
Severe pulmonary stenosis
Coronary artery disease
Atrial myxoma
Intramyocardial tumors
Myocarditis
QTsyndrome
Keams-Sayre syndrome
Refsum's disease

Though syncope due to circulatory dysregulation is more frequent than syncope

due to bradycardia or tachycardia, and fortunately more benign than the latter
onses, it is the occurrence of brady- or tachyarrhythmias that is responsible for the
sometimes grave prognosis of syncope. Recent evidence has shown that in patients
with syncope the presence of organic heart disease is the major factor that deter-
mines whether an underlying tachy- or bradyarrhythmia may be potentially lethal
[22,33,42,52,56,57]. A clinically important situation is found in patients recover-
ing from myocardial infarction in whom both types of events (syncope and sudden
death) are frequently due to ventricular tachyarrhythmias [16].
A suddenly and unexpectedly occurring death need not always be due to cardiac
arrhythmias but may also be due to other types of acute disease such as acute intra-
cerebral hemorrhage. In this context, the definition of sudden death is of great im-
portance. During the last two decades, the definition of sudden cardiac death with
regard to the interval between the onset of symptoms and death has received con-
siderable attention. The shorter this interval, the greater the probability of sudden
death being due to an arrhythmic event [47]. Goldstein et at. [31] considered 96.5%
of all deaths occurring within 1 h to be due to arrhythmia whereas this proportion
dropped to 75% in those with "intermediate" death. In this latter subgroup, 32% of
patients died from shock or congestive heart failure. Thus, the shorter the interval
between onset of symptoms (if at all) and death, the greater the probability that it is
due to some type of arrhythmia. The extreme would be instantaneous death, which
can be attributed to arrhythmia with a high degree of certainty.
The purpose of this chapter is to review the mechanisms of ventricular tachy-
arrhythmias that may lead to syncope and sudden (arrhythmic) death.

Pathophysiology of Sudden Death

In coronary artery disease, sudden cardiac death is one of the dominating modes of
death. Though death may also be sudden in other types of heart disease, the major
Mechanisms of Syncope and Sudden Death 167

Table 3. Mechanisms leading to ventricular tachyarrythmias and sudden cardiac death

Chronic electrophysiological abnormalities

Transient ischemic attacks
spasm (variant angina)
acute thrombotic coronary occlusion
- without previous myocardial infarction
- with previous myocardial infarction
"unstable" coronary artery lesion
(embolism into peripheral coronary artery without/with previous myocardial infarction)
exertional angina
Emotional stress (single factor?)

focus on patients with coronary artery disease is due to the great number of patients
dying suddenly. Clinical observations in patients on long-term ECG recording at the
time of sudden cardiac death have provided evidence that the great majority of these
events are due to some type of ventricular tachyarrhythmia [2, 58].
The mechanisms leading to ventricular tachyarrhythmias have been studied in
animal models and in man [13]. The methods used in these studies included the ex-
perimental assessment of the ventricular fibrillation threshold [35, 36], the repetitive
ventricular response index [46,50], and of models of acute or chronic ischemia [19,
24, 59], as well as clinical studies on the significance of spontaneous ventricular ar-
rhythmias during long-term ECG recording [1, 18, 51, 53, 54, 63], the role of left
ventricular dysfunction [30, 32, 39], detection of ventricular late potentials [3, 6, 7,
9,28,41,43,62,67], the role of inducible ventricular tachyarrhythmias [8,11,14,34,
61, 73], and the influence of the autonomic nervous system [48]. From these studies,
a multitude of possible mechanisms emerged that may underly the occurrence of
ventricular tachyarrhythmias (Table 3).

Primary Versus Secondary Sudden Cardiac Death

Sudden cardiac death may occur as the first manifestation of coronary artery disease
without antecedent complaints. In the context of this paper, this type of event will be
called "primary" sudden death. In other instances, a patient may die suddenly after
a previous myocardial infarction which has left some regional abnormalities of the
electrophysiological properties of the heart constituting the basis for the occurrence
of ventricular tachyarrhythmias. This type of event will be called "secondary" sud-
den death as it is secondary to preexisting damage of the myocardium. It is obvious
that any diagnostic or preventive measure can only be instituted early enough in pa-
tients with secondary cardiac death. Therefore, this differentiation is of major impor-
tance for any strategies for prevention of sudden cardiac death.

Arrhythmogenic Substrate

Experimental and clinical studies have provided evidence that myocardial infarction
may leave a zone of electrically abnormal ventricular myocardium that may be the
168 G. Breithardt et al.

AV 995
Fa 25 VECTOR Of 1 Ff 1. 2. 3
QRS 93- 241 OUR 148 V (4m) 12.97 V (5m) 17.55 V <TOTAL> 142. 2m
2mm
QRS / LP

QRS 93- 186 OUR 93

LP 186- 241 OUR 55
V<LP) 19. 16
V.ox (LP) 43.38
V (QRS) 179.41

Fig.t. Signal-averaged and filtered

recording of leads x, y, z (vector
magnitude) in a patient with ven-
m tricular tachycardia. QRS duration
~ ! r••] ~ in the highly amplified recording was
148ms. The program [67] automati-
cally identified the end of the total
QRS complex at 241 ms on the x axis.
[~Nl
The amplitude in the terminal 40 ms
was low [V(40) = 12.97,N], which
was automatically measured by the
program. Additionally, the onset of
low-amplitude activity was auto-
y matically identified at 186 ms on the
25[ horizontal axis. The program also
"
measured the mean voltage of the
late potential [V(LP) = 19.16 J.lV],
the maximal voltage of the late po-
tential [Vmax(LP) = 43.38 J.lV], and
the mean voltage of the true QRS
complex [V(QRS) = 179.41 J.lV]
gj lSI gj gj
lSI
N (Il
• rill.]
~
56'1

site of origin of ventricular tachycardia. This tissue is mostly located at the border
zone of a previous myocardial infarction and is characterized by islands of relatively
viable muscle alternating with areas of necrosis and later fibrosis. Such tissue may re-
sult in fragmentation of the propagating electromotive forces with the consequent
development of high-frequency components that can be recorded directly from these
areas [27,29,44,60]. Gardner et al. [29] were able to show that experimentally in-
duced slow conduction alone did not cause fragmented activity. Highly fractionated
electrograms could only be recorded in preparations from chronic infarcts with inter-
stitial fibrosis forming insulating boundaries between muscle bundles. The individual
components of fragmented electrograms, therefore, most probably represent asyn-
chronous electrical activity in each of the separate bundles of surviving muscle under
the electrode. The intrinsic asymmetry of cardiac activation due to fiber orientation
(anisotropy) may be accentuated by infarction and may predispose to reentry [29,
60]. The slow activation might result from conduction over circuitous pathways
caused by the separation and distortion of the myocardial fiber bundles. The low am-
plitude of the electrograms from these regions probably results from the paucity of
surviving muscle fibers under the electrode because of the large amounts of connec-
Mechanisms of Syncope and Sudden Death 169

tive tissue and not from depression of the action potentials. Therefore, the anatomic
substrate for reentry seems to be present in regions where fragmented electrograms
can be recorded which, thus, indicate slow inhomogeneous conduction (for a more
detailed discussion, see [7]).
Electrical recordings from these sites have demonstrated that the fractionated
low-amplitude activity may extend beyond normal ventricular activation into the ST
segment of the surface ECG. With conventional methods of ECG recording, these
signals cannot normally be registered on the body surface. Berbari et al. [3] in the ex-
perimental animal, and Fontaine et al. [28] in man, were the first to report that these
signals, subsequently called "late potentials" (Fig. 1), can be recorded from the body
surface by the use of high-gain amplification, appropriate filtering, and computer-
averaging techniques. A review on the clinical and prognostic significance of ven-
tricular late potentials has recently been published [7].

Acute Versus Chronic Arrhythmogenic Tissue

A zone with arrhythmogenic properties may arise acutely (and be present only tran-
siently) or it may exist chronically in the form of myocardium interspersed with fibro-
sis after a previous myocardial infarction (as described above). The classic example
of an acutely developing arrhythmogenic tissue leading to ventricular tachyarrhythmias
is acute myocardial infarction that is frequently accompanied by ventricular fibrilla-
tion. The changes that occur in this situation are frequently transient in nature and
may subside as soon as the tissue is completely necrotic. The electrophysiological ef-
fects of acute ischemia are complex and may depend on the time after its initiation
and the situation under which it is studied. The ischemia-related direct electrophysio-
logical changes of the membrane potential are markedly influenced by an inhomo-
geneous increase in extracellular potassium that may cause differences in resting
membrane potential. In addition, the increase in sinus rate during ischemia may
cause a marked increase in conduction delay in the ischemic zone. Regional activa-
tion of sympathetic nerves may cause heterogeneous changes in refractory periods
which thus increase the degree of dispersion of refractoriness [45, 59, 66]. Stimula-
tion of f3-receptors may increase the slow inward current and may cause facilitation
of afterdepolarizations and of triggered activity.
Besides acute myocardial infarction, ischemia may be confined to a small area of
the myocardium that may cause some type of abnormal automaticity, which may act
as a triggering factor (see below) influencing a chronic arrhythmogenic tissue after a
previous myocardial infarction.

Role of the Size of Myocardial Infarction

It has been well established that prognosis after myocardial infarction is largely deter-
mined by infarct size [32, 39, 54]. One might expect that impairment of left ventricular
contraction should lead to the development of heart failure and, thus, should be the
factor that determines prognosis. However, there is sufficient evidence that left ven-
tricular dysfunction is more strictly correlated to the occurrence of sudden and thus
170 G.Breithardt et al.

Arrhythmic complications after myocardial infarction

subacute and chronic stage
spontaneous
ventr. arrhythm.
t •
/infarctionl- arrhythmogenic -VT - recurrence
substrate ~ \

/ischemia I - primary - Fig. 2. Factors that influence the

VF arrhythmogenic substrate

possibly arrhythmic death. This suggests that the greater the impairment of left ven-
tricular function, the greater the chance for arrhythmias to occur. The greater chance
of arrhythmias to develop with larger infarcts is probably due to the fact that the
chance of an arrhythmogenic substrate to be present is increased. This is supported
by several studies having shown that ventricular late potentials are more frequent in
patients with more extensive left ventricular dysfunction [9], and that fragmented
electrograms are more frequently recorded in the circumference of the border zone
of an aneurysm in patients with ventricular tachycardia than in patients without [74].
In addition, induction of ventricular tachycardia can be more frequently achieved in
patients with more extensive left ventricular dysfunction [8, 11, 34, 61]. Long-term
follow-up of patients with left ventricular aneurysms has also shown that sudden ar-
rhythmic death is more prevalent than death due to heart failure [39].

Interaction of Various Factors with the Arrhythmogenic Substrate

The presence of an arrhythmogenic substrate in the form of regional slow ventricular

activation is not sufficient to develop ventricular tachycardia. Instead, some addi-
tional factor is necessary to initiate a reentrant type of ventricular tachycardia by
triggering the arrhythmogenic substrate. Once initiated, reentrant tachycardia may
perpetuate itself within the reentrant circuit if the conditions are adequate. Initiation
of ventricular tachycardia, however, can only be achieved if the prerequisites for the
occurrence of reentry (unidirectional block, regional slow activation, reexcitation of
the previously blocked tissue) are met within the arrhythmogenic substrate. The
great number of factors that interact with the arrhythmogenic substrate are pre-
sented in Fig. 2 in an attempt to present a unifying view on sudden cardiac death in
post-myocardial infarction patients.

Triggering Factors

Initiating factors causing reentry (thUS acting as a trigger) may be single premature
ventricular beats, ventricular couplets, or short runs of ventricular tachycardia (sal-
voes) (Fig.2) that may originate from outside the region of subsequent reentry.
Above all, frequent and repetitive ventricular arrhythmias are indicators of the risk
Mechanisms of Syncope and Sudden Death 171

of subsequently developing ventricular tachyarrhythmias after myocardial infarction

[54]. In addition, artificially induced premature beats using programmed stimulation
[8, 11, 13, 14, 34, 61, 73] (Fig. 4) or changes in the activation properties within the
potential reentrant circuit due to changes in basic heart rate [24, 59], may cause re-
entrant excitation. All these events may obviously be modulated by changes in auto-
nomic nervous innervation [45, 69].

Potential Role of Localized Ischemia as a Trigger Factor

Besides spontaneous ventricular arrhythmias, ischemia may also cause acute (tran-
sient) changes in the electrophysiological properties of the arrhythmogenic sub-
strate, either directly (as discussed above) or indirectly via ischemia-induced spon-
taneous repetitive ventricular activity acting as a trigger factor (Fig. 2). The role of
ischemia in the genesis of sudden cardiac death has been suggested by some recent
pathological studies [20, 25]. Davies et al. [20] were able to show that in patients who
died suddenly within 6 h after onset of symptoms, 50% had either isolated multifocal
microscopic necrosis or regional coagulative necrosis. This was frequently caused by
unstable plaques in the proximal coronary arteries due to rupture with or without
thrombus deposition. Partly, this thrombotic material might have been washed
downward into the coronary system. Though the necrotic areas found in these cases
were obviously too small to lead to ventricular fibrillation by themselves, they might
well have caused repetitive ventricular responses that, in the presence of an arrhythmo-
genic substrate after a previous myocardial infarction, might have triggered sus-
tained ventricular tachyarrhythmias.
In another study, Falk [25] reported on the significance of platelet microemboli
in sudden cardiac death occurring within 24 h. Most patients had unstable angina
with fatal outcome. There was autopsy evidence of recurrent mural thrombosis with-
in the coronary arteries with peripheral embolization. This culminated in total vascu-
lar occlusion leading to infarction and/or sudden death. Platelet microemboli were
frequently found in intramyocardial vessels. These aggregates were almost selec-
tively located in the microcirculation distal to thrombosed epicardial arteries.
Another factor that may cause transient episodes of regional ischemia serving as a
trigger factor may be the occurrence of coronary arterial spasm. Bertrand et al. [4]
reported that 20% of patients with recent transmural myocardial infarction showed
provocation of coronary spasm after intravenous injection of 0.4 mg methergine
compared with only 6.2% in patients studied later after myocardial infarction.
The relatively small areas of necrosis as found in the studies by Davies [20] and
Falk [25] are themselves not sufficient to cause ventricular fibrillation. However, by
acting as a trigger factor by locally inducing spontaneous ventricular activity, they
may induce ventricular tachyarrhythmias at sites distant from the one of ischemia.
This is supported by experimental studies. Patterson and co-workers [59] elegantly
showed that ventricular fibrillation could result from ischemia at a site remote from
previous myocardial infarction. Induction of stenosis of the left anterior descending
coronary artery caused ventricular fibrillation in 29 of 40 dogs if there was a previous
myocardial infarction in an area distant from the stenosed coronary artery, but only
in 2 of 10 dogs with similarly great stenosis but without a previous infarct (P<0.004).
172 G. Breithardt et al.

The authors hypothesized that a critical mass of myocardium is required for initiation
and maintenance of ventricular tachycardia. However, on the basis of the data pre-
sented above, we would like to hypothesize that a previous myocardial infarction
provides the arrhythmogenic substrate which is then triggered by ischemia-induced
ventricular arrhythmias occurring at a site distant from the infarction. This would be
another explanation for the observation by Patterson et al. [59] that transient isch-
emia caused ventricular fibrillation only if there was a previous myocardial infarc-
tion.
If the size of ischemic myocardium is sufficiently large (Fig. 2), ventricular fibril-
lation in the absence of previous myocardial infarction may occur and cause sudden
cardiac death. The true incidence of this mechanism is unknown. It would apply to
persons who die suddenly without an antecedent myocardial infarction (primary sud-
den death). Long-term ECG recording is rarely performed, if at all, in these asymp-
tomatic patients. Therefore, no information from patients studied out-of-hospital is
available. However, as this situation resembles the very early stage of acute myo-
cardial infarction, it can be assumed that in most of these cases, sudden cardiac death
is initiated by ventricular fibrillation. This is obviously in contrast to the mechanism
of sudden cardiac death in patients with previous myocardial infarction in whom the
initiating event is mostly a sustained monomorphic ventricular tachycardia [2, 58].

Clinical Course

If the first episode of sustained ventricular tachycardia is too fast and leads to severe
hemodynamic compromise, it may degenerate into ventricular fibrillation and thus
lead to sudden cardiac death. This course of events is suggested by a retrospective
analysis of long-term ECG tapes that were taken at the time of sudden cardiac death
in out-of-hospital patients. It was rare to have immediate ventricular fibrillation as
the initial tachyarrhythmia. Instead, most patients who died from ventricular tachy-
arrhythmias initially exhibited sustained monomorphic ventricular tachycardia,
mostly at rates below 270-300 beats/min [2, 58].
The degeneration of sustained monomorphic ventricular tachycardia into ven-
tricular fibrillation may be due to tachycardia-induced ischemia (Fig. 3). This may be
aggravated by the extent of tachycardia-induced hypotension and changes in reflex-
induced innervation of the myocardium. Another possibility might be that the myo-
cardium is from the onset more susceptible to degeneration into ventricular fibrilla-
tion due to a more extensive and diffuse scarring. Data by Stevenson et al. [68] have
shown differences in the clinical characteristics of patients with previous ventricular
fibrillation ("aborted" sudden death) vs patients with episodes of sustained ventricu-
lar tachycardia: 36% of patients with sudden death but no patient with ventricular
tachycardia had two separate areas of infarction (P < 0.05). Patients with previously
documented sustained ventricular tachycardia had a longer cycle length of induced
tachycardia (319 ± 69 ms) than patients with aborted sudden death (248 ± 31 ms)
(P < 0.05). The induced arrhythmia caused syncope during the electrophysiological
study in 1 of 19 patients with documented sustained ventricular tachycardia but in 6
of 15 patients with aborted sudden death (P<O.OI). In contrast, there were no dif-
ferences with regard to age, time interval from myocardial infarction to ventricular
Mechanisms of Syncope and Sudden Death 173

monanorphe VT (4 min) - KF
1000

V1
V6
HRA -J'----+---I----+--t----,f---+-+--+--t--
I I I I I .,
290 290 280 290

Fig. 3. Degeneration of sustained

monomorphic ventricular tachy-
cardia into ventricular fibrillation
Amiodaron H2797

tachycardia or sudden death, ejection fraction (0.31 % vs 0.29%), and the number of
patients with a major area of contracting myocardium supplied by an artery with a
50% or greater or a 70% or greater stenosis (84% vs 64% and 68% vs 41 %).
If the patient survives the first episode of ventricular tachycardia (occurring out-
side the acute phase of myocardial infarction), recurrences will occur over the long
term in a high percentage of cases [10). Patients who have survived a great number
of episodes of ventricular tachycardia obviously represent a positive selection. They
are nevertheless at risk of developing ventricular fibrillation later on.

Methods for Identification of Patients at Risk of Sndden Cardiac Death

Since the mechanisms of sudden death in general as well as in post-myocardial infarc-

tion patients are complex (Fig. 2 and Table 3), no single method will be able to identify
the patient at risk. Therefore, several techniques have been developed (Fig. 4).
For the analysis of spontaneous ventricular arrhythmias, ambulatory long-term
ECG recording has been used for many years. Multivariate statistical techniques
have shown that frequent and complex ventricular arrhythmias can be considered as
harbingers of sudden death in patients after myocardial infarction [30, 54, 63]. How-
ever, the large spontaneous variability of ventricular ectopic beats and above all of
complex forms such as pairs, salvoes, and runs of nonsustained ventricular tachycar-
dia [1, 51, 53, 75] demands extended periods of long-term ECG recording. Further-
more, the accuracy of the long-term ECG in correctly identifying high-risk patients
has recently been questioned [18] as spontaneous ventricular arrhythmias also occur
174 G.Breithardt et al.

Arrhythmic complications after myocardial infarction

subacute and chronic stage Fig. 4. Methods for diagnostic
24-h assessment of the arrhythmo-
Averaging spontaneous - ECG genic substrate and of those fac-
\ ventr. arrhythm.
t , _ PVS tors that may influence it.
(Abbreviations: EF = estima-
linfarctionl ...... arrhythmogenic -VT - recurrence tion of ejection fraction; PVS =
EtF substrate ~ \ programmed ventricular stimula-
tion; averaging = signal-aver-
lischemia I- primary - aging for detection of ventricular
t VF late potentials)
exercise
ECG

in a large proportion of patients who do not develop ventricular tachycardia or sud-

den death during follow-up. Thus, the large number of false-positive results limits
the clinical value of spontaneous ventricular arrhythmias, especially when it must be
decided who should or should not be treated with antiarrhythmic drugs. In addition,
long-term EeG recording, although noninvasive, is time-consuming, even if
semiautomatic or automatic arrhythmia detection devices are used.
Another factor that may initiate arrhythmias is the occurrence of ischemia. A
propensity to ischemia may be diagnosed using electrocardiographic exercise testing,
thallium myocardial scintigraphy, gated-blood pool imaging, as well as coronary
arteriography. However, though these methods may well be suitable to give an esti-
mate of the propensity to ischemia in an individual patient at the time of the study,
there is no way to predict whether and when a progression in the degree of stenosis
cir even the progression to occlusion will occur.
The effects of spontaneous ventricular arrhythmias and/or of ischemia may be
modulated by the autonomic nervous system. We are still in an early stage of under-
standing its role in the genesis of lethal arrhythmias in patients with coronary artery
disease [45, 69]. There is a need for appropriate techniques that allow analysis of
its role. Our knowledge of the contribution of increased sympathetic discharge to
ischemia-mediated increase in ventricular vulnerability has been based mainly on in-
direct observations. However, there is now growing evidence that autonomous re-
flexes likely play an important role in several clinical conditions such as transient
myocardial ischemia and spontaneous or induced myocardial reperfusion [45, 66,
69]. Preliminary studies in uncomplicated post-infarction patients showed that
measurable cardiac electrophysiological modifications occurred as a consequence
of psychoemotional stimulation even though this mental stress did not induce ar-
rhythmias [69].
Spontaneous ventricular arrhythmias only represent the trigger factor that may
influence the arrhythmogenic substrate and may induce sustained ventricular tachy-
cardia in the presence of predisposing electrophysiological conditions. Presently
used approaches include a more direct assessment of the electrophysiological ("ar-
rhythmogenic") substrate. A promising approach that has been introduced during
recent years is the detection of ventricular late potentials on the body surface [7].
These ventricular late potentials consist of low-amplitude fragmented electrical ac-
tivity that occurs at the end of QRS or during the ST segment and probably origi-
nates from areas of regional slow conduction in the border zone of old myocardial in-
Mechanisms of Syncope and Sudden Death 175

1000
~------.~-------~-------~------~~

D
V

• I " .

S, S, S,52 53 H1229
500 500 21.0 200
Fig.S. Programmed ventricular stimulation with induction of monomorphic sustained ventricular
tachycardia in a patient after recent myocardial infarction. (Abbreviations: t = time reference; I, II,
Vi = leads I, II, Vb resp; HRAJ and HRA2 = two bipolar recordings from the hight right atrium;
HBE = His bundle recording; RVA = recording and stimulating from the right ventricular apex)

farction (see above) [7, 74]. Their presence is closely correlated with the propensity
to ventricular tachycardia [11].
Our own experience is presently based on a total of 628 patients. Only patients
without a history of sustained ventricular tachycardia or fibrillation outside the acute
phase of myocardial infarction, without a history of syncope, and without complete
bundle-branch block were studied. During a mean duration of follow-up of 49 ± 15.0
months, the incidence of arrhythmic complications (sustained ventricular tachycardia
or sudden cardiac death within 1 h) increased as a function of the presence and dura-
tion of late potentials. The risk of major arrhythmic complications was 2.8 times
greater in patients with late potentials of less than 40 ms duration compared with
those without, and 9.3 times greater in those with a duration of 40ms or more.
Similar results were reported by others though the methodology for recording,
the algorithms used, and the entry criteria for the studies differed [21, 38, 40, 72, 73].
Programmed electrical stimulation of the ventricle (Figs.4 and 5) represents
another approach to assess the presence of an arrhythmogenic substrate. Since the
pioneering work by Wellens et al. [41], it has been well known that in patients with
previously documented ventricular tachycardia, identical tachycardias can be in-
duced in a high proportion of cases by appropriate ventricular stimulation tech-
niques. Furthermore, in patients resuscitated from out-of-hospital ventricular fibril-
lation, nonsustained and often sustained ventricular tachyarrhythmias can be in-
duced by programmed ventricular stimulation in about two-thirds of cases [17, 55,
64]. Recently, great interest has developed in the use of programmed ventricular
stimulation, not only for the diagnosis of ventricular tachycardia in patients with pre-
viously documented episodes, but also for assessing the prognosis of patients with
coronary artery disease.
176 G. Breithardt et al.

Our own prospective study includes 548 patients; 402 patients had a history of
previous myocardial infarction. In 269 patients, the study was done within 2 months
after myocardial infarction. Using single and double premature stimuli at different
cycle lengths of basic drive, nonsustained or sustained ventricular tachycardia was in-
duced in 34% and 11% of patients, respectively. During a mean follow-up duration
of 22 ± 18.5 months, 2.8% of patients died suddenly whereas a similar number of pa-
tients developed an episode of spontaneous symptomatic sustained ventricular
tachycardia requiring emergency intervention. Sensitivity for predicting sudden
death was 62%, and for predicting sustained ventricular tachycardia 92%. The high-
est predictive value for an abnormal result of programmed ventricular stimulation
was found in patients in whom sustained ventricular tachycardia at a rate below 270
beats/min was induced (predictive value: 39.3%). In contrast, induction of either no
ventricular echo beats or only nonsustained ventricular tachycardia indicated a good
prognosis as only 8 of 198 cases experienced an arrhythmic event. Relating the induc-
tion of sustained ventricular tachycardia at a rate below 270 beats/min to ejetion frac-
tion (cutoff point 40% ), the predictive value of a positive test increased to 71 %. No
patient without inducible sustained ventricular tachycardia and an ejection fraction
greater than 40% developed an arrhythmic event. A normal result of programmed
stimulation (induction of less than 4 consecutive echo beats) predicted a high chance
of not being at risk of sudden cardiac death or of symptomatic sustained ventricular
tachycardia. Thus, the results of our prospective study in patients after recent myo-
cardial infarction and in patients with chronic coronary artery disease have shown
that the subsequent occurrence of sudden cardiac death (within 1 h) or of symptomatic
sustained ventricular tachycardia requiring some type of emergency intervention
could be predicted by the results of programmed ventricular stimulation in a high
percentage of patients. If the type of induced ventricular arrhythmia (nonsustained
and sustained ventricular tachycardia), the rate of induced ventricular tachycardia,
the interval between myocardial infarction and electrophysiological study, and the
degree of left ventricular impairment (ejection fraction) are taken into account, the
predictive value of the test can be increased.
Similar results were reported by Hamer et al. [34] and Richards et al. [61]. In
contrast, Marchlinski et al. [49] as well as Santarelli et al. [65] were not able to pre-
dict the subsequent occurrence of arrhythmic events as they had a very low event
rate in their small patient cohorts.
In these studies, the predictive value of an abnormal result of programmed stimu-
lation (defined differently in the various studies) varied considerably. This partly de-
pended on the fact that in some studies there were no arrhythmic events or, due to
the small number of these events, no prediction could be made. The only two studies
(besides our study) that actually were able to predict the subsequent occurrence of
sudden cardiac death were those by Hamer et al. [34] and by Richards et al. [61].
These two studies included only patients after recent myocardial infarction. The pre-
dictive value of a positive test for sudden cardiac death was 33.3% and 2l.1 %, re-
spectively. Similar to our study, Richards et al. [61] also observed the occurrence of
sustained ventricular tachycardia after discharge from the hospital, whereas Hamer
et al. [34] did not report any episode of sustained ventricular tachycardia after hospi-
tal discharge.
Mechanisms of Syncope and Sudden Death 177

Clinical Approach to the Evaluation of Post-Myocardial Infarction Patients

From a clinical point of view, noninvasive procedures are obviously desirable for
screening purposes, whereas it would be acceptable to use more aggressive invasive
techniques in certain subsets of patients. A steplike approach using noninvasive re-
cording of late ventricular potentials as the initial step would allow one to preselect
patients for further evaluation by invasive electrophysiological techniques. Whether
this approach might be feasible was tested in 132 post-myocardial infarction patients
who were studied prospectively [8]. It could be demonstrated that the combined use
of signal averaging and programmed ventricular stimulation helped to identify sub-
groups of patients at markedly different risk of developing spontaneous symptomatic
sustained ventricular tachycardia. Patients at highest risk were characterized by (a)
the presence of ventricular late potentials, (b) an abnormal result of programmed
ventricular stimulation, and (c) a rate of induced ventricular arrhythmia less than 270
beats/min. With regard to establishing antiarrhythmic therapy, this might be a sub-
group of post-myocardial infarction patients that might benefit the most.
With respect to presently available information, signal averaging for the detec-
tion of late ventricular potentials seems to be a promising new technique for the
identification of patients at risk of ventricular tachyarrhythmias. However, the rela-
tive value of this technique for the prediction of ventricular tachycardias in compari-
son with sudden cardiac death demands further prospective study. With regard to the
significant number of false-positive results which is not only the case with signal
averaging, but also with, for instance, long-term ECG recording, it seems unjustified
to expect anyone method to be able to identify the individual patient at risk of sus-
tained ventricular tachycardia and/or sudden death. In this context, long-term ECG
recording and signal averaging might prove useful as screening methods, whereas
programmed ventricular stimulation might serve for further risk stratification. No
definite answer can presently be given with regard to the future role of programmed
ventricular stimulation for the identification of patients at risk of severe arrhythmic
events after myocardial infarction. Though the results of prospective studies in large
populations are encouraging and support the view that these arrhythmic events are
obviously related to chronic electrophysiological abnormalities, confirmation of
these data is urgently necessary.

Cardiovascular Mechanisms of Syncope

The various cardiovascular mechanisms that may lead to syncope have been listed in
Table 1. Elucidation of the cause of syncope is often a difficult and time-consuming
task. Several reports have recently focused on the usefulness of electrophysiological
studies for the evaluation of patients with syncope of unexplained origin (Table 4)
[5, 15, 23, 26, 33, 37, 76]. Different diagnostic yields were found in the various
studies, mainly reflecting differences in patient population. In an inhomogeneous
group of patients, DiMarco et al. [23] identified the potential cause of syncope in 17
of 25 patients (68%). Nine of these patients (36%) had inducible ventricular
tachycardia. Brandenburg et al. [5] pointed out that 59 of 92 patients (64% ) revealed
178 G. Breithardt et al.

Table 4. Results of electrophysiologic studies in syncope of unknown origin

No. %

Patients studied 552 100

Normal patients 258 47
Abnormal patients 294 53
Abnormal findings 345
Sinus node dysfunction 47 16
Supraventricular tachycardia 43 14.6
AV dysfunction 14 4.8
Hypersensitive carotid sinus reflex 29 9.9
Conduction disturbance in the His-Purkinje system 51 17.3
Ventricular tachycardia, fibrillation 143 48.6
Vagotonia 18 6.1

Summary from nine studies; see Teichman et al. [77].

an abnormal electrophysiological study. In this study, nine patients (9.7%) had in-
ducible supraventricular tachycardias and ten patients (10.8%) inducible ventricular
dysrhythmias. Gulamhusein et al. [33] found an electrophysiological abnormality in
only 17.6% of patients without clinical signs of structural heart disease. In 3 of 34 pa-
tients (8.8%), supraventricular tachycardias were induced, whereas sustained ven-
tricular tachycardia was not observed. Fisher [26] reported that of 51 patients with
syncope of unknown origin, 14 (27.5%) had inducible ventricular tachycardia. Hess
et al. [37] found a 34% incidence of inducible ventricular tachycardia in 32 patients
with various underlying cardiac abnormalities. Thus, the cause of syncope remains
unexplained in many cases despite extensive electrophysiological evaluation.

Syncope After Myocardial Infarction

Patients who present with syncope occurring for the first time after a myocardial in-
farction represent a specific subset. These patients seem to be at an increased risk of
sudden death [16]. In this situation, the underlying mechanism of syncope and of
sudden death seems to be a ventricular tachyarrhythmia which differs, however, with
regard to the duration of the attack. These patients have in a high majority of cases
the arrhythmogenic substrate for the occurrence of ventricular tachyarrhythmias as
is apparent from the high prevalence of ventricular late potentials [15].
In a previous study, we were able to show that patients with syncope after a
previous myocardial infarction have inducible ventricular tachyarrhythmias in a high
proportion of cases [16]. This study has now been extended to 34 patients. These pa-
tients had in common that their syncope occurred for the first time after a myocardial
infarction. Long-term ECG recordings had not been useful in assessing the cause of
syncope as no severe arrhythmias were noted and none of the patients developed
syncope or presyncope during the recording. The most important finding of the elec-
trophysiological study was the high degree of abnormal ventricular vulnerability as
assessed by programmed right ventricular stimulation: 88% of the patients had an
abnormal response to provocative testing; 16 patients (47%) had sustained and 14
Mechanisms of Syncope and Sudden Death 179

Syncope after mYocardial infarction

Resu as of PVS Fig. 6. Results of programmed

I
n=34 ventricular stimulation in pa-

I I I 1;<>1O~ '5~
~ 'V.flutter' 'sust. vr' Inon-sust. vrll 0 - 3 VEl tients with syncope after myo-
cardial infarction. (Abbrevia-
n=14 - - - - tions: VF = ventricular fibril-
I!ID lation; V. flutter = ventricular

-----
VEl
n=5 n=1 n=1O n=7 n=5 n=2 n=4 flutter; sust. VT = sustained
ventricular tachycardia; VE =
47 % 41.2 % 11.8 % ventricular echo beats)
1326 M

Syncope after mYocardial infarction

follow-up 8 months (median, 1-37)
_n=34 _ _ _
n-o-i:-n-'du-c.::i!!i=:bl=e::V=-A:;'--;'=non-su~taii1edVA1
,..- /SIISiained VAl

~ n=\ / n=14,\ / ' n=16 "-..

n=1
syncope
alive
n=3
no. event /
al1Ve
n=5
I'\. n=9
no. event
'\. ahve
n=1 n=15
non-cardi ac no event
death alive
n=2 n=1 n=2
syncope doc. sudden
alive VF death
1327 M
Fig. 7. Results of long-term follow-up in patients with syncope after myocardial infarction in relation
to the results of programmed ventricular stimulation. (Abbreviations: VA = ventricular arrhythmias;
doc. VF = documented ventricular fibrillation)

patients (41%) nonsustained ventricular tachyarrhythmias inducible (Fig. 6). Fur-

thermore, other coexisting electrophysiological abnormalities such as conduction
disturbances (5 of 27 patients) or abnormal sinus node function (5 of 27 patients)
were identified. Based on these results, we assumed that the syncope occurring for
the first time after myocardial infarction was due to some kind of short-lasting (self-
terminating) ventricular tachyarrhythmia [16]. However, as all patients had a history
of myocardial infarction antedating the occurrence of syncope, one might argue that
the results of programmed right ventricular stimulation may not be specific, but may
reflect a laboratory phenomenon related to the underlying cardiac abnormality
rather than to the symptoms of the patients. This explanation seems to be unlikely
when the results in these patients with syncope were compared with matched pa-
tients without a history of syncope but with a similar degree of left ventricular im-
pairment (unpublished data). Additionally, the clinical course of the patients after
initiation of therapy (Fig. 7) gives further support to the concept that the induced
ventricular tachyarrhythmia is of diagnostic importance and a specific finding in this
population. After initiation of antiarrhythmic therapy controlled by repeated elec-
trophysiological testing [12] or long-term EeG recording, none of 16 patients with
inducible sustained ventricular arrhythmias had recurrence of syncope. However,
spontaneous remission of syncope may play an important role, and successful treat-
ment may be attributed to this phenomenon [33, 76]. This observation has also been
made by Gulamhusein et al. [33] who found that 16 of 34 patients (47%) had no fur-
ther episode of syncope in the absence of any intervention.
180 G.Breithardt et al.

Blood pressure during VT - relation between syncope and

sudden dea th
~
• rate of VT
mean blood • LV function
pressure • body pos iti on
• counterregulat;on
• aa. drug therapy

J..------------ VF
-.1._---- ------- VF Fig. 8. Response of arterial blood
pressure at the onset of ven-
tricular tachycardia and its rela-
1__---41 non-sustained VT
tion to syncope and to sudden
cardiac death

t
time

onset of VT

syncope

Relation Between Syncope and Sudden Cardiac Death

Any type of brady- or tachyarrhythmia may lead to syncope if it is severe enough and
of sufficient duration to cause cerebral hypoperfusion. It is rare for supraventricular
tachycardias to cause syncope except for patients with WPW syndrome in whom
atrial fibrillation with rapid conduction over the accessory pathway develops. How-
ever, in the elderly patient, even episodes of intermittent atrial fibrillation, which
would otherwise be well tolerated by a younger patient, may lead to syncope, above
all, if there is coexistent cerebrovascular disease. The same principles' apply to syn-
cope due to bradyarrhythmias. If the brady- or tachyarrhythmia lasts long enough
and is sufficiently severe, above all in the presence of organic heart disease, sudden
death may occur.
An important clinical entity in which both syncope and sudden cardiac death may
occur is the occurrence of "syncope after myocardial infarction" as has been exten-
sively described above. Though bradyarrhythmias, e.g., due to intermittent third-
degree A V block, may occur as a consequence of infarction-related bundle-branch
block, our present experience suggests that there is a high probability of ventricular
tachyarrhythmias. Under these specific circumstances, syncope and sudden cardiac
death represent the extremes of a wide spectrum of responses to ventricular tachy-
arrhythmias (Fig. 8). Several factors are involved in this situation. These include the
rate and duration of the arrhythmia, the propensity to degeneration of monomorphic
ventricular tachycardia into ventricular fibrillation, the extent of the preexisting im-
pairment of left ventricular contraction, the extent and duration of the drop in blood
pressure, and the possible occurrence of ischemia, which might also lead to degener-
ation into ventricular fibrillation.
Mechanisms of Syncope and Sudden Death 181

Conclusions

Due to this interrelation between syncope and sudden cardiac death in patients after
myocardial infarction, vigorous efforts are obviously warranted to identify these pa-
tients as early as possible. A carefully taken history in patients who later died sud-
denly frequently reveals that syncope occurred during the preceding days or weeks
(personal experience). Thus, syncope could have served as a warning condition for
subsequent sudden death after which early intervention could have been instituted.
This emphasizes the great importance of identifying the patient with syncope after
myocardial infarction and of treating this entity as a seperate clinical syndrome. It is
obviously not sufficient to rely on a negative long-term ECG recording to exclude
the propensity to life-threatening ventricular arrhythmias in a patient with syncope
after myocardial infarction. Instead, we would advocate the more extensive use of
electrophysiological studies to identify the underlying mechanism of otherwise un-
explainable syncope and the use of this approach for the control of antiarrhythmic
drug efficacy as well. More experience is obviously needed before the best approach
to the management of these patients will be established.
Acknowledgements. This work was supported by the Johann A. Wiilfing-Stiftung, the Berta und
Ernst Grimmke-Stiftung, the Sonderforschungsbereich 30 (Kardiologie) and the Sonderforschungs-
bereich 242 (Koronare Herzkrankheit - Prophylaxe und Therapie akuter Komplikationen) of the
Deutsche Forschungsgemeinschaft, Bonn-Bad Godesberg (FRG). Dr. A. Martinez-Rubio was sup-
ported by a grant from the C.I.R.I.T. (Comissi6 Interdepartamental per la Recerca i l'Innovaci6
Tecnologica), Generalitat de Catalunya, Spain.

References

1. Andresen D, von Leitner E-R, Wegscheider K, Schroder R (1982) Nachweis komplexer tachy-
karder ventrikularer Rhythmusst5rongen im Langzeit-EKG. Dtsch Med Wochenschr 107:
571-574
2. Bayes de Luna A, Tomer P, Guindo J, Soler M, Oca F (1985) Holter ECG study of ambulatory
sudden death, review of 158 published cases. New Trends in Arrhythmias 1:293-297
3. Berbari EJ, Scherlag BJ, Hope RR, Lazzara R (1978) Recording from the body surface of ar-
rhythmogenic ventricular activity during the ST segment. Am J Cardiol 41 : 697 -702
4. Bertrand ME, Lablanche JM, Tilmant PY, Thieuleux FA, Delforge MR, Carre AG, Asseman
P, Berzin B, Libersa C, Laurent JM (1982) Frequency of provoked coronary arterial spasm in
1089 consecutive patients undergoing coronary arteriography. Circulation 65: 1299-1306
5. Brandenburg RO, Holmes GR, Hartzler GO (1981) The electrophysiologic assessment of pa-
tients with syncope. Am J Cardiol 47: 433
6. Breithardt G, Becker R, Seipel L, Abendroth RR, Ostermeyer J (1981) Non-invasive detection
of late potentials in man a new marker for ventricular tachycardia. Eur Heart J 2: 1-11
7. Breithardt G, Borggrefe M (1986) Pathophysiological mechanisms and clinical significance of
ventricular late potentials. Eur Heart J 7: 364-385
8. Breithardt G, Borggrefe M, Haerten K (1985) Role of programmed ventricular stimulation and
noninvasive recording of ventricular late potential for the identification of patients at risk of ven-
tricular tachyarrhythmias after acute myocardial infarction. In: Zipes DP, Jalife J (eds) Cardiac
electrophysiology and arrhythmias. Grone and Stratton, Orlando, pp 553-561
9. Breithardt G, Borggrefe M, Karbenn U, Abendroth RR, Yeh HL, Seipel L (1982) Prevalence
of late potentials in patients with and without ventricular tachycardia: correlation to angio-
graphic findings. Am J Cardiol 49: 1932-1937
182 G. Breithardt et al.

10. Breithardt G, Borggrefe M, Podczeck A (1988) Has the clinical presentation of ventricular
tachycardia changed? In: Breithardt G, Borggrefe M, Zipes D (eds) Non-pharmacological ther-
apy of tachyarrhythmias. Futura, New York (in press)
11. Breithardt G, Borggrefe M, Quantius B, Karbenn U, Seipel L (1983) Ventricular vulnerability
assessed by programmed ventricular stimulation in patients with and without late potentials.
Circulation 68: 275-281
12. Breithardt G, Borggrefe M, Seipel L (1984) Selection of optimal drug treatment of ventricular
tachycardia by programmed electrical stimulation of the heart. In: Clinical aspects of life-
threatening arrhythmias. Ann NY Acad Sci 427: 49-66
13. Breithardt G, Seipel L, Loogen F (1980) Der akute Herztod. Bedeutung elektrophysiologischer
Stimulationverfahren. Verh Dtsch Ges Herz-Kreislaufforschg 46: 38-64
14. Breithardt G, Seipel L, Meyer T, Abendroth R-R (1982) Prognostic significance of repetitive
ventricular response during programmed ventricular stimulation. Am J Cardiol 49 : 693-698
15. Borggrefe M, Breithardt G, Karbenn U (1985) Clinical value of late potentials in patients with
syncope of unknown origin. Eur Heart J [Suppl 1] 6: 18
16. Borggrefe M, Breithardt G, Seipel L (1983) Syncope after myocardial infarction. Electrophysio-
logic findings. Circulation [Suppl III] 68 : III-55
17. Borggrefe M, Breithardt G, Yeh HL, Seipel L (1982) Klinisch-elektrophysiologische Befunde
bei Patienten nach Kammerflimmern. Z Kardiol 71 : 643-648
18. Cats VM, Lie KJ, van Capelle FJL, Durrer D (1979) Limitations of 24 hour ambulatory electro-
cardiographic recording in predicting coronary events after acute myocardial infarction. Am J
Cardiol 44: 1257-1262
19. Daniel T, Boineau J, Sabiston D (1971) Comparison of human ventricular activation with canine
model in chronic myocardial infarction. Circulation 44: 74-89
20. Davies MJ, Path PRC, Thomas AC, Path MRC, Knapman PA, Hangartner JR (1986) Intra-
myocardial platelet aggregation in patients with instable angina pectoris suffering sudden isch-
emic cardiac death. Circulation 73: 418-427
21. Denniss AR, Cody DV, Fenton SM, Richards DA, Ross DL, Russell PA, Young AA, Uther
JB (1983) Significance of delayed activation potentials in survivors of myocardial infarction.
J Am Coll Cardiol 1: 582
22. Dhingra RC, Denes P, Wu D, Chuquimia R, Amat-y-Leon F, Wyndham CH, Rosen KM (1974)
Syncope in patients with chronic bifascicular block. Significance, causative mechanisms, and
clinical implications. Ann Inter Med 81: 302-306
23. DiMarco JP, Garan H, Harthorne JW, Ruskin IN (1981) Intracardiac electrophysiologic tech-
niques in recurrent syncope of unknown cause. Ann Intern Med 95 : 542-548
24. El-Sherif N, Scherlag BJ, Lazzara R, Hope RR (1977) Reentrant ventricular arrhythmias in the
late myocardial infarction period. I. Conduction characteristics in the infarction zone. Circula-
tion 55 : 686-702
25. Falk E (1985) Unstable angina with fatal outcome: dynamic coronary thrombosis leading to in-
farction and/or sudden death. Autopsy evidence of recurrent mural thrombosis with peripheral
embolization culminating in total vascular occlusion. Circulation 71: 699-708
26. Fisher JD (1981) Role of electrophysiologic testing in the diagnosis and treatment of patients
with known and suspected bradycardias and tachycardias. Progr Cardiovasc Dis 24: 25-90
27. Flowers NC, Horan LG, Thomas JR, Tolleson WJ (1969) The anatomic basis for high frequency
components in the electrocardiogram. Circulation 39: 531-539
28. Fontaine G, Frank R, Gallais-Hamonno F, Allali I, Phan-Thuc H, Grosgogeat Y (1978) Electro-
cardiographie des potentiels tardifs du syndrome de post-exitation. Arch Mal Coeur 71: 854-864
29. Gardner PI, Ursell PC, Fenoglio JJ, Wit AL (1985) Electrophysiologic and anatomic basis for
fractionated electrograms recorded from healed myocardial infarcts. Circulation 72: 596-611
30. Geltman EM, Ehsani AA, Campbell MK, Schechtman K, Roberts R, Sobel BE (1979) The in-
fluence of location and extent of myocardial infarction on long-term dysrhythmia and mortality.
Circulation 60: 805-814
31. Goldstein S, Friedman L, Hutchinson R, Canner P, Romhilt D, Schlant R, Sobrino R, Verter
J, Wasserman A, The Aspirin Myocardial Infarction Study Research Group (1984) Timing,
mechanism and clinical setting of witnessed death in postmyocardial infarction patients. J Am
Coll Cardiol3: 1111-1117
32. Grande P, Pedersen A (1984) Myocardial infarct size: correlation with cardiac arrhythmias and
sudden death. Eur Heart J 5 : 622-627
Mechanisms of Syncope and Sudden Death 183

33. Gulamhusein S, Naccarelli GV, Prystowsky EN, Zipes DP, Barnett HJ, Heger JJ, Klein GJ
(1982) Value and limitations of clinical electrophysiologic study in assessment of patients with
unexplained syncope. Am J Med 73: 700-706
34. Hamer A, Vohra J, Hunt J, Sloman G (1982) Prediction of sudden death by electrophysiologic
studies in high risk patients surviving acute myocardial infarction. Am J Cardiol50: 223-229
35. Han J (1969) Ventricular vulnerability during acute coronary occlusion. Am J CardioI24:857-
864
36. Han J, Garcia de Jalon PD, Moe GK (1966) Fibrillation threshold of premature ventricular re-
sponses. Circ Res 18: 18-26
37. Hess DS, Morady F, Scheinman MM (1982) Electrophysiologic testing in the evaluation of pa-
tients with syncope of undetermined origin. Am J Cardiol 50: 1309-1315
38. Hopp HW, Hombach V, Osterspey A, Deutsch H, Winter U, Behrenbeck DW, Tauchert M,
Hilger HH (1985) Clinical and prognostic significance of ventricular arrhythmias and ventricular
late potentials in patients with coronary heart disease. In: Holter monitoring technique. Techni-
cal aspects and clinical applications. Schattauer, Stuttgart, pp 297-307
39. Jehle J, Heerdt M, Spiller P, Loogen F (1982) VerJaufsbeobachtungen bei Patienten mit
Aneurysma des linken Ventrikels nach konservativer und chirurgischer Therapie. Z Kardiol
71:566-575
40. Kacet S, Libersa C, Caron J, Boudoux B, d'Haute Feuille X, Marchand J, Dagano J, Lekieffre
J (1988) The prognostic value of averaged late potentials in patients suffering from coronary
artery disease. In: Aliot E, Lazzara R (eds) Ventricular tachycardias. Martinus Nijhoff Pub-
lishers, Dordrecht-Boston-Lancaster (in press)
41. Kanovsky MS, Simson MB, Falcone R, Dresden C, Josephson ME (1983) The late potential is
an independent marker for ventricular tachycardia. J Am Coli Cardiol1 : 582
42. Kapoor WN, Karpf M, Wieand S, Peterson JR, Levey GS (1983) A prospective evaluation and
follow-up of patients with syncope. N Eng! J Med 309: 197-204
43. Kuchar D, Thorburn C, Sammel N (1985) Natural history and clinical significance of late poten-
tials after myocardial infarction. Circulation [Suppl III] 72: III-477
44. Langner Jr PH, Gese10witz DB, Briller SA (1973) Wide band recording of the electrocardio-
gram and coronary heart disease. Am Heart J 86: 308-317
45. Lombardi F (1986) Acute myocardial ischemia, neural reflexes and ventricular arrhythmias. Eur
Heart J [Suppl A] 7:91-97
46. Lown B (1968) Electrical stimulation to estimate the degree of digitalization. II. Experimental
studies. Am J CardioI22:251-259
47. Madsen JK (1985) Ischaemic heart disease and prodromes of sudden cardiac death. Is it possible
to identify high risk groups for sudden cardiac death. Br Heart J 54: 27-32
48. Ma11iani A, Schwartz PJ, Zanchetti A (1980) Neural mechanisms in life-threatening arrhythmias.
Am Heart J 100: 705-715
49. Marchlinski FE, Buxton AE, Waxman HL, Josephson ME (1983) Identifying patients at risk of
sudden death after myocardial infarction: Value of the response to programmed stimulation,
degree of ventricular ectopic activity and severity of left ventricular dysfunction. Am J Cardiol
52: 1190-1196
50. Matta RJ, Verrier RL, Lown B (1976) The repetitive extrasystole as an index of vulnerability to
ventricular fibrillation. Am J Physiol 230: 1469-1473
51. Michelson El, Morganroth J (1980) Spontaneous variability of complex ventricular arrhythmias
detected by long-term electrocardiographic recording. Circulation 61: 690-695
52. Morady F, Shen E, Schwartz A, Hess D, Bhandari A, Sung RJ, Scheinman MM (1983) Long-
term follow-up of patients with recurrent unexplained syncope evaluated by electrophysiologic
testing. J Am Coli Cardiol 2: 1053-1059
53. Morganroth J, Michelson EL, Horowitz LN, Josephson ME, Pearlman AS, Dunkman WB
(1978) Limitations of routine long-term electrocardiographic monitoring to assess ventricular
ectopic frequency. Circulation 58:408-414
54. Moss AJ (1980) Clinical significance of ventricular arrhythmias in patients with and without
coronary artery disease. Progr Cardiovasc Dis 23: 33-52
55. Myerburg RJ, Conde CA, Sung RJ, Mayorga-Cortes A, Mallon SM, Sheps DS, Appel RA,
Castellanos A (1980) Clinical, electrophysiologic and hemodynamic profile of patients resus-
citated from prehospital cardiac arrest. Am J Med 68: 568-576
184 G. Breithardt et al.: Mechanisms of Syncope and Sudden Death

56. Nalos PC, Gang ES, Mandel WJ, Ladenheim ML, Lass Y, Peter T (1987) The signal-averaged
electrocardiogram as a screening test for inducibility of sustained ventricular tachycardia in high
risk patients: a prospective study. J Am Coll Cardiol 9: 539-548
57. Olshansky B, Mazuz M, Martins JB (1985) Significance of inducible tachycardia in patients with
syncope of unknown origin: a long-term follow-up. J Am Coll Cardiol 5: 216-223
58. Olshausen K, Pop T, Treese N, Meyer J (1985) Sudden death during Holter monitoring. Eur
Heart J [Suppl I] 6: 45
59. Patterson E, Gibson JK, Lucchesi BR (1982) Electrophysiologic actions of lidocaine in a canine
model of chronic myocardial ischemic damage. Arrhythmogenic actions of lidocaine. J Cardio
PharmacoI4:925-934
60. Richards DA, Blake GJ, Spear JF, Moore EN (1984) Electrophysiologic substrate for ventricu-
lar tachycardia: correlation of properties in vivo and in vitro. Circulation 69: 369-381
61. Richards DA, Cody DV, Denniss AR, Russell PA, Young AA, Uther JB (1983) Ventricular
electrical unstability: a predictor of death after myocardial infarction. Am J Cardiol 51: 75-80
62. Rozanski JJ, Mortara D, Myerburg RJ, Castellanos A (1981) Body surface detection of delayed
depolarizations in patients with recurrent ventricular tachycardia and left ventricular aneurysm.
Circulation 63 : 1172-1178
63. Ruberman W, Weinblatt E, Goldberg JD, Frank CW, Shapiro S, Chaudhary BS (1980) Ven-
tricular premature complexes in prognosis of angina. Circulation 61: 1172-1178
64. Ruskin IN, DiMarco JP, Garan H (1980) Out-of-hospital cardiac arrest. Electrophysiologic ob-
servations and selection of long-term antiarrhythmic therapy. N Engl J Med 303: 607-613
65. Santarelli P, Bellocci F, Loperfido F, Mazzari M, Mongiardo R, Montenero AS, Manzoli U,
Denes P (1985) Ventricular arrhythmia induced by programmed ventricular stimulation after
acute myocardial infarction. Am J Cardiol 72: 487-494
66. Schwartz PJ, Zaza A (1986) The rational basis and the clinical value of selective cardiac sym-
pathetic denervation in the prevention of malignant arrhythmias. Eur Heart J [Suppl A] 7:
107-118
67. Simson MB (1981) Use of signals in the terminal QRS complex to identify patients with ventricu-
lar tachycardia after myocardial infarction. Circulation 64: 235-242
68. Stevenson WG, Brugada P, Waldecker B, Zehender M, Wellens HJJ (1985) Clinical angio-
graphic and electrophysiologic findings in patients with aborted sudden death as compared with
patients with sustained ventricular tachycardia after myocardial infarction. Circulation 71:
1146-1152
69. Tavazzi L, Zotti AM, Rondanelli R (1986) The role of psychologic stress in the genesis of lethal
arrhythmias in patients with coronary artery disease. Eur Heart J [Suppl A]7 :99-106
70. Van Durme JP (1975) Tachyarrhythmias and transient cerebral ischemic attacks. Am Heart J
Annotations 89: 538
71. Von Leitner ER, Andresen A, Meyer V (1977) Untersuchung von Patienten mit unklaren Syn-
kopen mittels Langzeit-EKD und His-Bundel-Elektrokardiographie. Verh Dtsch Ges Kreislauf-
forschg 43: 382-383
72. Von Leitner R, Oeff M, Loock D, Jahns B, SchrOder R (1983) Value of non-invasively detected
delayed ventricular depolarizations to predict prognosis in post myocardial infarction patients.
Circulation [Suppl III] 68: III-83
73. Wellens HJJ, Duren DR, Lie KI (1976) Observations on mechanisms of ventricular tachycardia
in man. Circulation 54:237-244
74. Wiener I, Mindich B, Pitchon R, Pichard A, Kupersmith J, Estioko M, Jurado R, Camunas J,
Lifwak R (1982) Epicardial activation in patients with coronary artery disease: Effects of regional
contraction abnormalities. Circulation 65: 154-160
75. Winkle RA (1980) Antiarrhythmic drug effect mimicked by spontaneous variability of ventricu-
lar ectopy. Circulation 61 : 690-696
76. Yeh H-L, Breithardt G, Seipel L, Borggrefe M, Loogen F (1982) Elektrophysiologische Be-
funde und Langzeitbeobachtung bei Patienten mit Synkopen. Z Kardiol 71 : 263-270
Chapter 13

Possible Role of the Fear Paralysis Reflex

in Sudden Cardiac Death
B.KAADA

Introduction

Sudden cardiac death is generally assumed to be due to ventricular fibrillation occur-

ring in a setting of coronary heart disease [11]. The role of a neural factor in the
mechanism of sudden cardiac death was first advanced by MacWilliam in 1887 [52],
who also suggested that sympathetic discharges may be the important factor in initiat-
ing the fatal arrhythmia. More recently, evidence for the participation of the brain in
inducing ventricular fibrillation through sympathetic discharges in emotional stress,
in many instances also in the absence of coronary disease, has in particular been pro-
vided by Lown and associates [46,47,49,50,62], Corbalan et al. [8] and DeSilva [11],
although psychical factors have been given much less attention than somatic ones.
The traditional somatic risk factors fail to account for half of the mortality from coro-
nary heart disease encountered on a worldwide basis [6]. There can be no doubt that
higher nervous activity frequently is implicated in the altered ventricular ectopic
activity. It has been estimated that sudden cardiac death may be preceded by acute
psychical disturbances in 20%-40% of cases [11, 62]. Structural damage ofthe heart
is not a prerequisite for the occurrence of ventricular fibrillation [11]. The latter may
be caused by transient impulses to the heart which momentarily disorganize cardiac
electrical activity [12, 45, 47, 48]. A dominant role of the left sympathetic nerves in
arrhythmogenesis and the antifibrillatory effect on its lesioning have been well docu-
mented [66].
Cardiac damage can, however, also result from parasympathetic overstimulation.
Ventricular ectopic rhythms followed vagal stimulation in dogs with acute myocar-
dial infarction [43], and the same was observed in chronic vagal stimulation in dogs
[53]. In monkeys similar cardiac necrosis resulted from vagal stimulation as did
stimulation of hypothalamic nuclei [27, 54, 55].
Cardiac arrest in diastole resulting from direct or reflex vagal stimulation in healthy
persons is a well recognized cause of sudden death in humans [1, 29, 42]. Kayssi [42]
emphasized that death by cardiac inhibition is accepted as a separate identity in
forensic medicine. Death occurs in a matter of minutes and may not show cardiac
changes at postmortem.
There are a number of reports on sudden cardiac deaths in adults precipitated by
psychical stimuli and associated with bradycardia to which we shall return later.
As emphasized by Dimsdale [13], psychically induced sudden cardiac deaths are
not a uniform phenomenon. It occurs in a number of different contexts, each of
which has its own pathophysiology. It is not sufficient to state that the individual was
186 B.Kaada

under "psychical stress." The psychical variables implicated and the cardiac mediator
need to be defined. Psychical stress may mean such diverse conditions as increased
mental activity, anxiety, anger, fear, frustration, grief, and the cardiac mediator may
be sympathetic or parasympathetic. What are the factors that in some situations
make the heart rate increase under psychical stress and in other situations decrease?
For instance, mental stress (arithmetic to be executed aloud rapidly) [49,71], anxi-
ety, and anger ordinarily induce cardiovascular arousal with tachycardia, whereas
bradycardia may occur in certain conditions of threat and fear.
The purpose of the present report is to draw attention to the latter psychological
variable, threat-induced fear, which activates the "fear paralysis reflex" and is associ-
ated with profound bradycardia due to excessive cardiac parasympathetic discharges.
The latter may lead to asystole, fatal cardiac arrhythmia, and death. This reflex has
been demonstrated in all animals tested. It has been extensively studied in animals
by ethologists and in physiological psychology, but is less known within the medical
profession. It has recently been proposed that the sudden infant death syndrome
may be triggered by this inborn atavistic fear paralysis reflex [20, 34-36].
It is here postulated that the fear paralysis reflex may also be a cause of sudden
cardiac death in adults. By identifying and delineating this fixed behavior pattern as
one cause of sudden cardiac death, much can be learned about the cardiac catastrophy
due to the elicitation of this reflex, as one may profit from the great number of previ-
ous studies performed in other connections.
First, a brief review is given of the main relevant features of the fear paralysis re-
flex in animals. Thereafter, its possible relation to certain categories of sudden car-
diac death in adults is discussed.

Characteristics of the Fear Paralysis Reflex

For a more comprehensive description of the fear paralysis reflex and references, the
reader is referred to the monograph by the author [34].

Stimuli

The fear paralysis reflex is elicited by any threatening stimulus perceived as a danger,
as in confrontation with a potential predator. The prey animal may react either with
flight/fight (active avoidance) or, when such actions are not possible, with a sudden
and prolonged state of immobility (passive avoidance). This may be observed either
as a "freezing" on approach of the predator or as "fear paralysis" after the predator
has seized or touched the prey [60]. Fear paralysis may, however, also occur without
physical contact.
The actual stimulus which triggers this reflex has been under discussion, but there
is now general agreement that it is fear prompted by aversive environmental events
[22, 60]. Fear refers to a present situation, while anxiety refers to what is not immedi-
ately present, but to a future or past state. An important feature of the induction of
fear paralysis is a condition of helplessness and hopelessness [64]. The subject is in
Fear Paralysis Reflex and Sudden Cardiac Death 187

an inescapable, overwhelming situation for which no solution is available and which

it is unable to cope.
Stimuli which in animals may evoke the innate fear response are, besides predator
confrontation, restraint of movement (preventing flight), loss of support, unex-
pected and sudden noises, distress, and alarm calls. The response is greatly reduced
in the presence of the mother or companions. Stimuli which may be of particular im-
portance in human sudden infant deaths are restraint of movement, sudden and un-
familiar noises, separation from the mother, and exposure to unfamiliar environ-
ments and strangers.
The fear paralysis hypothesis as an explanation of sudden human infant death has
found support in three recent studies: (1) Norvenius [57] considered restriction of
movements as a contributing factor to the death in nearly one-third of the cases [cf.
34], (2) Kaada and Gundersen [38] have confirmed that there is a clear excess of sud-
den infant deaths during weekend/holidays, and (3) Kariks [41] has brought evi-
dence from postmortem studies that the sudden human infant death is cardiogenic
and not respiratory (cf. below). The two first mentioned observations strongly indi-
cate environmental trigger mechanisms, thus complying with the fear paralysis hypo-
thesis.

Response

The reflex response in animals and man is an immediate motor and autonomic "para-
lysis" associated with excessive parasympathetic activation with profound bradycar-
dia. In animals this takes place without struggling or agonal outcry. Besides the
generalized somatomotor inhibition, there is reduced muscular tone and apnea in ex-
piration. These responses are due to an active central inhibition, including activation
of a pain-suppressing mechanism. Blood pressure rises are due to peripheral vaso-
constriction. The reflex follows acute exposure with an immediate and relatively
short-lasting effect with a duration ofless than 1 min [61].

Functional Significance

The reflex has survival value: the motor quieting makes it more difficult to discover
the prey; the suppression of pain is favorable in case of wounding; the bradycardia is
part of a complex oxygen-conserving cardiovascular response involving regional
shifts of blood supply with constriction of most vascular beds, maintaining central
organ perfusion pressure, and reserving the limited oxygen stores for the vital or-
gans. Darwin [10] advanced his "death feigning" theory that the immobility reflects
an adaptive behavior to predation.

Terminology

Throughout the years this innate fixed behavior pattern has been variously termed
[69], such as animal hypnosis, death feigning, tonic immobility, hypotonic akinesia,
188 B.Kaada

Totstell Reflex, still reaction, playing dead reaction, play opossum, passive fear, fear
paralysis, and others. The term "fear paralysis reflex" or the shortened version "fear
paralysis," introduced by Leyhausen [44], has been preferred in our studies because
it emphasizes the nature of the stimulus, the immediate reflex and inborn character
of the response, and the essentially inhibitory response pattern.

Age Distributiou

The fear paralysis reflex has an age peak distribution in the first weeks or months of
life coinciding with the development of emotional and fearful responsiveness. In
small animals such as mice, rats, and chickens, the sensitive period is at the age of
1-2 weeks, and in dogs and rhesus monkeys at the age of 2-4 months [34, 35]. The
latter peak fits the unique age distribution of the sudden infant death syndrome,
which has a similar peak at age 2-4 months.
The attenuation of the fear paralysis after this critical period has been attributed
in mammals to the rapid development of the neocortex [51], in particular the frontal
lobes. Following neodecortication the reflex is enhanced [51]. Thus, the neural sub-
strate of the reflex is probably present throughout adult life, although it is subjected
to inhibitory influences from higher levels of the central nervous system. Presumably
in adult life it may be released by its proper stimulus, acute threat causing fear,
through a process of removal of inhibition, i.e., disinhibition.

Examples of the Fear Paralysis Reflex in Animals

There is abundant literature describing the reflex in a number of animals in natural

environmental and laboratory settings [reviews, see 20, 22, 34-36]. A few examples
will serve to illustrate the variability of the stimuli which may evoke the reflex.
A profound bradycardia and immobility were recorded in wild willow grouse
[21], red deer calves [18], and white-tailed deer fawns [33] when exposed to threaten-
ing stimuli such as unfamiliar events, dog's barking, lightening and thunder, and the
approach of an unfamiliar observer (Fig. 1).
During testing of captured wild rodents in an open field arena with the animals
exposed to a natural predator (snake), a sudden visual (stylized hawk silhouette) or
a loud auditory stimulation, the rodents exhibited periods of prolonged immobility
and unresponsiveness of 2-60 min duration and severe bradycardia [30]. Two-thirds
of the animals developed cardiac arrhythmias, and one-fifth of the trapped rodents
died within the 1st week in the laboratory.
Similar responses were observed in preweanling rat pups when exposed to sud-
den noise in unfamiliar surroundings [31]. Exposure to loud noises was also found to
be an effective means of prolonging fear paralysis in domestic fowl [22, 23], lizards
[14], and frog [56]. A number of deaths were witnessed when birds were exposed to
fearful stimuli [22].
Richter [64] observed that in wild rats fatal cardiac arrest in diastole occurred
simply when being held in the hand; this restraint of movement suddenly and finally
abolished all hope of escape. Bradycardia and sudden death were also provoked by
Fear Paralysis Reflex and Sudden Cardiac Death 189

Phase I II III IV C V
Event A
Breathing" •
(normal) (apnea)
220
•
•~ • •• ••
iQ.
~
180
'.
. . . e!. ··1
'".
•• Yo• .: '.
"/I
~""~
-.,,- ~

!oJ
~
0:
140
"
~
0:
«
!oJ 100 "
:J:

•
60 •• ••• • • .. ••••• •
• •
••
•
20

-10 0 10 20 30 40 50 60 70
ELAPSED TIME (seel
Fig. 1. Representative example of the changes in the instantaneous heart rate and pattern of breath-
ing during the prone response to alarm stimuli observed in a I-day-old female white-tailed deer
fawn. Event A, abrupt bradycardia (from average 177 beats/min at rest to 60 beats/min) and apnea
evoked by approach to within 1 m of the fawn by an unfamiliar observer; Event B, a nearby motion
or noise caused a reversal of recovery; Event C, return of heart rate toward resting levels as the ob-
server left the immediate area of the fawn. (Jacobsen [33])

immersion in a swimming jar under a jet of water and by trimming of the whiskers.
Richter concluded that the situation of restraint, of being hopelessly trapped, leads
to intense fear with excessive parasympathetic discharges and slowing of the heart
rather than acceleration. He also evoked the concept of a "giving up" response.
Atropine prevented and cholinergic drugs hastened these rapid deaths.
The cardiovascular and respiratory responses of the fear paralysis reflex are
reminiscent of those seen in the "dive" and "smoke" reflexes, both initiated via
trigeminal and glossopharyngeal afferents. The dive reflex, which previously has
been discussed in relation to sudden death [4, 64, 73-75], was also said to be pro-
foundly influenced from higher levels of the central nervous system. Thus, severe
bradycardia may occur merely in anticipation of immersion [4, 19,28,59], as well as
a result of frightening stimuli, such as when forced to submerge [40]. In these cases,
however, one is more likely dealing with the fear paralysis reflex. The smoke reflex
is elicited by allowing the smoke of a cigarette to envelop the rabbit's nostrils, which
causes a severe bradycardia and general motor inhibition with loss of muscular tone.
These reflexes have been used in studies of the physiology and pharmacology of the
fear paralysis reflex [34, 37].
Severe bradycardia which might progress to ventricular asystole and death was
seen in monkeys which experienced unpredictable electric shocks without the ability
190 B.Kaada

to cope with the shocks [9]. Such responses were absent in a control group of animals
trained to manipulate a lever and thus to avoid shock (active avoidance).
The relationship of these observations to the fear paralysis reflex was not dis-
cussed, nor were the following observations in animals and humans, although such a
connection seems apparent.
A model in dogs for analyzing the neurophysiological pathways by which psychi-
cal stress may alter electrical instability of the heart has been provided by Lown and
associates [8]. The behavioral and cardiac responses of eight animals were compared
when in a stressful (restrained in a sling) or nonstressful (free to move in a cage) en-
vironment. The restrained dogs exhibited a significantly lower threshold for repeti-
tive ventricular extrasystoles. Enhanced sympathetic tone suggested by somatic
tremor and tachycardia in the restrained animals was also accompanied by marked
vagotonia, as indicated by anal sphincter relaxation and, in three dogs, by the occur-
rence of profound bradycardia. This followed a brief period of tachycardia upon pre-
sentation of stressful cues anticipating shock. No explanation was given for this vari-
ation, but it is assumed that the bradycardiac response in some individuals in this set-
ting is comparable to the bradycardia seen in the orienting (see below) and the fear
paralysis response to sudden, unexpected stimuli signaling threat. The conditioning
stimulus may in some individuals be interpreted as a threat to a dangerous or un-
pleasant event, in other individuals not. Whether an event is stressful or not depends
on a number of factors such as experience and a lifelong process of conditioning.
A more prolonged bradycardia developed in animals subjected to repeated ses-
sions of shock avoidance [2]. Thus, dogs manifested a progressive decrease in heart
rate and cardiac output accompanied by a progressive increase in total peripheral re-
sistance during daily sessions of an l-h anticipation period followed by 1 h of unsig-
naled shock avoidance.
These varied examples illustrate that severe bradycardia and sudden death may
occur through threatening stimuli evoking fear and in the absence of any cardiac
disease. Both the stimuli and response patterns are indicative of the fear paralysis
reflex.
The death of new-born whelps among farm-raised fur animals (mink and fox) dur-
ing sonic booms is possibly due to the elicitation of the fear paralysis reflex in the cubs.

Orienting Reaction

The fear paralysis reflex with slowed breathing and slow heart rates is in the author's
opinion a further development of the orienting response to unexpected stimuli sig-
naling threat [7,26,34,39,58]. On the other hand, tachycardia is the usual response
when the stimulus is expected. Two examples will serve to illustrate the situation:
1. Expecting an air-inflated balloon bursting evoked tachycardia when the bal-
loon was burst, whereas an immediate bradycardia ensued when the balloon was
burst unexpectedly [32]. Slowing of pulse rate also followed an unexpected auto horn
and other sharp stimuli [67, 68].
2. Clifton [7] recorded heart rate in conditioning and extinction experiments in
newborn during feeding (Fig.2). Six seconds after start of an 8-s tone (the con-
ditioned stimulus), the nipple was presented (the unconditioned stimulus). This was
associated with tachycardia (Fig. 2). However, a large sustained decrease in heart
Fear Paralysis Reflex and Sudden Cardiac Death 191

.10

.8

.6

II: .4 p..o.~
:I: I
.5 .2 ;I
.
!
CL 0 I
P
:>
c -2 '<f
'E
l -4

! -6

.s.a -8

g -10 .........
1-12 "......
\
\
Fig. 2. Orienting reactions elicited by a threat-
-14 \
\
ening or unexpected stimulus. Heart rate re-
\ sponse of conditioning and control groups on
-16 .,Conditionlng Group, Conditioning Trl. ~ the last conditioning trial and on first extinction
.Condltlonlng Group, Extinction Trl. \
.18 oControl Group, Conditioning Trl. \ • trial in newborns. Increased heart rate on pre-
OIControl Group. Extinction Trl. \ ~'
-20 ~ sentation of the nipple and deceleration on non-
<, ! , I ! I ! , , I I , I I I! I! f
appearance of the nipple (see text) (Clifton [7])
-5:+5 2.5 4.5 6.5 8.5 10.5 12.5 14.5 16.5
Post Onset Seconds

rate to the unsuspected nonappearance of the nipple was recorded, i.e., a decelera-
tion of the heart rate as a response to the threat of not obtaining food.
To this category also belong the pallid breathholding attacks associated with
bradycardia in the awake state in infants. The attacks are elicited by a sudden unex-
pected stimulus of potentially threatening nature [5]. The pallor is the result of
peripheral vasoconstriction. Cyanotic breathholding spells with acceleration of heart
rates, on the other hand, are evoked by rage reactions after frustration or a bellow
of pain with cerebral ischemia caused by a Valsalva-like maneuver [5].
To summarize these examples, it may be stated that slowing of the heart rate ap-
pears to be elicited in a subject's everyday situation by stimuli perceived as poten-
tially threatening. Depending upon the stimulus strength, the situation, and the abil-
ity to cope with the situation, the response manifests itself in orienting reactions, pal-
lid breathholding spells, or the fear paralysis reflex, and eventually sudden cardiac
death in extreme life-threatening situations [34].

Neuroanatomy and Pharmacology

The fear paralysis reflex is dependent on the integrity of nervous structures at the
midbrain-pontine level [refs. in 34]. The reflex is influenced from more rostral levels:
facilitation from limbic areas and inhibition from neocortex, in particular the frontal
lobes. The fear paralysis response is funneled through the medioventral hypo-
thalamus, the medial bulboreticular formation, and the nucleus of the solitary tract.
At this level the reflex may potentiate cardiac inhibition and apnea produced by
chemo- and baroreceptor stimulation and by impulses from the face, respiratory and
gastrointestinal tracts, and by hypoxia [refs. in 34].
192 B.Kaada

- -- Clonldlne

-SMOKE
I ! ! ; !

SEC
! )

Fig. 3. A Immediate and profound decrease in heart rate (HR) and inhibition of respiratory (RESP)
and other somatomotor movements associated with inhibition of muscular electrical activity (EMG)
t , , !

in response to a single 3-s blow of tobacco smoke into the rabbit's nostrils. Heart rate decreased from
207 to 42 beats/min, i.e., a reduction to 20% of the prestimulatory baseline level. The electromyo-
gram was recorded from flexor muscles of one hind limb . B Partial blocking of all three components
of the smoke reflex by clonidine (0.3mg/kg i.m.). (Kaada [34])

Atropine selectively blocks the bradycardia, whereas the entire reflex appears to
be mediated via a noradrenergic system and is suppressed by the aradrenoceptor
agonist clonidine and the tricyclic antidepressant amitriptyline and enhanced by the
a2-adrenoceptor inhibitor yohimbine [34, 37] (Fig. 3).
The fear paralysis reflex is associated with a temporary prolongation of the QT-
interval, a condition which makes the ventricles more vulnerable to fibrillation . This
prolongation is suppressed by atropin and /J-adrenoceptor blockers [Kaada, unpub-
lished].

Sudden Cardiac Death Triggered by Psychical Stimuli in Adults

To attribute sudden cardiac death to the fear paralysis reflex, both the stimulus and
the response should possess the characteristics of this reflex. On the stimulus side,
death should primarily be precipitated by fear evoked by a threatening situation with
which one is unable to cope. On the response side, the essential feature of the fear
paralysis is recognized first and foremost by a profound bradycardia.
Excessive parasympathetic discharges are assumed to be the cause of voodoo
death in primitive societies, resulting from superstitious beliefs and fears , when men
were condemned and sentenced by the so-called medicine man [25 , 64, 75].
Fear Paralysis Reflex and Sudden Cardiac Death 193

Richter [64] compared this death with the sudden death provoked in wild rats by
subjecting them to various life-threatening stimuli (cf. above). The ensuing hope-
lessness and helplessness or giving up involved overactivity of the parasympathetic
system.
Wolf [74-76] similarly stressed that a striking bradycardia may occur not only
in response to intense noxious stimulation of any part of the body or instrumental
manipulation of air passages, esophagus, pleura, and peritoneum, but also under cir-
cumstances interpreted as sudden fear, overwhelming, total social exclusion, or
hopeless dejection, similar to those in voodoo death.
Similarly, Szasz [70], in a bioanalytic approach to the problem of psychogenesis
of somatic change, emphasized how the multitude of physiological changes brought
about by sympathetic hyperactivity as a result of fear in response to external danger
were replaced by a quite different chain of events if the organism cannot deal with
a stressful situation through appropriate adaptive behavior. A breakdown in the
homeostatic mechanism then results in functionally specific, localized parasympa-
thetic hyperfunctions, which he called "regressive innervation." Among these
maladaptive end results is myocardial infarction.
Engel [15-17] distinguished two primary biological modes of response to danger
and unpleasure and of corresponding primal affects: (1) "flight or fight," the term
most expressive of the attitude of anxiety and (2) giving up, the attitude and expres-
sion most indicative of depression-withdrawal, associated with hopelessness or help-
lessness. The latter quiescent and withdrawn patterns of frustration have to a large
degree been overlooked [15]. These two types of response actually represent active
and passive avoidance, respectively.
Engel collected anecdotes of 170 emotional sudden deaths and classified the life
settings in which death may occur into eight categories. Settings of various kinds of
losses, overwhelming grief, or humiliation accounted for 59%: on the impact of the
collapse, death of a close person or threat of such loss, during acute grief, during
mourning, on an anniversary, and on loss of status or self-esteem. Settings of per-
sonal danger or threat of injury implying fear, or after the dangerous situation is
over, accounted for 34%, and some paradoxical situations of a "happy ending,"
reunion, or triumph accounted for the remaining 6%. Most deaths occurred within
an hour of the event reported. Common to these situations were stress-related fac-
tors: (a) being impossible for the victim to ignore because of their abrupt, unex-
pected, or dramatic quality implying threat with overwhelming excitation; (b) evok-
ing intense emotions; and (c) involving an uncontrollable situation, as inferred from
the frantic, disorganized, or paralytic behavior exhibited with failure to cope with the
situation.
Gellhorn [24] emphasized that during intense emotional excitement there may be
a change from sympathetic responses to parasympathetic dominance. A shift from
tachycardia to bradycardia with vasodepressor syncope characteristically occurs in
settings in which the impulse to flee is inhibited [17].
Anisman et al. [3] suggested that the effects of prolonged stressful experiences on
affective state may be related to depletion of monamines. A major element in deter-
mining these neurochemical changes is the organism's ability to cope with the aver-
sive stimuli through behavioral means. Aversive stimuli which are perceived as un-
controllable exacerbate affective disorder and contribute to behavioral depression.
194 B.Kaada

Parasympathetic dominance does not necessarily mean excessive parasympa-

thetic discharges. A depletion of monamines may similarly result in a parasympa-
thetic dominance by reducing the sympathetic outflow and thereby interfere with the
normal balance between the two autonomic divisions.
Schmale [65], like Engel and others, attempted to single out the psychological
settings in which illness develops. In actual loss, or threat of loss of an object, differ-
ent states of displeasure immediately preceding the onset of the disease were de-
scribed: anxiety, anger, fear, guilt, shame ("inadequacy," "failure"), helplessness,
and hopelessness. In 41 of 42 patients the investigator felt evidence for feelings of
helplessness or hopelessness prior to the onset of disease.
More recently, Reich et al. [62] observed among 117 arrhythmic patients 25 who
experienced acute emotional and psychological disturbances during the 24h preced-
ing the cardiac attack. Among the situations leading to psychological disturbances
were interpersonal conflicts, public humiliation, threat of or actual marital separa-
tion, bereavement, business failure, anger, acute depression, fear, anticipatory ex-
citement, grief, extreme agitation, and tension.
Dimsdale [13] similarly summarized a number of deaths in response to psychical
stimuli. The latter involved overwhelming stress, helplessness and hopelessness, pre-
dilection to death among surgical patients, longing for death, loss of will to live, and
life changes. The bradycardiac response to these psychical stressors in many in-
stances was discussed.
In the author's opinion such instances of sudden death evoked by acute threaten-
ing, frightful stimuli, helplessness and hopelessness, and associated with a parasym-
pathetic dominance in fact represent the atavistic innate fear paralysis reflex tem-
porarily released from cortical inhibition. The bradycardia, like the other compo-
nents of the fear paralysis response, are thus assumed to represent a disinhibitory
phenomenon.

The Terminal Stages of Cardiac Death

A condition of fear with hopelessness and helplessness is also encountered during the
terminal stages of cardiac death. Irreversible asystole most often follows ventricular
fibrillation, although primary bradyarrhythmias also have been reported [refs. in 72].
Wang et al. [72] made a survey of the English literature and found reports of a total
of 55 patients in whom sudden death occurred during continuous ambulatory Holter
monitoring. In most patients (79%) with coronary artery disease, ventricular
tachyarrhythmias dominated. The authors analyzed the terminal electrical events in
similar recordings in 23 hospitalized adults who died without apparent heart disease.
In most of them there was a gradual slowing of heart rate with shifting of cardiac
pacemaker from the sinus node and atria to the atrioventricular junction and ventri-
cles, resulting in cardiac asystole. In these patients bradyarrhythmia was the domi-
nant terminal rhythm in 83% , whereas ventricular tachyarrhythmia occurred in 17% .
These terminal electrical events became manifest from 1 to 450 min (mean 62) before
irreversible asystole. The mechanism of bradycardiac asystole in patients with no ap-
parent cardiac disease was attributed to a generalized anoxic depression, together
with neurogenic suppression of the heart. It appears likely that many of these cases
with vagal overactivity induced by fear represent the fear paralysis reflex.
Fear Paralysis Reflex and Sudden Cardiac Death 195

It also seems likely that the fear paralysis reflex may operate in the following two
situations of catastrophy: in people locked up in a fire with no possibilities to escape
and with no control of the situation and in people buried in snow in an avalanche
with no possibility to move.

Postmortem Findings

Recently, Kariks [41] reported lesions in the heart of 179 sudden infant deaths. The
lesions were described as focal myofibrillar degeneration and were present in all
cases. They represent the final manifestations of a common pathway to shock. Iden-
tical lesions have been widely described in cases of shock regardless of causative fac-
tors, and also in sudden unexpected death in human adults, in head trauma, pro-
longed stress situation, in poisoning with various drugs, in calves who collapsed and
died suddenly, and in experimental animals. What triggers or initiates this shock in
sudden deaths in infants, in human adults, and cattle is left open to research. It might
well be the psychical shock encountered in fear paralysis.

Closing Comments

The proposal advanced in the present communication that a defined group of sudden
cardiac deaths in human adults is caused by the inborn fear paralysis reflex, as re-
cently also suggested for sudden infant death, is not proven, although the indices are
strong. It is relevant, however, to quote a passage by Reichenbach and Benditt [63],
who studied the pathogenesis of heart myofibrillar degeneration in human adults:
In modern biology, hypotheses provide the framework, the target, for experi-
ment. They are not developed merely for the satisfaction of authors. Hopefully
this communication will serve such a purpose. We have no System to defend, but
only Nature with which to contend.
Finally, it may seem a paradox that the fear paralysis reflex, a behavior pattern
of survival value, should also be the cause of death. However, as emphasized by
Engel [15]:
... the responses to giving-up may be maladaptive or unsuccessful, in which case
they seem to be potent provoking conditions for major psychic and somatic disor-
ganizations, sometimes leading to death ... There is some evidence to show that
object loss, real, threatened, or phantasied, when followed by "giving-up" and
the characteristic affects of helplessness or hopelessness, may be a necessary, but
not sufficient condition for disease.
196 B.Kaada

References
1. Alderman EL (1978) Control of myocardial function by the autonomic nervous system. Adv
Cardiol 23 : 93-102
2. Anderson DE, Tosheff JG (1973) Cardiac output and total peripheral resistance changes during
preavoidance periods in the dog. J Appl Physiol 34: 650-654
3. Anisman H, Zacharko RM (1982) Depression. The predisposing influence of stress. Behav
Brain Sci 5: 89-137
4. Blix AS, Folkow B (1983) Cardiovascular adjustments to diving in mammals and birds. In:
Handbook of physiology - The cardiovascular system III, pp 917-945
5. Bower BD (1984) Pallid syncope (reflex anoxic seizures). Arch Dis Child 59: 1118-1119
6. Buell JC, Eliot RS (1980) Psychosocial and behavioral influences in the pathogenesis of acquired
cardiovascular disease. Am Heart J 100: 723-740
7. Clifton RK (1974) Cardiac conditioning and orienting in the infant. In: Obrist PA et al (eds)
Cardiovascular psychophysiology. Aldine, Chicago, pp 479-504
8. Corbalan R, Verrier RL, Lown B (1974) Psychologic stress and ventricular arrhythmia during
myocardial infarction in the conscious dog. Am J Cardiol 34: 692-696
9. Corley KC, Mauck HP, Shiel F O'M (1975) Cardiac responses associated with "yoked-chair"
shock avoidance in squirrel monkeys. Psychophysiology 12: 439-444
10. Darwin C (1900) A posthumous essay on instinct. In: Romanes GJ (ed) Mental evolution in ani-
mals. Appleton, New York, pp 360-364
11. DeSilva RA (1982) Central nervous system risk factors for sudden cardiac death. Ann NY Acad
Sci 382: 143-161
12. DeSilva RA, Lown B (1978) Ventricular premature beats, stress and sudden death. Psycho-
somatics 19: 649-659
13. Dimsdale JE (1977) Emotional causes of sudden death. Am J Psychiatry 134: 1361-1365
14. Edson PH, Gallup GG Jr (1972) Tonic immobility as a fear response in lizards (Anolis
carolinensis). Psychonom Sci 26 : 27 -28
15. Engel GL (1962) Anxiety and depression-withdrawal: the primary affects of unpleasure. Int J
PsychoanaI43:89-97
16. Engel GL (1968) A life setting conductive to illness, the giving-up complex. Ann Intern Med
69:293-300
17. Engel GL (1971) Sudden and rapid death during psychological stress. Folklore or folk wisdom?
Ann Intern Med 74:771-782
18. Espmark Y, Langvatn R (1979) Cardiac responses in alarmed red deer calves. Behav Processes
4: 179-186
19. Gabrielsen GW (1985) Free and forced diving in ducks: habituation of the initial dive response.
Acta Physiol Scand 123: 67-72
20. Gabrielsen GW (1986) Passiv frykt - en mulig arsak til plutselig uventet spedbarndj.}d? Tidsskr
Nor Lregeforen 106: 898-902,962
21. Gabrielsen GW, Kanwisher J, Steen JB (1977) "Emotional" bradycardia: a telemetry study on
incubating willow grouse (Lagopus lagopus). Acta Physiol Scand 100: 255-257
22. Gallup GG Jr (1977) Tonic immobility: the role of fear and predation. Psychol Record 1 :41-61
23. Gallup GG Jr, Nash RF, Potter RJ, Donegan NH (1970) Effect of varying conditions of fear on
immobility reactions in domestic chickens (Gallus gallus). J Comp Physiol Psychol 73: 442-445
24. Gellhorn E (1970) The emotions and the ergotropic and trophotropic systems. Psychol Forsch
34:48-94
25. Goodfriend M, Wolpert EA (1976) Clinical case report: death from fright. Report of a case and
literature review. Psychosom Med 38: 348-356
26. Graham FK, Clifton RK (1966) Heart-rate change as a component of the orienting response.
Psychol Bull 65: 305-320
27. Groover M, Stout C (1966) Neurogenic myocardial necroses. In: Raab W (ed) Prevention of
ischemic heart disease. Thomas, Springfield, pp 57-66
28. Harrison RJ, Ridgway SH, Joyce PL (1972) Telemetry of heart rate in diving seals. Nature 238:
280
29. Riggings CB, Vatner SF, Braunwald E (1973) Parasympathetic control of the heart. Pharmacol
Rev 25 : 119-155
Fear Paralysis Reflex and Sudden Cardiac Death 197

30. Hofer MA (1970) Cardiac and respiratory function during sudden prolonged immobility in wild
rodents. Psychosom Med 32: 633-647
31. Hofer MA, Reiser MF (1969) The development of cardiac rate regulation in preweanling rats.
Psychosom Med 31: 372-388
32. Hunt TJ (1970) Tachycardia and bradycardia associated with particular forms of stimuli. J Physiol
(Lond) 21O:64P
33. Jacobsen NK (1979) Alarm bradycardia in the white-tailed deer fawns. J Mammal 60:343-349
34. Kaada B (1986) Sudden infant death syndrome - the possible role of "the fear paralysis reflex".
Norwegian University Press, Oslo, 56 pp
35. Kaada B (1986) Krybbedjild: Hjertedjild utljilst ved "fryktparalyserefleksen." Tidsskr Nor
Lregeforen 106: 894-898, 962
36. Kaada B (1987) The sudden infant death syndrome induced by "the fear paralysis reflex"? Med
Hypotheses 22: 347-356
37. Kaada B (1987) Neurotransmitters in "the smoke reflex" in rabbits. Gen PharmacoI18:61-68
38. Kaada B, Gundersen R (1987) 0kt risiko for krybbedjild i helgene? Tidsskr Nor Lregeforen 107:
755-757
39. Kagan J, Lewis M (1965) Studies of attention in the human infant. Merill-Palmer Quarterly 11:
95-127
40. Kanwisher JW, Gabrielsen G, Kanwisher N (1981) Free and forced diving in birds. Science 211:
717-719
41. Kariks J (1985) Is shock the mode of death in SIDS? Med Hypotheses 18:331-349
42. Kayssi AI (1948) Death from inhibition and its relation to shock. Br Med J 2: 131-135
43. Kerzner J, Wolf M, Kosowsky B et al. (1973) Ventricular ectopic rhythms following vagal stimu-
lation in dogs with acute myocardial infarction. Circulation 47: 44-50
44. Leyhausen P (1967) On the natural history of fear. In: Lorenz K, Leyhausen P (eds) Motivation
of human and animal behavior. Van Nostrand Reinhold, New York
45. Lombardi F (1986) Acute myocardial ischaemia, neural reflexes and ventricular arrhythmias.
Eur Heart J [Suppl A] 7: 91-97
46. Lown B, Tykocinske M, Garfein A, Brooks P (1973) Sleep and ventricular premature beats.
Circulation 48: 691-701
47. Lown B, Temte JV, Reich P, Gaughan C, Regestein Q, Hai H (1976) Basis for recurring ven-
tricular fibrillation in the absence of coronary heart disease and its management. N Engl J Med
294:623-629
48. Lown B, Verrier RL, Rabinowitz SH (1977) Neural and psychologic mechanisms and the prob-
lem of sudden cardiac death. Am J Cardiol 39: 890-902
49. Lown B, DeSilva RA, Lenson R (1978) Roles of psychological stress and autonomic nervous
system changes in provocation of ventricular premature complexes. Am J Cardiol 41: 979-985
50. Lown B, DeSilva RA, Reich P, Murawski BJ (1980) Psychophysiologic factors in sudden cardiac
death. Am J Psychol137: 1325-1335
51. McGraw CP, Klemm WR (1969) Mechanisms of the immobility reflex ("animal hypnosis"). III.
Neocortical inhibition in rats. Commun Behav Bioi 3: 53-59
52. MacWilliam JA (1887) Fibrillar contraction of the heart. J Physiol (Lond) 8: 296-310 [See also
Br Med J (1973) 2: 215-219]
53. Manning GW, Hall GE, Banting FG (1937) Vagus stimulation and the production of myocardial
damage. Can Med Assoc J 37:314-317
54. Manning JW, Cotton M deV (1962) Mechanism of cardiac arrhythmias induced by diencephalic
stimulation. Am J Physiol 203: 1120-1124
55. Melville K (1966) Cardiac ischemic changes induced by CNS stimulation. In: Raab W (ed) Pre-
vention of ischemic heart disease. Thomas, Springfield, Ill, pp 31-38
56. Nash RF, Gallup GG Jr, McClure MK (1970) The immobility reaction in leopard frogs (Rana
pipiens) as a function of noise-induced fear. Psychonom Sci 21: 155-156
57. Norvenius SG (1984) The contribution of SIDS to infant mortality trends in Sweden. Depart-
ment of Pediatrics, University of Goteborg, Report 1984. Goteborg, Sweden, 21 pp
58. Obrist PA (1968) Heart rate and somatic-motor coupling during classical aversive conditioning
in humans. J Exp Psychol 77: 180-193
59. Olsen CR, Fanestil DD, Scholander PF (1962) Some effects of apneic underwater diving on
blood gases, lactate, and pressure in man. J Appl Physiol17: 938-942
198 B. Kaada: Fear Paralysis Reflex and Sudden Cardiac Death

60. Ratner SC (1967) Comparative aspects of hypnosis. In: Gordon JE (ed) Handbook of clinical
and experimental hypnosis. Macmillan, New York
61. Ratner SC, Thompson RW (1960) Immobility reactions (fear) on domestic fowl as a function of
age and prior experience. Anim Behav 8: 186-191
62. Reich P, DeSilva RA, Lown B, Murawski BJ (1981) Acute psychological disturbances preced-
ing life-threatening ventricular arrhythmias. JAmMed Assoc 246: 233-235
63. Reichenbach DD, Benditt EP (1970) Catecholamines and cardiomyopathy: the pathogenesis
and potential importance of myofibrillar degeneration. Human Pathol1 : 125-150
64. Richter CP (1957) On the phenomenon of sudden death in animals and man. Psychosom Med
19:193-200
65. Schmale A (1958) Relationship of separation to disease. Psychosom Med 20: 259-277
66. Schwartz PJ, Zaza A (1986) The rational basis and the clinical value of selective cardiac sym-
pathetic denervation in the prevention of malignant arrhythmias. Eur Heart J 7: 107-118
67. Shock NW, SchUitter MJ (1942) Pulse rate response of adolescents to auditory stimuli. J Exp
PsychoI30:414-42S
68. Skaggs EB (1926) Changes in pulse, breathing and steadiness under conditions of startledness
and excited expectancy. J Comp PsychoI6:303-318
69. Steininger F (1936) Die Biologie der sogenannten "tierischen Hypnose". Ergeb Bioi 13 : 348-451
70. Szasz TS (1952) Psychoanalysis and the autonomic nervous system. Psychoanal Rev 39: 115-151
71. Tavazzi L, Zotti AM, Rodanelli R (1986) The role of psychologic stress in the genesis of lethal
arrhythmias in patients with coronary artery disease. Eur Heart J [Suppl A] 7:96-106
72. Wang FS, Lien WP, Fong TE, Lin JL, Cherng JJ, Chen JR, Chen JJ (1986) Terminal cardiac
electrical activity in adults who die without apparent cardiac disease. Am J Cardiol 58: 491-495
73. Whayne TF Jr, Killip T (1967) Simulated diving in man: comparison of facial stimuli and re-
sponse in arrhythmia. J Appl Physiol 22: 800-807
74. Wolf S (1964) The bradycardia of the dive reflex - a possible mechanism of sudden death. Trans
Am Clin Climatol Assoc 76: 192-200
75. Wolf S (1967) The end of the rope: the role of the brain in cardiac death. Can Med Assoc J 97:
1022-1025
76. Wolf S (1969) Psychosocial forces in myocardial infarction and sudden death. Circulation [SuppI4]
49:74-83
Chapter 14

Some Clinical Neurological Aspects of Syncope

O. J OAKIMSEN

Introduction

The etiology of syncope, defined as a transitory loss of consciousness due to a tempo-

rary fall of cerebral blood flow, is usually easy to recognize. In about 90% of the
cases it is due to various well-known reflex mechanisms [2]. Consequently, such pa-
tients are normally taken care of by general practitioners. However, when the etiol-
ogy is unknown, and the circumstances of the syncope are uncertain, the diagnostic
problems might be a great challenge to the medical staff. For neurologists the most
frequent problem in that respect is to distinguish syncopes from attacks of neurologi-
cal nature - particularly epileptic seizures. This short survey presents some of the
most important aspects of neurological differential diagnoses of syncope (Table 1).

Differential Diagnosis

An ordinary syncope without presyncopal symptoms can sometimes be confused

with atonic epileptic seizures. However, such seizures are very rare, and when a
careful history is taken, the patients often have a known history of epilepsy, and they
have usually experienced one or a few myoclonic jerks just before falling to the
ground.

Table 1. Some neurological differential diagnoses of syncope

Epilepsy
Complex partial seizures
Atonic/ akinetic
GTC
Narcolepsy/cataplexy
TIA (post. territory)
Loss of consciousness
Drop attacks
Basilar artery migraine
"Bulbar syncopes"
Psychogenic
200 o. Joakimsen
A convulsive syncope, sometimes combined with tongue-biting or urinary incon-
tinence, can be difficult to distinguish from a general tonic-clonic (GTC) epileptic
seizure. The former usually occurs only after more than 10 s of lost or disturbed con-
sciousness [2], while a typical GTC seizure starts with convulsions, followed by an
atonic state which is the reversed order of events compared with syncope.
Another differential diagnostic problem is how to distinguish between presyn-
copal symptoms and manifestations of complex partial epileptic seizures. In both
states there can be epigastrial discomfort, dizziness, excessive sweating, pallor or
flushing of the skin, and disturbed consciousness. However, the typical epileptic
symptoms of complex partial seizures, for example, focal motor, sensory, and ver-
sive manifestations, are usually not present in the presyncopal states. The same con-
cerns characteristic epileptic symptoms such as olfactory and gustatory hallucina-
tions, deja vu manifestations, dreamy states, and automatisms with semi- or
quasipurposeful movements. Blurring of vision and buzzing in the ears, typical of
presyncope, are not like the more vivid visual and auditive hallucinations of complex
partial seizures. The disturbed consciousness is more like a confusion and disorienta-
tion than the quantitatively reduced level of consciousness typical of a presyncope.
Hence, a patient undergoing a complex partial seizure often does not respond to
questions or other attempts at communication despite being apparently awake and
responsive.
Some recent reports from long-term ambulatory ECG/EEG recordings on pa-
tients undergoing complex partial seizures have shown that almost all patients had
ictal tachycardia, and approximately one-half of them had cardiac arrhythmias dur-
ing the seizure as well [1]. Such secondary autonomic effects of epilepsy may well be
the cause of loss of consciousness in some epileptic patients - and may be one of the
explanations of sudden unexpected death of epileptics [1, 4, 6]? Conversely, it is im-
portant to note what neurologists not infrequently experience - that any cerebral
ischemic event can be responsible for trigging an epileptic focus and thereby provoke
a seizure.
Both narcoleptic and cataplectic attacks are sometimes confused with syncope.
The typical history and clinical examination usually make the correct diagnosis pos-
sible.
Sudden loss of consciousness can in some cases be the only manifestation of tran-
sient ischemic attacks from the territory of the vertebrobasilar arteries. Other ictal
and postictal focal signs and symptoms from this region of the central nervous system
usually indicate the correct diagnosis. Drop attacks are not uncommon events. They
are considered to be caused by ischemic episodes of the brainstem. The conscious-
ness is not disturbed.
Syncope associated with basilar artery migraine is probably caused by constric-
tion of arteries of the posterior cerebral circulation. The other typical concomitant
migrainous symptoms usually cause no diagnostic problems versus ordinary syncope.
So-called bulbar syncopes are caused by various brain lesions involving the bul-
bar centers that regulate heart rate and vasomotor functions (neoplasms, amyotrophic
lateral sclerosis, syrinx, sequelae of vascular insults, etc.).
Unconsciousness of psychogenic origin frequently occurs and is likely to be con-
fused with syncope. Such attacks are often vivid and dramatic with a wide range of
bizarre events and usually occur in the presence of an audience. The subjects almost
Some Clinical Neurological Aspects of Syncope 201

never hurt themselves when falling, and there is no skin pallor or flushing or other
autoriomic disturbances. The history often reveals a psychiatric etiology.

Conclusion

Although syncope usually is a benign disorder, a sudden loss of consciousness some-

times heralds serious disturbances varying from life-threatening heart arrhythmias to
severely stenosed neck arteries. Investigations have suggested that the history and
physical examination are the most useful aids in establishing a cause of syncope [3],
and also in distinguishing between syncopes and similar episodes of neurological
origin. Syncopes occurring in the absence of prior or recurrent neurological, coro-
nary, or other overt cardiovascular disease stigmata are not associated with increased
morbidity and mortality and do not normally need full cardiological or neurological/
neurophysiological investigations [3, 5].
Sometimes, however, long-term ambulatory EeG/EEG recordings may be valu-
able in establishing the underlying cause when the episodes of unconsciousness are
of uncertain origin and the seizures are a nuisance to the patient.

References

1. Blumhardt LD (1986) Electrocardiographic accompaniments of temporal lobe epileptic seizures.

Lancet 1:1051-1056
2. Gastaut H (1974) Syncopes: generalized anoxic cerebral seizures. In: Vinken PS, Bruyn GW
(eds) Handbook of neurology, vol 15. North-Holland, Amsterdam, pp 815-835
3. Kapoor WN, Karpf M, Wieand S, Peterson JR, Levey GS (1983) A prospective evaluation and
follow-up of patients with syncope. N Engl J Med 309: 197-204
4. Kiok MC, Terrence CF, Fromm GH, Lavine S (1986) Sinus arrest in epilepsy. Neurology 36:
115-116
5. Savage DD, Corwin L, McGhee DL, Kannel WB, WolfPA (1985) Epidemiologic features of iso-
lated syncope: the Framingham Study. Stroke 16:626-629
6. Terrence CF, Rao GR, Perper JA (1981) Neurogenic pulmonary edema in unexpected, un-
explained death of epileptic patients. Ann Neurol 9: 458-464
Chapter 15

Differential Diagnosis in Syncope and Epilepsy:

Clinical Neurophysiological and Cardiological Aspects
LA.SULG

Most paroxysmal disturbances of consciousness can be directly diagnosed either as

epilepsy or as cardiogenic syncopes. There remain, however, a considerable number
of unspecified syncopes without conclusive findings in the primary clinical evalua-
tion. In these cases usually the history is not specific and neither ECG nor EEG ex-
aminations have resulted in significant findings. Usually at this stage neither epilepsy
nor cardiogenic syncope can be excluded. Many of these patients have passed several
admissions to hospital, making them quite expensive for the public health service
[51].
In recent decades there has been a move toward increasing application of long-
term monitoring techniques [4, 5, 49, 51]. This approach has been quite rewarding,
particularly in documenting cardiac arrhythmias and cerebral seizures not seen in
casual recordings. However, there is still some lack of efficacy in the applications due
to the practice of conducting ECG and EEG monitoring separately from each other.
The most rational way to achieve a definite diagnosis would be a simultaneous
monitoring of both variables. Furthermore, the monitoring procedure may in some
cases be shortened by means of appropriate activation or provocation procedures,
such as exercise and orthostatic tests for the cardiovascular system, and prolonged
sleep deprivation combined with pharmacological provocation of latent epileptogenic
activity.
Much could be learned if both cardiac and cerebral activity were recorded simul-
taneously. This would for instance create opportunities to observe that some cardiac
arrhythmias, which had been assumed to be primary cardiac, in reality were auto-
nomic paraphenomena of cerebral discharges in focal epilepsy. It could also be
learned that it would be wrong always to try to cover all symptoms with only one
diagnosis. Some patients may have both epileptic and nonepileptic syncopes.
At present clinical neurophysiology seems to have the best resources for long-
term multivariable monitoring (MVM). Diagnostic problems in differentiating non-
epileptic syncopes from epileptic fits, as well as problems in sleep diagnostics and re-
search have resulted in laboratories equipped for long-term MVM including EEG,
ECG, noninvasive recording of respiration, blood pressure, O2 saturation, CO2 pro-
duction, as well as eye and limb movements. For this purpose a portable 4-8 channel
tape recorder is used. Recorded variables can be visualized by means of a special
video unit and analyzed with up to six times the recording speed. This equipment has
been further developed to perform automatic detection of epileptogenic events in
EEG and to identify/quantify different sleep stages.
Differential Diagnosis in Syncope 203

Interactions Between Cerebral and Cardiac Activity

in Physiological and Pathological Conditions

The interdependence of the brain and heart is intimate. All energy required by the
brain can only be delivered by and through the heart. Optimal function of the cardio-
vascular system is possible only by means of proper neuronal and hormonal control.
Thus, along the vital channels between brain and heart there is a continuous flow of
physiological commands, parallel with feedback information from the target organs.
The central autonomic regulation of the cardiovascular system operates from a
neural hierarchy resembling that which governs the somatic sensory motor system.
Neural systems lying in the forebrain, hypothalamus, brainstem reticular formation,
spinal cord, and peripheral ganglia all influence cardiac rhythm and rate.
At the spinal level, sympathetic preganglionic fibers arise in the intermediolateral
columns of the first five thoracic segments and pass from there to innervate the three
cervical and upper four or five thoracic sympathetic ganglia. From these ganglia post-
ganglionic fibers proceed to the cardiac plexus. There they produce predominantly
cardiac acceleration and arterial vasodilatation. Parasympathetic fibers arise in man
primarily or entirely from the dorsal motor nucleus of the vagus nerve in the medulla
oblongata. The descending vagus nerves innervate the intercardiac ganglia of the
atria and parts of the ventricles. Vagal action produces cardiac slowing by suppres-
sing the excitability of the SA node; when intense or in the presence of hypoxia, such
parasympathetic stimulation can produce an asystole [29, 62, 78].
At the medullary level of the brainstem, afferent fibers from the vagus as well as
from the baroreceptors and chemoreceptors of the carotid sinus project largely onto
the nucleus solitarius complex. The presence of tachycardia, hypertension and liabil-
ity to arrhythmias in patients with severe peripheral neuropathy can generally be at-
tributed to involvement of these afferent nerves [62]. Major sympathomimetic vaso-
motor influences arise from neuronal areas located in the reticular formation of the
pontomedullary region extending over an area from the lower end of the pons down
to approximately the level of the medullary obex. Within this region, a median zone
subserves largely vasodepressor influences, while stimulation of a somewhat less de-
fined more dorsal and lateral region induces systemic vasoconstriction and cardiac
acceleration [29, 62, 78]. Some evidence indicates that the lateral vasomotor area
projects heavily toward the medial region which relays the ultimate vasomotor out-
flow to the spinal cord. Tiny regions within medial and lateral pontomedullary areas
possess substantial functional discrimination. These structures project via the lateral
spinal funiculus to the intermediolateral columns of the spinal cord with excitatory
fibers lying somewhat more dorsal than vasodepressor fibers [62, 78].
Descending pathways from the forebrain converge in the hypothalamus in the
ventromedial (inhibitory) and ventrolateral (excitatory) areas where interneurons
integrate these impulses with other afferent messages received from the diencepha-
lon and brainstem and, in tum, discharge their descending fibers to the lower brain-
stem and cerebellum. As mentioned, the main integrating structures and efferent
relay from the brainstem to the periphery are located in the reticular formation of
the pontomedullary hindbrain [26, 29, 62, 78].
Veda [109] demonstrated in dogs that nearly all varieties of cardiac arrhythmia
can be induced by electrical stimulation of the brain. The more rostral the stimula-
204 I.A.Suig

tion target, the more dominant are tachyarrhythmias. The lower the stimulation
level, the more depressive and inhibitive are the effects, occasionally causing cardiac
arrest. Some cardiac reactions appeared immediately upon stimulation, while others
occurred after some latency. This bimodal appearance of arrhythmias was inter-
preted as follows: (1) the immediate reaction is a direct central effect along neural
pathways (predominantly sympathetic), and (2) the delayed reaction is an indirect
effect, mediated by reflexes and released catecholamines. This assumption was
tested and confirmed .by analysis of catecholamines in the blood. Changes in blood
pressure and other hemodynamic parameters also showed a bimodal trend. The
brain thus seems to possess resources to affect and control cardiovascular functions
either neurally with short latency, or more smoothly via humoral control. Both ways
are probably utilized in parallel in physiological control of cardiovascular and hemo-
dynamic adaptations.
Abundant clinical evidence also points to the capacity of the central nervous sys-
tem to alter the cardiac rate and rhythm, especially during periods of stress or injury.
At the forebrain level, modest changes in cardiac rate and rhythm are an everyday
subjective experience during anxiety and other conditioned responses. More serious
arrhythmias, including permanent asystole, have also been reported and are presumed
to represent severe and inappropriate conditioned reflex responses to real or imagined
external threats [63]. An extensive literature describes the association between in-
flammatory and destructive diseases of the brainstem and severe arrhythmias [25].
Clinicians have also recognized that most patients with severe subarachnoid
hemorrhage have abnormal ECG tracings. Arrhythmias are most frequent in case of
concurrent hypoxemia. Ventricular tachycardia has been recorded, and neurogenic
cardiac arrhythmia has been incriminated as the probable mechanism of syncope and
sudden death in patients with acute subarachnoid hemorrhage [74]. Intense parasym-
pathetic-sympathetic stimulation combined with an increased level of circulating cat-
echolamines is responsible for these changes [29,78, 109]. Findings in experimental
animals suggest that arrhythmias accompanying acute subarachnoid hemorrhage re-
flect massive autonomic discharges resulting from a sudden increase in intracranial
pressure, while more delayed changes in the ECG and cardiac rhythm are the result
of increased circulating catecholamines [29, 64, 78, 109].
Cardiac arrhythmias are also found in association with acute stroke, partly as a
consequence and partly as a cause [25, 35]. A study by Dimant and Grob [16] indi-
cated that 90 of 100 patients with acute stroke had abnormal ECGs. Of these, 14
demonstrated conduction defects and 21 auricular fibrillation. Only 5% of the pa-
tients had ECG changes compatible with recent myocardial infarction. It may be as-
sumed that many of the rhythm changes were neurogenic in origin.

Primary Cardiogenic Unconsciousness

In different series cardiac arrhythmias cause between 30% and 60% of syncopes.
This variation is primarily a result of the definition of syncope applied in each series.
The more selective the series and the more severe the syncopes, the greater the pro-
portion of cardiac arrhythmias. The role of ventricular tachyarrhythmias as a cause
of syncopes and sudden death is reviewed by Breithardt et al. (Chapter 12) and some
Differential Diagnosis in Syncope 205

aspects of supraventricular arrhythmias are discussed by Ross et al. (Chapter 9). In

this chapter the relationship between cerebral function and arrhythmias is discussed
and some aspects of bradyarrhythmias and syncope are briefly reviewed.

Cerebral Function and Cardiac Arrhythmias

When the heart rate either increases or descreases, a series of adaptive mechanisms
both in the heart and the peripheral circulation keep the cardiac output from falling.
When the heart rate exceeds 180-200 or becomes lower than 30-40, a reduction in
cardiac output will ensue. However, if the hemodynamic cardiac reserves are re-
duced, the zone with a maintained cardiac output is greatly reduced.
In order to influence consciousness a reduction in cardiac output must cause a
subsequent fall in the cerebral blood flow (CBF). Again, a number of compensating
mechanisms may come into play. Normally, the cerebral arterial bed possesses a con-
siderable capacity for relaxing its resistance vessels and thereby maintaining flow.
During an acute increase in demand in brain metabolism, CBF can increase by as
much as 400% -500%. During chronic conditions CBF rarely increases to more than
twice normal. The studies by Eisenberg et al. [20] in patients with heart failure indi-
cate that in a chronic state, CBF may be reduced by as much as 20% without produc-
ing cerebral insufficiency. Reductions of 50% or more were consistently associated
with mental confusion, syncope, and a poor prognosis.
The brain has essentially no metabolic reserves and requires a constant supply of
glucose and oxygen to maintain its normal functions. The net cerebral oxygen re-
quirement is remarkedly constant in normals at about 3.5 cc O 2/100 mg brain/min.
In order to maintain this rate of metabolism about 55 cc of arterial blood must per-
fuse each 100 g of brain/min. The total brain blood flow of 700-800 cc/min represents
between 15% -20% of the normal resting cardiac output. A fall in oxygen uptake by
the brain exceeding 20% is usually associated with loss of consciousness [78].
In addition to the heart rate and the adequacy of the adaptive mechanisms, the
time factor is well known to be crucial for the development of cerebral symptoms
during arrhythmias. Both the adaptive mechanisms protecting the cardiac output and
those protecting CBF are highly time dependent. In addition, the position of the sub-
ject may be important. In the supine position an asystole lasting 6-10 s usually causes
syncope, in erect subjects the time is substantially shorter. In contrast to much previ-
ous and current thinking, the mechanism of the arrhythmia is not essential for the
consequences on cerebral function. Regardless of the mechanism causing bradycardia
or tachycardia, if the rates are identical the cerebral consequences will be so.
One of the few studies of CBF in arrhythmias has been done by Stem and Lavy
[98]. They studied 16 patients with chronic atrial fibrillation and a normal heart rate.
They claimed that the CBF was reduced compared with age matched normal controls
and reported a normalization of CBF following reversion to sinus rhythm. These re-
sults need substantiation, and similar studies ought to be done in other arrhythmias.

Role of Bradyarrhythmias in Syncopes

Atrioventricular (AV) block as a cause of syncope in Adams-Stokes syndrome is well
described, including its natural cause of disease without pacing and the improvement
206 I. A . Sulg

Per cen t
time

//
20

IS
Fig. I. Spectral distribution
of quantitated electroencephalo-
graphic activity in patients suffer-
10 ing from A V block before (dotted
line) and after (solid line) implan-
tation of pacemaker. The shaded
area indicates EEG activity in
healthy controls. From Sulg et al.
[101]
4 5 6 7 8 9 10 II 12 13 l4 15 16 17 18 19 20
Frequ ency in ci s

caused by adequate pacing [48, 49, 97]. Significant hypoxic EEG changes have been
described in conjunction with the turning off of cardiac pacemakers in such patients
[98]. In another study six patients with complete A V block and heart rate between
30 and 45 per min were subjected to regional CBF and quantitated EEG studies [99,
101] . The regional CBF in the group was reduced by 21 .5% and the hemispheric cor-
tical flow by 25.9%. Also, the quantified EEG showed clear-cut deviations from nor-
mal. Following the implantation of a fixed rate ventricular pacemaker with a rate of
72 per min the flow measurements increased by about 10% and the quantified EEG
showed a shift towards the normal pattern (Fig. 1) . However, in both types of mea-
surements residual abnormalities were still observed, indicating permanent cerebral
damage due to the previous bradycardia.
In a number of patients with bradycardia and chronic cerebral symptoms, it may
be difficult to judge whether pacemaker implantation may improve the symptoms.
Since improvements due to pacing occur rapidly, such a dilemma can be solved
through a period of extended test pacing.
Sinoatrial (SA) disease as a cause of syncope is well defined [80]. It is also well
established that, due to the presence of alternative pacemakers when the SA node
fails , the prognosis is much better than in A V block [81 , 94]. Thus in series of pa-
tients with syncope, SA disease is found mainly in the benign end of the spectrum.
The cerebral consequences are therefore less severe. A particular feature of SA dis-
ease is the tendency to concomitant episodes of supraventricular tachycardia. Thus,
in individual patients it may be difficult to know whether it is the bradycardic or the
tachycardic component which induces most symptoms. Since the tachyarrhythmic
component of the syndrome may increase the risk of cerebral emboli from the heart
[24], it should be recalled that neurological findings in such patients are not always
caused by the bradycardia as such.
The methods of investigating patients with syncopes and possible bradyarrhyth-
mias are essentially the same as those used in tachyarrhythmias. However, the role
of invasive electrophysiology in deciding whether to implant a pacemaker or not is
currently rather small in such patients. Thus, various ECG monitoring techniques
Differential Diagnosis in Syncope 207

and the correlation of long pauses with symptoms remain the essential methods of
evaluation.
A complication of ventricular pacing in patients with SA disease is the so-called
"pacemaker syndrome" [27, 66]. The syndrome is found predominantly in patients
with intact retrograde A V conduction causing atrial contractions against closed A V
valves. This gives rise to cannon waves in the atria and the large veins, probably
releasing an atrial vasodilatory reflex. Thus, the syndrome is another interesting
example of the role of the nervous system in the modulation of symptoms in cardiac
disease.

Atrial Natriuretic Factor

The recent discovery of atrial natriuretic factor (ANF) has opened a new view of
physiological interactions between the heart and other organ systems [87]. Increased
pressure within the atria is the main factor causing ANF release from the heart. ANF
opposes the action of vasopressin in the kidneys, an action matched centrally by its
potent inhibitory effects on vasopressin release. Its renal effects are supplemented by
an effect on the brain which inhibits water and salt intake. Recent studies have re-
vealed that ANF also interacts with brain dopamine and with different neuropeptide
systems. In total ANF is a powerful and flexible controller of electrolytes and body
fluid volume. Its effects are mediated by specific ANF-binding receptors in the ner-
vous system, mainly in the periventricular thalamus, the suprachiasmatic nucleus,
the paraventricular nucleus, and afferent vagal nerves. The role of ANF as a
mediator of diureses in acute supraventricular tachyarrhythmias is already substan-
tiated. Changes in plasma ANF concentrations as a consequence of changing from
sequential to ventricular cardiac pacing have also been described [22]. Further work
is warranted in order to establish whether an overshoot of ANF release may contri-
bute to the mechanisms of syncope and vasovagal responses.

Primary Neurogenic Unconscousness

Epilepsy

The most decisive diagnostic finding for epilepsy is either a typical, clinically witnes-
sed seizure or epileptiform discharges in EEG. In cases of grand mal, there will al-
most never be any diagnostic uncertainty. A type of epileptic seizure that may give
rise to problems in differential diagnosis versus syncope is an astatic or akinetic sei-
zure, characterized by a sudden slumping to the ground, combined with short unre-
sponsiveness. If these seizures occur in a supine position, they are accompanied by a
generalized muscular hypotonia, unresponsiveness, and eventually also brief body
spasm. The epileptogenic discharges in EEG are diagnostic and may be manifest also
interictally.
Petit mal seizure may also sometimes be confusing, but the EEG always shows
typical seizure pattern with bilateral synchronous 3/s spike-and-wave complexes.
208 I.A.Sulg

Pallor, sweating, and cardiac bradyarrhythmias, typical for syncope, are absent in
astatic and petit mal seizures.
Focal epilepsies hardly ever cause seizure patterns which could be mistaken for
syncope. In complex focal seizures some automatisms usually point to the character-
istic release of behaviors associated with a lower level of brain functions, emanating
mainly from limbic structures. Ictal EEG shows nearly always epileptogenic abnor-
malities, and often interictal discharges are present.
It is essential to emphasize that a normal EEG during a seizure never can be
taken as absolute evidence against epilepsy. There are rare epileptic seizures in
which the epileptogenic discharges in the depth of the brain do not reach scalp elec-
trodes [112].
Generally, epilepsy is often over-diagnosed and many patients, especially chil-
dren, have the label wrongly applied to them. Jeavons [46] found that 93 of 470 pa-
tients seen at epilepsy clinics in Birmingham did not have the suspected epilepsy. Of
these, 35 had cardiogenic syncope and 39 had attacks of psychological origin. Other
had migraine, breathholding spells, night terrors, or narcolepsy [46]. Routine short-
term EEG examination rarely gives the final diagnosis, and therefore long-term
monitoring is needed. For correct classification of epileptic seizure documentation of
behaviors by cine- or video recording is also important. Several EEG laboratories
now employ long-term cassette and/or video monitoring routinely.

Narcolepsy

The cataplectic fraction of the narcoleptic syndrome can be mistaken for syncope.
Cataplexy, a sudden generalized muscle weakness, is often precipitated by a strong
emotional reaction, such as surprise, mirth, anger, or a sudden need to laugh. Some-
times a narcoleptic sleep attack may follow the cataplexy or mimic a syncope. An in-
dividual in narcoleptic sleep can be easily awakened to full consciousness. An indi-
vidual in epileptic postictal unconsciousness awakes slowly and is usually dis-
oriented. In narcoleptic sleep and in cataplexy EEG often shows a pattern character-
istic of rapid eye movement (REM) sleep and hardly ever epileptogenic discharges.
Also the presence of other fractions of the syndrome (sleep attacks, sleep paralysis,
hypnagogic hallucinations) in the case history rules out syncope. However, a patient
may have both narcolepsy and epilepsy.

Ischemic-Hypoxic Syncopes Secondary to Neurogenic InOuences

Epileptogenic Events
The effect of epileptogenic events in the brain on the cardiovascular functions in-
cludes vasomotor reactions, blood pressure changes, and alterations in cardiac rate
and rhythm [44, 77, 110]. These secondary effects of epilepsy are not widely known.
We recorded EEG, ECG, and cardiotachygram simultaneously during 160 epileptic
seizures and a much larger number of subclinical epileptiform EEG discharges [99,
100, 102-104]. The experiences can be summarized as follows: Nearly all seizures,
especially in temporal lobe epilepsy, induced alterations in heart rate and rhythm,
Differential Diagnosis in Syncope 209

most often tachyarrhythmias, but also extrasystoles, and SA and A V conduction

blocks. This is in accordance with other published reports [7, 44, 77, 110]. Most of
these arrhythmias do not have any clinical relevance. In temporal lobe epilepsy,
however, occasionally more pronounced changes in heart rhythm including brady-
arrhythmias may occur, but asystole is rare [4, 15]. In experimental studies cortical
stimulation usually evokes cardiovascular reactions with tachycardia and vasopressor
trends, rarely bradycardia, and never asystole. Therefore it seems probable that trig-
gering mechanisms for the latter type of arrhythmias in epileptics with cortical focus
in temporal lobe are mediated to the heart via limbic, subthalamic, and medullary
structures. The lower the final interfering structures in CNS, the more pronounced
the trend to bradyarrhythmias. Thus one has to take into account the possibility that
a patient can have both an epileptic seizure and a secondary cardiogenic syncope fol-
lowing the seizure. This must, however, be rare.
A more frequently occurring combination is seen in epileptics who have interictal
syncopes independent of epileptogenic events in the brain, often not recognized as
different from the epileptic fits. The fits, which do not disappear by medication, are
perhaps seen as "residual" epileptic seizures, intractable by antiepileptic drugs. In
my series there were four patients of this type. Two of them had intermittent A V
block, one had SA block, and the fourth had attacks of glossopharyngeal neuralgia,
accompanied by extreme bradyarrhythmias. In all four patients polygraphic record-
ing was central for diagnosis. A more serious situation may arise when status epilep-
ticus is complicated by asystole or tachyarrhythmia. Such a dangerous coupling was
observed only once, and simultaneous recording of EEG and ECG enabled detec-
tion of the asystole in time.
In conclusion, EEG examination without simultaneous ECG recording is no
longer "state of the art." Epileptics have a high risk of sudden death and up to 17%
of such deaths are unexplained at autopsy [38, 45, 55, 106]. Young patients with sub-
therapeutic serum levels of anticonvulsant drugs have been said to be at particular
risk. Some of these deaths may be due to arrhythmias. The fact that epilepsy can
cause cardiac arrhythmias has some important implications for cardiologists. They
should be cautious in concluding from long-term ECG monitoring alone that cardiac
arrhythmias are the primary and only cause of actual symptoms. Furthermore, it may
not be necessary to proceed to complicated electrophysiological studies of the heart
if the coupling between EEG and ECG events has been documented.

Respiration
The effect of respiration on heart rate and rhythm must also be taken into account.
Cardiac arrhythmias may be neither primary cardiogenic nor neurogenic, but may be
caused by respiratory disturbances, especially in sleep apnea syndrome. Patients may
have various arrhythmias in connection with obstructive apneas lasting 1-2 min (see
Chaps. 21 and 22). Also entirely normal individuals may have transitory arrhythmias
in REM sleep. It is a common experience that sudden, unexpected sensory impact,
especially sudden acoustic effects, may cause changes in heart rate and rhythm. The
so-called startle reaction in small children is nearly always accompanied by brady-
cardia and short apnea. Before the evoked potential era I used repetitive auditive
stimulation (clicks of varying loudness) during polygraphic recording for screening in
210 LA.Sulg

suspected loss of hearing. None of 30 infants, tested awake as well as in sleep,

reacted with any severe arrhythmia or apnea. However, infants and children with
prolonged QT interval in their ECG may even react with fatal arrhythmia [88].

Vasovagal Syncope
Acute vasovagal (vasodepressor) syncope is the most common cause of fainting and
a well-known expression of the influence of the autonomic nervous system. Fear, dis-
gust, anxiety and pain precipitate discharges from the limbic system that activate
medullary vasodepressor centers and perhaps project directly onto the spinal cord,
resulting in reduced sympathetic tone [79]. In the vasodepressor phase, heart rate
and blood pressure fall, cardiac output declines, and the CBF drops below the levels
required to maintain consciousness. Such subjects appear pale, and accompanying
parasympathetic hyperactivity induces piloerection and sweating. EEG shows suc-
cessive slowing to delta frequencies. Awareness usually returns promptly parallel to
EEG normalization once the subject becomes supine. With any of the disorders
causing syncope, the degree of impaired consciousness depends upon the severity of
reduction of CBF. Rarely, with prolonged cerebral ischemia brief convulsive move-
ments may result. In vasovagal syncope both sympathetic and parasympathetic sys-
tems are involved, since atropine prevents the bradycardia but not always the depres-
sor reaction.

Long Q-T Syndrome

The ability of the CNS to trigger severe ventricular arrhythmias including ventricular
fibrillation has been especially well documented in patients with prolonged Q-T
interval. Syncope as well as sudden death can be triggered by physical or emotional
stress [63, 89, 90]. Hereditary prolongation of the Q-T interval was first described in
association with congenital neural deafness, syncopes and sudden death (Jervell and
Lange-Nielsen syndrome) [28, 47]. A similar syndrome without deafness has also
been described (Romano-Ward syndrome) [85, 113]. Favorable responses to stell-
ectomy and pharmacologic interventions in these patients have been documented
(see Chapter 7). Fraser et al. [28] have given a detailed clinical description of the at-
tacks in this syndrome. If in exercise, they stop moving, lie or sit down, grasp their
breast or stomach, moan or cry. After some minutes the attack is over. The attacks
have the character of Adams-Stokes syncopes and are caused by ventricular tachy-
cardia of the torsade de pointes type, which may degrade into ventricular fibrillation.

Carotid Sinus Syndrome [42, 58, 68, 107]

The carotid sinus reflex (CSR) arc comprises an afferent limb from the carotid sinus
baroreceptors via the glossopharyngeal and vagus nerves to the midbrain. The pre-
cise central connections of the arc are not known in man, but it is generally consid-
ered that the vagus nucleus and the vasomotor center in the midbrain form a central
processing unit for the cardioinhibitory and vasodepressor CSR respectively. The
efferent limb of the cardioinhibitory reflex is via the vagus nerve and the parasym-
pathetic ganglia to the cholinergic receptors of the SA and A V nodes. The efferent
Differential Diagnosis in Syncope 211

limb of the vasodepressor reflex is via the sympathetic outflow to the heart, capaci-
tance and resistance vessels, affecting contractility, preload and afterload respec-
tively [68]. There has been much controversy concerning the significance of CSR
hypersensitivity. Evidence from a prospective study of carotid sinus massage in
elderly people indicated that CSR hypersensitivity was present only in patients with
overt cardiac disease or previously unexplained dizziness or syncope, but could not
be demonstrated in "normal" individuals [36, 68]. Carotid sinus and sick sinus syn-
dromes may coexist in patients with symptomatic bradyarrhythmias but the majority
of patients with carotid sinus syndrome (CSS) show no evidence of sinoatrial disease
[68].
In CSS, as well as in other syncopes triggered from reflexogenic structures, an
appropriate provocation test is essential for a proper diagnosis [70]. ECG, blood
pressure, and EEG have to be recorded continuously during the procedure. This en-
ables one to differentiate between vagovasal, vagocardial, and central reaction types
on carotid sinus stimulation [68].

Nonepileptic Syncopes, Sobbing Spasms and Breath-holding Spells

In small children manual pressure on the eye bulbs may provoke significant changes
in heart activity, respiration, and EEG. This has been classified by Gastaut [32] as
either "syncopes blanches" due to cardiac inhibition and vasodepression, or "syn-
copes blues" due to respiratory difficulties. The latter includes the so-called sobbing
syncopes and breath-holding syncopes. In case of prolonged cardiac arrest there was
loss of consciousness, fall, and jerks followed by a tonic spasm in opistotonus. EEG
showed typical hypoxic slowing down to transitory suppression, but never epilepti-
form events. Gastaut interpreted these findings as a reflex reaction in an exaggerated
vagotonia, common to all children who have syncopes. The "syncopal reflex" weakens
with age and syncopes disappear usually at about the age of 3-5 years [32).

Glossopharyngeal Syncope
More than 30 cases of glossopharyngeal syncope (GPS) with cardiac arrest have been
reported [53, 84, 86]. Most of these patients have primarily been misdiagnosed as
epileptics. Polygraphic monitoring may demonstrate the true nature of the mecha-
nism leading to syncope. One of my cases also had epilepsy, and EEG showed inter-
ictal discharges without any significant effect on heart rate. An attack of pain, how-
ever, was accompanied by profound bradycardia and EEG slowing, but there were
no epileptogenic discharges [103].
Syncope by swallowing (deglutition syncope) probably has the same neurogenic
background as syncope in glossopharyngeal neuralgia. The syncope is apparently the
result of a reflex in which both the afferent limb from the esophagus and the efferent
limb to the heart pass through the vagal nerve [43, 50, 60]. Most of the previously re-
ported cases had structural abnormalities of esophagus and heart, and some had a
combination of glossopharyngeal and vagal pathology as a result of infiltration by in-
vasive tumor. Usually cold drinks are the triggering factors [53]. Some patients have
syncope with swallowing as well as with vomiting. Most of these patients have also
increased sensitivity of the carotid sinus [43].
212 I. A.Sulg

Ischemic.Hypoxic Syncopes Secondary to Insufficiency

in Neural Control of Cardiovascular Functions

Postural Hypotension

In this category of syncopes, insufficient regulation of blood pressure and cardiac

output to postural demand is the primary cause. The insufficiency of neuroregulation
is on the sympathetic side of the autonomic nervous system. By postural venous
pooling there is an insufficiency or inability to compensate for the peripheral vascu-
lar resistance and for the decreased cardiac output [3, 37]. Any tendency to impaired
sympathetic reflex control is accentuated by reduced blood volume such as in chronic
use of diuretics.
Postural hypotension may arise with acute onset of sympathetic dysfunction in
severe infection, e.g., in Guillain-Barre syndrome [19, 39]. Chronic sympathetic dys-
function with postural hypotension can also result from drugs. A widespread degen-
eration and atrophy of mUltiple neuron systems (mainly sympathetic pre- and post-
ganglionic structures) may lead to the syndrome of multiple system atrophy (Shy-
Drager Syndrome) [3, 62, 9S]. Similar syndromes may develop in syringobulbia,
tabes dorsalis, familial dysautonomia (Riley-Day Syndrome), polyneuropathy (espe-
cially in diabetics), and other conditions gathered under the diagnosis of chronic
idiopathic postural hypotension. Intrinsic and extrinsic lesions in the brain stem may
also interrupt the control of blood pressure and cause postural hypotension [SO].
In patients with postural hypotension associated with multiple system atrophy, par-
kinsonian features are common. Occasionally parkinsonism may precede the dys-
autonomia and lead to an erroneous diagnosis of Parkinson's disease [SO].
About S%-lS% of geriatric patients have significant postural falls in blood pres-
sure that may lead to dizziness and syncope [50]. Syncope is a prevalent and poten-
tially dangerous event in elderly and may remain unexplained in up to SO% of cases
[96].

Postprandial hypotension may be an additional factor [91]. In 40 elderly with

histories of syncope the blood pressure and heart rates were monitored in sitting
position before and after a standardized breakfast [61]. In half of these individuals
the sitting systolic blood pressure declined by an average of 24 mm Hg within 3S min
after breakfast. Compensatory cardioacceleration was minimal, suggesting impair-
ment in baroreflexes.

Micturition syncope, usually seen in men when arising from bed at night to uri-
nate, is probably a special type of postural syncope. It has been suggested that vaso-
motor reflexes from the bladder playa contributory part inducing vagally mediated
bradycardia [17]. A full bladder causes vasoconstriction and rise of blood pressure by
a reflex through the spinal cord. When it is rapidly emptied this reflex ceases ab-
ruptly and blood pressure may fall so low that syncope results [SO]. One patient has
been reported to repeatedly faint about 30 s after voiding at night. The first signs
were 2: 1 A V block with ventricular rate of 28/min and drop in blood pressure, fol-
lowed by profound slowing in EEG [17].
Differential Diagnosis in Syncope 213

Drug-Induced Transitory Loss of Consciousness

All drugs affecting the autonomic nervous system and cardiac function (e.g., beta-
blockers, quinidine, and digitalis) may, if overdosed or in interference with other
drugs, cause near-syncopal or syncopal symptoms. Other drugs that interfere with
cardiovascular reflexes include chlorpromazine, meprobamate, barbiturates, and
benzodiazepines. Large doses of these drugs or moderate doses in subjects with an
impaired tolerance may result in severe hypotension.
The tricyclic antidepressants are known to prolong the QRS complex and the
Q-T interval. Usually this increase is small and harmless. If overdosed or in patients
with long Q-T interval, however, administration of these drugs may cause ventricular
arrhythmias of the torsade des pointes type [34, 41, 44, 69].
A few cases of carbamazepine-induced cardiac conduction disturbances have
been reported [56]. Misinterpretation of cardiac syncopes as being caused by epileptic
fits may lead to increased dosage with life-threatening consequences. Therefore it is
recommended that ECG is always recorded of patients before carbamazepine therapy
is initiated [56].
The most commonly used drug in cardiology, digitalis, also has some important
neurotropic effects. Of those, three have been well documented in animal studies:
vasomimetic actions, sensitization of baroreceptors (e.g., in sinus caroticus syn-
drome) and, in large doses, sympathetic stimulation. Only the first of these has been
widely appreciated in the clinical literature [33]. The third and most controversial
neural effect is the sympathomimetic action. While low doses inhibit sympathetic
outflow because of the baroreceptor sensitizing effect, large digitalis doses have been
found to excite the central nervous system, resulting in enhanced sympathetic out-
flow and cardiac arrhyhmias. The CNS excitatory effects are not confined only to
cardiac sympathetic nerves, but can also be demonstrated in phrenic and peripheral
sympathetic nerves, suggesting generalized excitation of brain stem nuclei [33, 54].
Based on this fact, one approach that has been effective experimentally is the use of
CNS depressants to counteract digitalis-induced tachyarrhythmias [33].

Syncopes of MisceUaneous Etiology

Tussive syncope, occurring with paroxysms of severe coughing, probably has a com-
plex etiologic mechanism. The main causes are primarily increased intrathoracic and
intra-abdominal pressures, which are then transmitted via the aorta partly to the
baroreceptors and partly to the cerebrospinal fluid compartment, causing a corre-
sponding increase in intracranial pressure. Both these effects reduce blood flow to
the brain, resulting in transient cerebral ischemia and syncope [29, 50, 65]. EEG re-
corded during cough syncope does not show any epileptiform activity [13, 29, 50, 65].
Severe aortic stenosis is associated with a high incidence of syncopal or near-
syncopal episodes, commonly during exercise. The cardiac output is reduced and
cannot increase during strain; dilated muscular beds create a low peripheral resis-
tance and "steal" blood from the cerebral circulation [79]. Continuous recording of
systemic and pulmonary arterial pressure and heart rate in patients with aortic
stenosis showed a subnormal systemic arterial pressure response to graded submaxi-
214 I.A.Sulg

mal exercise. Near-syncopal episodes occurring during abrupt symptom-limited

exercise were associated with acute falls in systemic and pulmonary arterial pressure
in the absence of cardiac arrhythmia [83]. Acute reflexogenic peripheral arterial and
venous vasodilatation, mediated by the effects of sudden changes in left ventricular
pressure on ventricular baroreceptor activity, plays a central role [12, 83].
In subclavian steal syndrome there is an obliteration of the proximal subclavian
artery, causing reversal of blood flow in the ipsilateral vertebral artery, which
siphons blood from the contralateral vertebral artery. This results in "steal" of blood
from the basilar artery. If ipsilateral arm muscles are exercised the stealing can be so
pronounced that vertebrobasilar TIA symptoms (drop attacks, more seldom pre-
syncopal or syncopal symptoms) may appear [8, 21, 82]. Bornstein and Norris [6],
however, are of the opinion that subclavian steal is a harmless hemodynamic phe-
nomenon in the majority of cases.
Intracerebral steal of blood from structures controlling wakefulness and con-
sciousness (i.e., the reticular formation) to other activated parts of the brain may
also result in near-syncopal symptoms or in syncope. At this point I would like to
propose a new hypothesis concerning epilepsy: the loss of consciousness in epileptic
fit may be an intracerebral steal phenomenon. A pronounced increase (more than
500%) of the regional CBF within discharging neural regions may steal blood from
the reticular formation.
Drop attacks due to subclavian steal or arteriosclerotic stenosis of vertebral/basilar
arteries produce brief insufficiency in corticospinal pathways in the pons. The subject
has a sudden atonia in the lower extremities, resulting in fall. There is usually no
cerebral hypoxia and EEG is not affected [79].
Arteriospastic basilar artery syncope has been described in association with mi-
graine. The syncopes, most often of the postural and vasodepressor types, may ap-
pear before or during migraine attacks [64].
Dissecting aneurysma of the ascendent aorta may have relapsing syncope as one
of the presenting symptoms [12].
High altitude syncope is a hypoxic syncope, which can also happen in an aircraft
when the cabin pressure drops. A rapid decompression (1-2s) from 8000 feet to
40000 feet leads to an impairment of performance in 12-15 s and to unconsciousness
in 20s [92]. Diving or immersion syncope is, unfortunately, not known by all people
exercising short distance diving. They usually hyperventilate before entering the
water; already hypocapnic and hypoxic at immersion, they exhaust their oxygen re-
serves by physical strain before producing a sufficient amount of CO 2 to result in sub-
jective dyspnea. The ensuing cerebral hypoxia can then induce loss of consciousness,
sometimes ending fatally [67].
Hyperventilation syndrome is a result of psychogenically triggered hyperventila-
tion leading progressively to hypocapnic vasospasms, preferably in the brain, result-
ing in paresthesias, giddiness, fainting, and clouding of consciousness. A real syn-
cope rarely occurs without some additional precipitating maneuver.

Pseudoseizures and Pseudosyncopes

The coexistence of epileptic and pseudoseizures ("hysterical seizures") in the same

patient is not rare and creates problems in diagnosis and management. The diagnosis
Differential Diagnosis in Syncope 215

of pseudoseizures and their differentiation from epileptic seizures can be made on

clinical grounds alone in most cases [14, 52, 59]. Pseudoseizure can be sudden or
gradual in onset and is usually more prolonged than an epileptic seizure. It occurs in
the presence of an audience and is usually in response to environmental stress. The
attack is bizarre and uncoordinated, unlike the epileptic seizure. Eye deviation,
autonomic overactivity, Babinski sign, tongue biting, incontincence, and postictal
confusion are absent. Physical injury is rare. Suggestion can often dramatically pre-
cipitate or terminate an attack [9, 79]. Often, however, the diagnosis is difficult even
though the seizure may have been witnessed by the physician. The patient is usually
unaware of the psychogenic nature of the symptom and is oblivious to the connection
between the symptom and his or her life situation. Many times the physician does not
get an opportunity to observe the actual episode and has to depend on the descrip-
tion of the seizures given by a witness. This can be very misleading because the wit-
ness in most cases is not a trained observer. It is difficult to obtain an EEG during an
actual episode and the physician is therefore forced to depend on interictal EEG,
with its serious limitations. A possible source of additional information is ECG dur-
ing attacks. As referred to earlier, the majority of complex partial seizures are ac-
companied by abrupt changes in heart rate and rhythm. The most striking difference
between these two groups was the time to reach maximum heart rate from onset of
ictal acceleration: mean time 205 s in pseudoseizures and 23 s in epileptic seizures
[40].
A further contribution to differential diagnosis can be blood analysis of prolactin.
It has been shown that transient hyperprolactinemia follows epileptic seizures, but
not pseudoseizures [1,10,11,57,108]. However, prolactin rise follows seizures orig-
inating only from the temporal lobe until generalization occurs. There is no signifi-
cant difference between grand mal seizures of various origins. This suggests that
once the entire brain is involved, there is interference with the regulation of prolactin.
In patients with myoclonic or akinetic seizures, no postictal increase of prolactin has
been found. In "hysterical seizures" an increase in prolactin is uncommon [57, 108].
In pseudosyncope (hysterical fainting) EEG shows a normal awake record, and
there is slight tachycardia. Signs of vegetative overactivity are lacking. Breathing is
eupneic or hyperpneic, especially in hyperventilation syndrome. The pseudofaint
subject is also incapable of slow closure of passively opened eyelid as in true fainting,
and more or less actively resists passive eye opening. The pupils are briskly respon-
sive.

Tests for Functions of the Autonomic Nervous System

Symptoms of autonomic dysfunction are common among patients with neurological

and metabolic disorders. Objective assessment of this dysfunction requires appropri-
ate tests. If EEG examination is performed as multivariable recording, this proce-
dure can offer valuable diagnostic information about the autonomic reactivity on the
cerebral as well as on the cardiovascular level [50].
Hyperventilation (HV) and photic stimulation are routine procedures in EEG ex-
aminations. A 3-5 min HV may provoke latent EEG abnormalities, especially epi-
216 I. A.Sulg

leptogenic ones through hypocapnic cerebral hypoxia. The HV has also some effects
on the heart. Arrhythmias, caused by coronary insufficiency, may be exaggerated or
provoked. Functional or neurogenic disturbances in heart rhythm may be reduced
[102]. Respiratory sinus arrhythmia, a physiological phenomenon, is always accen-
tuated by HV, especially during the first minute of forced hyperventilation [102].
This offers the opportunity to calculate some indices of heart rate variability as an
index of the function of the autonomic nervous system. During forced breathing the
vagus index can be calculated: the 3 longest and 3 shortest R-R intervals are mea-
sured during the first HV minute. The ratio between mean length of long intervals
and mean of short intervals is the vagus index, which is a rough measure of vagal in-
tactness. This index decreases with age [105]. A more pronounced decay results from
inflammatory and degenerative diseases in the autonomic nervous system [23, 75,
76]. In Shy-Drager and in acute Guillain-Barre syndromes, for example, there is
often hardly any respiratory arrhythmia. The variability in R-R intervals can also be
expressed in percentages [75]. In severe head trauma and in electrocerebral silence of
miscellaneous genesis there is usually an absolute abolition of respiratory arrhythmia
[102].
Valsalva maneuver can be performed in the same examination procedure. In
a standardized Valsalva test the subject blows into a mouthpiece connected to a
manometer, maintaining the pressure at 40mmHg for lOs. Here the vagus index is
the ratio between the longest interval after Valsalva and the shortest one during Val-
salva, normally ~ 1.21. There is usually also a 10-30 mm Hg rise in mean blood pres-
sure [2, 18, 93].
In stimulation of carotid sinus by massage continuous multivariable recording
(EEG, ECG, blood pressure) secures documentation of possible reactions: vagovasal,
vagocardial, and central responses [50,70,71].
The ocular pressure test can also be used for testing of vagal reactivity [30, 31].
Pressure on the eyeballs has been used to terminate attacks of paroxysmal atrial
tachycardia. Occasionally, however, it may cause profound bradycardia with syncope
[50].
Carotid and ocular compression are not without danger for the patient [72]. Usual-
ly, however, severe bradycardia can be easily controlled by drugs or pacing.
Orthostatic test is a mandatory diagnostic maneuver in evaluation of suspected
postural hypotension. This test too has to be monitored by multivariable recording.
Blood pressure and heart rate are here the most important variables. Most appropri-
ate would be an automatic noninvasive device regularly performing timed repeated
measurements. Such devices usually also calculate heart rate, which reflects the abil-
ity of the autonomic nervous system to compensate in postural hypotension [73]. An
abnormal response consists of a rapid and progressive decline in mean blood pres-
sure of more than 15 mmHg. Autonomic insufficiency can be inferred by observing
an unchanged pulse rate despite hypotension. A 30/15 ratio has been advocated by
Appenzeller [2]: the ratio between R-R intervals at beats 15 and 30; normal value
> 1.03. Simultaneous recording of EEG offers an opportunity to document the cere-
bral tolerance limit by eventual postural hypotension.
Applying these tests, one has to follow the principle that the test conditions have
to mimic the conditions of syncope as closely as possible. Therefore, the standing
position is the optimal one and will most often give considerably more positive re-
Differential Diagnosis in Syncope 217

suIts than with the subject lying or sitting. One has to take the necessary precautions,
of course, to avoid accidental injuries. A tilt table might be useful.

Conclusion

Diagnostic evaluation should be as effective as possible by means of rational and

reliable examination techniques. The time used for traditional work-up by a cardiol-
ogist and then by a neurologist, separately, could be reduced considerably and the
diagnostic precision increased, if an interdisciplinary team ("syncope team") had the
responsibility for an optimally coordinated diagnostic evaluation. Once epilepsy or
nonepileptic fit has been diagnosed as the cause of symptoms, clinical evaluation has
to be completed along these two alternative lines in order to find the primary etiology.
With this kind of evaluation time will be saved and appropriate treatment guaranteed.

Acknowledgement. Sincere thanks to Margaret Srether for typing the manuscript.

References

1. Abbot R, Browning M, Davidson D (1980) Serum prolactin and cortisol concentrations after
grand mal seizures. J Neurol Neurosurg Psychiatry 43: 173-177
2. Appenzeller 0 (1982) The autonomic nervous system. An introduction to basic and clinical
concepts, 3rd edn. North-Holland, Amsterdam
3. Bannister R (1979) Chronic autonomic failure with postural hypotension. Lancet II: 404-406
4. Blumhardt LD, Smith PEM, Owen L (1986) Electrocardiographic accompaniments of tem-
porallobe epileptic seizures. Lancet II: 1051-1057
5. Boudoulas H (1979) Superiority of 24 hour outpatient monitoring over multistage exercise test-
ing for the evaluation of syncope. J Electrocardiol12: 103-108
6. Bornstein NM, Norris JW (1986) Subclavian steal: a harmless haemodynamic phenomenon?
Lancet II: 303-305
7. Buren 1M, Ajmone-Marsan C (1960) A correlation of autonomic and EEG components in tem-
porallobe epilepsy. Arch Neurol3: 683-703
8. Cattaino G, Querin F, Pomes A, Piazza P (1984) Transient global amnesia. Acta Neurol Scand
70:385-390
9. Cohen R, SUter C (1982) Hysterical seizures: suggestion as a provocative EEG test. Ann
Neurol11:391-395
10. Collins WCJ, Lanigan 0, Callaghan N (1983) Plasma prolactin concentrations following epi-
leptic and pseudo seizures. J Neurol Neurosurg Psychiatry 46: 505-508
11. Dana-Haeri J, Trimble M, Oxby J (1983) Prolactin and gonadotrophin changes following
generalized and partial seizures. J Neurol Neurosurg Psychiatry 46: 331-335
12. DeBarey ME, McCollum CM, Crawfort ES, Morris GC, Howell J, Noon GD, Lawrie G (1982)
Dissection and dissecting aneurysms of the aorta: twenty-year follow-up of 527 patients treated
surgically. Surgery 92: 118-134
13. DeMaria AA, Westmoreland BF, Sharbrough FW (1984) EEG in cough syncope. Neurology
34:371-374
14. Desai BT, Porter R, Penry JK (1979) The psychogenic seizure by video tape analysis: a study
of 42 attacks in 6 patients. Neurology 29: 602
15. Devinsky 0, Price BH, Cohen SI (1986) Cardiac manifestations of complex partial seizures.
Am J Med 80: 195-202
218 I.A.Sulg

16. Dimant J, Grob D (1977) Electrocardiographic changes in myocardial damage in patients with
acute cerebrovascular accidents. Stroke 8: 448-455
17. Donker DNJ, Robles de Medina EO, Kieft J (1972) Micturition syncope. Electroencephalogr
Clin NeurophysioI33:328-331
18. Duvoisin RC (1962) Convulsive syncope induced by the Weber maneuver. Arch Neurol 7:
219-226
19. Edmonds ME, Sturrock RD (1979) Autonomic neuropathy in the Guillain-Barre syndrome. Br
Med J 2: 668-670
20. Eisenberg S, Madison L, Sensenbach W (1960) Cerebral hemodynamic and metabolic studies
in patients with congestive heart failure. Circulation 21 : 704-709
21. Eklof B, Schwartz SI (1970) Effects of subclavian steal and compromised cephalic blood flow
on cerebral circulation. Surgery 68:431-441
22. Erne P, Raine AEG, Burgisser E, Gradel E, Burkart F, Buhler FR (1987) Paradoxical inhibi-
tion of atrial natriuretic peptide release during pacing-induced hypotension. Clin Sci 73: 459-
462
23. Ewing DJ, Campbell IW, Clarke BF (1980) The natural history of diabetic autonomic neu-
ropathy. Q J Med 49: 95-108
24. Fairfax AI, Lambert CD, Leatham A (1976) Systemic embolism in chronic sinoatrial disorder.
N Engl J Med 295 : 190-192
25. Fischer T (1964) Extreme Sinusbradykardie als Folge eines lateralen oberen Oblongatasyn-
droms (Wallenberg-Foixsces Syndrom). Psychiatr Neuro1147: 397-401
26. Folkow B, Lisander B, Tuttle RS, Wang SG (1968) Changes in cardiac output upon stimulation
of the hypothalamic defence area and the medullary depressor area in the cat. Acta Physiol
Scand 72: 220-233
27. Forand 1M, Schweiss IF (1984) Pacemaker syndrome during anesthesia. Anesthesiology 60:
588-590
28. Fraser GR, Froggatt P, Murphy T (1964) Genetical aspects of the cardio-auditory syndrome of
Jervell and Lange-Nielsen. Ann Hum Genet 28: 133-141
29. Fuxe K, Agnati A, Hlirfstrand M, Kalia TH, Svensson TH, Neumeyer A, Zoli M, Lang R,
Ganten D, Terenius L, Goldstein M (1986) Morphological and biochemical studies on neuro-
peptide Y (NPY) and adrenergic mechanisms and their interactions in central cardiovascular
regulation. In: Stober T, Schimrigk K, Ganten D, Sherman DG (eds) Central nervous system
control of the heart. Proceedings of the 3rd International Brain Heart Conference. Martinus
Nijhoff, Boston, pp 38-82
30. Gastaut H, Fischer-Williams M (1957) Electroencephalographic study of syncope, its differenti-
ation from epilepsy. Lancet II: 1018-1025
31. Gastaut H, Gastaut Y (1958) La syncope vaso-vagale reflexe. Sandoz, Paris, p 44
32. Gastaut H (1964) A physiopathogenic study of reflex anoxic cerebral seizures in children (syn-
copes, sobbing spasms and breath-holding spells). In: Kellaway P, Petersen I (eds) Clinical
electroencephalography of children. Grune and Stratton, New York, pp 257-274
33. Gillis RA, Pearle DL, Levitt B (1975) Digitalis: a neuroexcitatory drug. Circulation 52: 739-
742
34. Glassmann AH, Bigger IT, Giardina EV, Kantor SJ, Perel JM, Davies M (1979) Clinical char-
acteristics of imipramine-induced orthostatic hypotension. Lancet I: 468-472
35. Goldberg AD, Raftery EB, Cashman PMM (1975) Ambulatory electrocardiographic recording
in patients with transient cerebral attacks or palpitation. Br Med J 4: 569-571
36. Gould L, Reddy CV, Becker WH, Keunchan 0, Soo Guym K (1978) Usefulness of carotid
sinus pressure in detecting the sick sinus syndrome. J Electrocardiolll: 261-268
37. Hilton SM, Spyer KM (1980) Central nervous regulation of vascular resistance. Annu Rev
Physiol 42: 399-411
38. Hirsh CS, Martin DL (1971) Unexpected death in young epileptics. Neurology 21: 682-690
39. Hokkins A, Neville B, Bannister R (1974) Autonomic neuropathy of acute onset. Lancet II:
769-771
40. Howell SJL, Blumhardt LD, Smith PEM (1988) Changes in heart rate associated with pseudo-
seizures. International Congress on Long Term EEG Monitoring, Bad Hofgastein, Austria
41. Hulten B-A, Heath A (1983) Clinical aspects of tricyclic antidepressant poisoning. Acta Med
Scand 213: 275-278
Differential Diagnosis in Syncope 219

42. Hutchinson EC, Stock JPP (1960) The carotid sinus syndrome. Lancet II: 445-446
43. Inglauer S, Schwartz BA (1936) Heart-block periodically induced by the swallowing of food in
a patient with cardiospasm (vasovagal syncope). Ann Otol Rhinol Laryngol 45: 875-880
44. Jackson JH (1931) Clinical and physiological researches on the nervous system. In: Selected
writings of JH Jackson. Hodder and Stoughton, London
45. Jay FW, Leestma JE (1981) Sudden death in epilepsy. Acta Neurol Scand [SuppI82] 63: 1-66
46. Jeavons PM (1975) The practical management of epilepsy. In: Brett EM (ed) Paediatric neurol-
ogy. Churchill Livingstone, Edinburgh, pp 304-305
47. Jervell A, Lange-Nielsen F (1957) Congenital deaf-mutism, functional heart disease with pro-
longation of the Q-T interval and sudden death. Am Heart J 54: 59-69
48. Johansson BW (1966) Complete heart block (thesis). Acta Med Scand [Suppl]: 180
49. Johansson BW (1977) Long-term ECG in ambulatory clinical practice. Eur J Cardiol 5[1]:
39-48
50. Johnson RH, Lambie DG, Spalding JKM (1986) The autonomic nervous system. In: Baker
AB, Joynt RJ (eds) Clinical neurology, vol 4. Harper and Row, Cambridge, pp 14-48
51. Kapnoor WN, Karpf M, Maher Y, Miller RA, Levy GS (1982) Syncope of unknown origin: the
need for a more cost-effective approach to its diagnostic evaluation. JAMA 247: 2687-2691
52. King DW, Gallagher BB, Murvin AJ, Smith DB, Markus DJ, Hartlage LC, Ward LC (1982)
Pseudoseizures: diagnostic evaluation. Neurology 32: 18-23
53. Kjellin K, Muller R, Widen L (1959) Glossopharyngeal neuralgia associated with cardiac arrest
and hypersecretion from the ipsilateral parotid gland. Neurology 9: 527-532
54. Korth C (1937) Uber die Wirkung des Strophantins auf das Zentralnervensystem. Arch Exp
PathoIPharmakoI185:42-46
55. Krohn W (1963) Causes of death among epileptics. Epilepsia 4: 315-321
56. Ladefoged SD, M(lJgelvang JC (1982) Total atrioventricular block with syncopes complicating
carbamazepine therapy. Acta Med Scand 212: 185-186
57. Laxer KD, Mullooly JP, Howell B (1985) Prolactin changes after seizures classified by EEG
monitoring. Neurology 35: 31-35
58. Leatham A (1982) Carotid sinus syncope. Br Heart J 47: 409-410
59. Lesser RP (1985) Psychogenic seizures. In: Pedley TA, Meldrum BS (eds) Recent advances in
epilepsy, no. 2. Churchill Livingstone, Edinburgh, pp 273-288
60. Levin B, Posner JB (1972) Swallow syncope. Neurology 22: 1086-1093
61. Lipsitz LA, Nyquist RP, Wei JY, Rowe JW (1983) Postprandial reduction in blood pressure in
elderly. N Eng! J Med 309: 81-83
62. Loewy AD (1986) Anatomic aspects of central nervous cardiovascular regulation. In: Stober T,
Schimrigk K, Ganten D, Sherman DG (eds) Central nervous system control of the heart.
Proceedings of the 3rd International Brain Heart Conference. Nijhoff, Boston, pp 3-18
63. Lown B (1979) Sudden cardiac death: the major challenge confronting contemporary cardiol-
ogy. Am J Cardiol43: 313-328
64. Lund M (1962) Arteriospastic syncope in association with migraine (in Danish). Nord Med 70:2-7
65. McIntosh HD, Estes EH, Warren JV (1956) The mechanism of cough syncope. Am Heart J 52:
70-82
66. Miler M, Fox S, Jenkins R, Schwartz J, Toonder FG (1981) Pacemaker syndrome: a non-
invasive means to its diagnosis and treatment. PACE 4:503-506
67. Miles S (1972) Underwater medicine. Staples, London, p 363
68. Morley CA, Hudson WM, Kwok HT, Sutton R (1983) Is there a difference between carotid
sinus and sick sinus syndromes? Br Heart J 49 : 620-621
69. Motte G, Coumel PH, Abitbal G, Dessertenne F, Salma R (1970) Le Syndrome Q-T long et
syncopes par "torsade de pointes". Arch Mal Coeur 63: 831-853
70. Nager F (1961) Die Kardioinhibition und Carotissinusdruck Effecte im ECG bei Herz- und
Kreislaufgesunden? Helv Med Acta 28 : 42-62
71. Nathanson MH (1946) Hyperactive cardioinhibitory carotid sinus reflex. Arch Intern Med 54:
111-130
72. Nelson DA, Mahrn MM (1963) Death following digital carotid artery occlusion. Arch Neurol
8:640-646
73. Norris A, Nathan BA, Schock W, Yiengst MJ (1955) Are changes in heart rate and blood pres-
sure response to tilting and standardized exercise? Circulation 8 : 521-526
220 I. A.Sulg

74. Parizel G (1973) Life-threatening arrhythmias in subarachnoid hemorrhage. Angiology 24:

17-21
75. Persson A, Solders G (1983) R-R variations, a test of autonomic dysfunction. Acta Neurol
Scand 67:285-293
76. Persson A, Solders G (1983) R-R variations in Guillain-Barre syndrome: a test of autonomic
dysfunction. Acta Neurol Scand 67: 294-300
77. Phizackerley PJR, Poole EW, Whitly CVM (1954) Sinoauricular heart block as an epileptic
manifestation. Epilepsia 3: 89-91
78. Plum F (1979) Cardiac arrhythmias and neurological dysfunction. In: Busse E (ed) Cerebral
manifestations of episodic cardiac dysrhythmias. Excerpta Medica, Amsterdam, pp 11-17
79. Plum F (1982) Brief or episodic loss of consciousness. In: Wyngaarden J, Smith LH (eds)
Textbook of medicine. Saunders, Philadelphia, pp 1926-1930
SO. Rasmussen K (1971) Chronic sinoatrial heart block. Am Heart J 81: 38-47
81. Rasmussen K (1981) Chronic sinus node disease: natural course and indications for pacing. Eur
Heart J 2: 455-459
82. Reivich M, Holling HE, Took JF (1961) Reversal of blood flow through the vertebral artery
and its effect on cerebral circulation. N Engl J Med 265 : 878-885
83. Richards AM, Nicholls MG, Ikram H, Hamilton EJ, Richards RD (1984) Syncope in aortic
valvular stenosis. Lancet II: 1113-1114
84. Riley HA, German WJ, Wortis H, Herbert C, Zern D, Eichna L (1942) Glossopharyngeal
neuralgia initiating or associated with cardiac arrest. Trans Am Neurol Assoc 68: 28-30
85. Romano C, Gemme G, Pongiglione R (1963) Aritme cardiache rare dell'eta pediatrica. Clin
Pediatr 45 : 656-683
86. Rushton I, Stevens JC, Miller RH (1981) Glossopharyngeal (vagoglossopharyngeal) neuralgia.
A study of 217 cases. Arch NeuroI38:201-205
87. Samson WK (1987) Atrial natriuretic factor and the central nervous system. In: Rosenblatt M,
Jacobs JW (eds) Atrial natriuretic factor. Endocrinology and metabolism. Clinics of North
America, Saunders, Philadelphia, pp 145-161
88. Scholt GD, McLeod AA, Jewitt DE (1977) Cardiac arrhythmias that masquerade epilepsy. Br
Med J 2: 1454-1457
89. Schwartz PJ (1976) Cardiac sympathetic innervation and the sudden infant death syndrome.
Am J Med 60:161-172
90. Schwartz PJ, Stone HL (1978) Unilateral stellectomy and sudden death. In: Schwartz PJ,
Brown AM, Melliani A, Zanchetti A (eds) Neural mechanisms in cardiac arrhythmias. Raven
Press, New York, pp 107-122
91. Seyer-Hansen K (1977) Postprandial hypotension. Br Med J 2: 1262-1263
92. Sharp GR, Dhenin G, Ernsting J (19SO) Aviation medicine, physiology and human factors.
Tri-Med, London, p 172
93. Sharpey-Schafer EP (1955) Effects of Valsalva maneuver on the normal and failing circulation.
Br J Med 1: 693-695
94. Shaw DB, Holman RR, Gowers JI (1980) Sinoatrial disorder (sick sinus syndrome). Br Med J
1: 139-141
95. Shy G, Drager GA (1960) A neurological syndrome associated with orthostatic hypotension.
Arch NeuroI2:511-527
96. Silverstein MD, Singer DE, Mulley AG, Thibaut GE, Barnett GO (1982) Patients with syn-
cope admitted to medical intensive care units. JAMA 248: 1185-1189
97. Stem S, Tzivoni D (1976) Atypical causes of Adams-Stokes attacks. Chest 69:249-255
98. Stem S, Lavy S (1979) Frequency and type of cerebral ischemic attacks in patients with tachy-
arrhythmias and bradyarrhythmias. In: Busse E (ed) Cerebral manifestations of episodic car-
diac dysrhythmia. Excerpta Medica, Amsterdam, pp 30-47
99. Sulg IA (1965) Relationships between EEG and the cardiac rhythm (in Swedish). Nord Med
74:707
100. Sulg IA (1967) Polygraphy in differential diagnosis of paroxysmal loss of consciousness. Elec-
troencephalogr Clin Neurophysiol 23: 370
101. Sulg lA, Chronquist S, Schilller H, Ingvar DH (1969) The effect of intracardial pacemaker
therapy in cerebral blood flow and electroencephalogram in patients with complete atrio-
ventricular block. Circulation 39: 487-494
Differential Diagnosis in Syncope 221

102. Sulg IA (1969) Interval analysis of heart rhythm. Ann Clin Research 1: 150-155
103. Sulg IA (1979) Dependency between cerebral and cardiac activity in physiological as well as in
abnormal conditions. In: Busse E (ed) Cerebral manifestations of episodic cardiac dysrhyth-
mias. Excerpta Medica, Amsterdam, pp 30-47
104. Sulg IA (1984) Neurogen Kardiopati (in Swedish). Liikartidningen 81: 1060-1061
105. Sulg lA, Saunte C, Sjaastad 0, Tyssvang T (1986) Heart rate and the index of respiratory ar-
rhythmia as a measure of cardiac vagus-sympathicus equilibrium. Ups J Med Sci [Suppl] 43: 75
106. Terrence CF, Wisotzkey HM, Perper TA (1975) Unexpected, unexplained death in epileptic
patients. Neurology 25: 594-598
107. Thomas JE (1972) Disease ofthe carotid sinus syncope. In: Vinken PJ, Bruyn GW (eds) Hand-
book of clinical neurology, vol1l. North-Holland, Amsterdam, pp 523-551
108. Trimble M (1978) Serum prolactin in epilepsy and hysteria. Br Med J 2: 1682-1685
109. Ueda H (1962) Arrhythmias produced by cerebral stimulation. Jpn Circ J 26: 225-230
110. Van Burn 1M (1958) Some autonomic concomitants of ictal automatism. Brain 81 : 505-529
111. Van Durme JP (1975) Tachyarrhythmias and transient cerebral ischemic attacks. Am Heart J
89:538-540
112. Wada JA (1985) Differential diagnosis of epilepsy. Electroencephalogr Clin Neurophysiol
[Suppl] 37: 285-311
113. Ward OC (1964) A new familial cardiac syndrome in children. J Ir Med Assoc 54: 103-106
Part IV
Thromboembolism and Ischemia
Chapter 16
Thromboembolic Complications in Atrial Fibrillation
J. GOD1FREDSEN and P. PETERSEN

Introduction

Both stroke and atrial fibrillation (AF) are common and important clinical entities
with significant morbidity and mortality. They are also linked together through a
wide variety of possible pathogenetic mechanisms. The classic theory is that in AF
many, if not most, cases of ischemic cerebral infarcts are caused by embolism origi-
nating from thrombus formation in the noncontracting left atrium.
AF is present in a very heterogeneous group of cardiac patients regarding age,
degree of congestive heart failure (CHF), previous high blood pressure (HBP), and
other disorders. Since the same is the case for many types of stroke, clinically and
pathologically as well as in CT scan characteristics, it is not surprising that unanimity
regarding risk factors and pathogenesis of stroke in AF is not easily obtained.
This review deals with this complex issue, its pathological, clinical, and therapeutic
features, based on the renewed interest in the subject and recent important findings
in the cardioneurologicalliterature. Some important review articles on the subject
have recently been published [6, 9]. It seems reasonable to distinguish between the
occurrence of stroke in AF on the one side (the cardiologist's view) and the occur-
rence of AF in stroke on the other (the neurologist's view). The percentage of both
total numbers of stroke in AF and total numbers of patients with AF in stroke are
very similar, namely, about 20%-25% [4,11].

Pathology and Pathogenesis

The pathogenetic background for thrombus formation in the atria in AF includes en-
larged atria, injured mitral valve [as in rheumatic heart disease (RHD)], akinetic
atrial wall, injured atrial endothelium, and stagnant and sluggish blood flow in the
atria [5, 13, 19]. All these mechanisms need not be present simultaneously. Each of
them, and especially their interplay, is incompletely understood, but still regarded
necessary for embolism even though most studies fail to find more than about 50%
atrial thrombi as compared with the number of emboli [1, 2, 16]. In one patho-
anatomic study, however, this ratio was reversed, i.e., atrial thrombi outnumbered
emboli by 2: 1 [7].
Thus, even though it is established that atrial thrombi are more frequent in pa-
tients with AF, and especially in patients with RHD, than in patients with sinus
226 J. Godtfredsen and P. Petersen

rhythm (SR), this is no proof that stroke in AF is mainly of embolic origin. Although
some progress seems on its way, one difficulty is the lack of a long-term animal model
as well as studies with labeled clots in the atria [29].
In two recent studies with rather strict neurological criteria for embolic stroke,
the proportion of embolic origin vs localized atherothrombotic mechanisms varied
between 20% and 63% [4, 14]. Another important mechanism that may contribute
to stroke during AF is decreased cerebral blood flow due to low cardiac output and
perhaps also due to changes in coagulation factors leading to regional localized
thrombus formation [12, 18].
Left atrial size seems to increase gradually in the course of AF [24] and the occur-
rence of stroke is apparently an early event after the onset of AF (see below), speak-
ing against the atrial thrombus/embolus theory at least in nonvalvular AF. Injured
atrial endothelium in the presence of low cardiac output may of course play an im-
portant role in thrombus formation, but only very few data exist [9].
Finally, it could be that in many cases of stroke AF is the consequence and not
the indirect cause of the acute cerebral event. Clinical evidence of such a mechanism
is the very high degree of CHF in these elderly patients [4, 11], and usually larger
cerebral infarcts and high mortality [9]. From a cardiologist's viewpoint, it is no sur-
prise that a severe cerebral disaster, whatever the cause, could precipitate overt de-
compensated CHF and AF in elderly, frail patients with concomitant, unstable com-
pensated heart failure. Based on careful clinical analysis of the circumstances at the
onset of AF, it was shown that in 70 of 1212 patients (5.6%) AF was possibly precipi-
tated by stroke, which is about 40% of the entire number of strokes in this series
[11].
Pointing in the same direction are the results of a recent Danish clinical and
echocardiographic study of 70 patients with AF where the risk of emboli varied with
the underlying etiology: lone AF: primary myocardial disease: primary mitral valve
disease with a ratio of 1: 2 : 7 [8]. The present situation regarding emboli as the cause
of stroke in AF is thus very similar to the situation regarding the pathogenesis of AF
itself: most likely the cause is multifactorial with several closely interwoven mecha-
nisms, many of which are not yet clearly defined. Still, a common denominator is
thrombogenesis, making treatment with anticoagulants attractive.

Epidemiological and Clinical Factors

The two main sources of data on stroke in AF are the Framingham Study and the
British Whitehall and Regional Heart Study [10, 30]. In the Framingham Study the
risk of stroke was increased 17-fold in RHD and 5.6-fold in nonvalvular AF. The
British study, with its two parts, showed a relative risk of stroke in rheumatic and
nonrheumatic AF of 6.9% and 2.3%, respectively. The total stroke rate was lower,
however, namely, 1.5% and 0.4% per year vs 5% per year 12 months after the onset
of nonvalvular AF in the Framingham Study. Both these prospective studies dealt
with individuals in the younger age groups (30-69 years at entry) and can therefore
hardly be considered representative of the much more common situation in the clini-
cal setting, where the median age of the typical AF patient is 70 years [11].
Thromboembolic Complications in Atrial Fibrillation 227

Table 1. Incidence rates of thromboembolic complications in P AF and CAF

related to the duration of arrhythmia

< 1 month Up to 1 year 2-5 years

PAF (n=426) 6.8% 2% 2%

CAF (n=786) 8.2% 4% 4%

It had previously been demonstrated in the Danish study, and later confirmed in
the Framingham Study, that the thromboembolic complications cluster shortly after
the onset of AF or after the transition from paroxysmal AF (PAF) to chronic AF
(CAF) [11, 15, 21, 31]. Incidence rates ofthromboembolic complications in PAF and
CAF are shown in a selected clinical series in Table 1 [21]. In CAF it has been shown
by multifactorial Cox's analysis that the only significant risk factor for stroke in AF
is the underlying etiology (e.g., RHD) of AF, but not age, sex, or degree of CHF at
the onset of AF [22]. In the British prospective study of men with nonvalvular AF,
hypertension was also shown to be a risk factor for stroke [10]. In contrast, the
pathoanatomic study from Boston [16] showed only a minor difference in the rate of
systemic embolism between AF patients with RHD and those with ischemic heart
disease (IHD): 41 % vs 35%.
By means of multiple logistic regression analysis it was shown in the latest report
from the Framingham Study [31] that also age, systolic blood pressure, IHD, and
CHF were significant risk factors for stroke, though it is not entirely clear whether
the multifactorial analysis comprised all strokes or only those that occurred in the 59
patients with AF at entry.
Not only is there quite a high total rate of thromboembolic complications during
the entire course of AF, i.e., about 17% to 50% depending on the underlying etiol-
ogy, but also the recurrence rate and the mortality resulting from complications are
both about 25% [10,11,14,16,26,31].
Thus, the important features of thromboembolic complications in AF are: (a)
definitely increased risk of stroke which accounts for about 60%-80% of such com-
plications, (b) clustering of events shortly after onset of AF, (c) increased rate of
embolic complications in RHD, and (d) recurrence rate and mortality due to compli-
cations of about 25% .
Even when a stroke in AF is survived for more than 1 month, the mortality is still
increased [10, 22] and depends primarily upon age, degree of CHF, and HBP.
Besides the more crude clinical methods of diagnosing stroke in hospitalized pa-
tients in the series quoted above, a recent study in ambulatory, cerebrally asymp-
tomatic patients with CAF using CT scan showed that 48% of these patients had cor-
tical areas of low density compared with only 28% in patients with SR. This finding
may be an indication of silent cerebral embolism in AF patients [23].
Lone AF usually denotes PAF or CAF in patients without any overt underlying
heart disease. The Framingham Study of these patients [3] demonstrated a fourfold
increase in the stroke rate compared with controls, suggesting that lone AF is not
benign. However, a study from the Mayo Clinic [17] in patients with lone AF <60
years of age, 91% of whom had PAF, showed a very low incidence of thromboem-
bolism: only 6% at the lO-year follow-up.
228 J. Godtfredsen and P. Petersen

Prevention and Treatment of Thromboembolism

No randomized, clinical trial has been performed to solve the question of a possible
benefit from anticoagulation treatment in AF. Most authors agree that AF in RHD,
particularly mitral stenosis, is a clear-cut indication for life-long anticoagulation
treatment, as is an episode of sustained stroke in nonvalvular AF patients [20].
A recent retrospective study recommended anticoagulation treatment prophy-
lactically also in nonvalvular AF patients [25].
However, the use of anticoagulants in AF after a sustained stroke has been seri-
ously questioned on the basis of methodological problems and as a result of evalua-
tion of the survival rate and the recurrence of stroke in this setting as compared with
patients with SR [27].
Presently, several randomized studies of anticoagulation treatment in AF are
underway. Until the results from these are known, a prudent attitude seems to be an
individualized evaluation in each case, and probably not to anticoagulate for more
than 1-2 years, especially in older patients where this treatment carries a significant
risk of serious side effects.
The role of antiplatelet-aggregating agents in AF is also an entirely unsettled
matter, but they are used by some clinics on pragmatic grounds [28].

References

1. Aberg H (1969) Atrial fibrillation. I. A study of atrial thrombus and systemic embolism in a
necropsy material. Acta Med Scand 185: 373-379
2. Beer DT, Ghitman B (1961) Embolization from the atria in arteriosclerotic heart disease. JAMA
177:287-290
3. Brand FN, Abbott RD, Kannel WB, Wolf PA (1985) Characteristics and prognosis of lone atrial
fibrillation. JAMA 254:3449-3453
4. Britton M, Gustafsson C (1985) Non-rheumatic atrial fibrillation as a risk factor for stroke.
Stroke 16: 182-188
5. Caplan LR, D'Cruz I, Hier DB, Reddy H, Shah S (1986) Atrial size, atrial fibrillation, and
stroke. Ann Neurol19: 158-161
6. Cerebral embolism task force (1986) Cardiogenic brain embolism. Arch Neuro143 :71-84
7. Davies MJ, Pomerance A (1972) Pathology of atrial fibrillation in man. Br Heart J 34: 520-525
8. Egeblad H (1987) Intracardiac thrombus - systemic embolism. Contribution of echocardiogra-
phy. University of Copenhagen (in press)
9. Fisher CM (1982) Embolism in atrial fibrillation. In: Kulbertus HE, Olsson SB, Schlepper M
(eds) Atrial fibrillation. Hassle, Molndal, pp 192-207
10. Flegel KM, Shipley MJ, Rose G (1987) Risk of stroke in nonrheumatic atrial fibrillation. Lancet
1:526-529
11. Godtfredsen J (1975) Atrial fibrillation. Etiology, course and prognosis. A follow-up study of
1212 cases. Munksgaard, Copenhagen
12. Gustafsson C, Blomback M, Britton M, Hamsten A, Svensson J (1987) Haemostatic function
and the increased risk of stroke in nonvalvular atrial fibrillation. Stroke 18: 25
13. Hall CA, Clubb S, Cabin HS (1987) Atrial fibrillation and systemic embolization: left atrial size
strongly predicts risk of embolization in patients without mitral stenosis. J Am Coll Cardiol 9:
220A
14. Hart RG, Coull BM, Hart D (1983) Early recurrent embolism associated with nonvalvular atrial
fibrillation: a retrospective study. Stroke 14:688-693
Thromboembolic Complications in Atrial Fibrillation 229

15. Hart RG, Easton JD, Sherman DG (1987) Duration of nonvalvular atrial fibrillation and stroke.
Stroke 14:827
16. Hinton RC, Kistler JP, Fallon IT, Friedlich AL, Fisher CM (1977) Influence of etiology of atrial
fibrillation on incidence of systemic embolism. Am J Cardiol 40: 509-513
17. Kopecky SL, Gersh BJ, McGoon MD, Holmes DR, Whisnant JP, Frye RL (1985) The natural
history of idiopathic "lone" atrial fibrillation: a 3 decade population study. Circulation [SuppI3]
72:1
18. Lavy S, Stern S, Melamed E, Cooper G, Keren A, Levy P (1980) Effect of chronic atrial fibrilla-
tion on regional cerebral blood flow. Stroke 11 : 35-38
19. Matsuzaki M, Toma Y, Suetsugu M, Ono S, Okada K, Hiro J, Anno Y, Morimoto K, Kusu-
kawa R (1987) Analysis of blood flow velocity and thrombogenesis in left atrial appendage by
transesophageal two-dimensional echocardiography. J Am Coli Cardiol 9: 212A
20. Nordlander R (1982) Anticoagulation in atrial fibrillation. In: Kulbertus HE, Olsson SB,
Schlepper M (eds) Atrial fibrillation. Hlissle, Molndal, pp 251-259
21. Petersen P, Godtfredsen J (1986) Embolic complications in paroxysmal atrial fibrillation. Stroke
17:622-626
22. Petersen P, Godtfredsen J (1986) Risk factors for stroke in chronic atrial fibrillation. Circulation
[Suppl 2]74: 152
23. Petersen P, Madsen EB, Brun B, Pedersen F, Gyldensted C, Boysen G (1987) Silent cerebral
infarction in chronic atrial fibrillation. Stroke (in press)
24. Petersen P, Kastrup J, Brinch K, Godtfredsen J, Boysen G (1987) Relation between left atrial
dimension and duration of atrial fibrillation. Am J Cardiol (in press)
25. Roy D, Marchand E, Gagne P, Chabot M, Cartier P (1986) Usefulness of anticoagulant therapy
in the prevention of embolic complications of atrial fibrillation. Am Heart J 112: 1039-1043
26. Sage 11, VanUitert RI (1983) Risk of recurrent stroke in patients with atrial fibrillation and non-
valvular heart disease. Stroke 14:537-540
27. Sandercock P, Warlow C, Bamford J, Peto R, Starkey I (1986) Is a controlled trial oflong-term
oral anticoagulants in patients with stroke and non-rheumatic atrial fibrillation worthwhile?
Lancet 1: 788-792
28. Sherman DG, Hart RG, Easton JD (1986) The secondary prevention of stroke in patients with
atrial fibrillation. Arch N eurol 43 : 68-70
29. Vandenberg B, Seabold J, Kieso R, Hunt M, Comad G, Johnson J, Kerber RE (1986) Indium-
111 platelet scintigraphy and two-dimensional echocardiography for the detection of experimen-
tal left atrial appendage thrombi. Circulation [SuppI2] 74 : 513
30. Wolf PA, Dawber TR, Thomas HE, Kannel WB (1978) Epidemiologic assessment of chronic
atrial fibrillation and risk of stroke: the Framingham Study. Neurology 28:973-977
31. Wolf PA, Kannel WB, McGee DL, Meeks SL, Bharucha NE, McNamara PM (1983) Duration
of atrial fibrillation and imminence of stroke: the Framingham Study. Stroke 14: 664-667
Chapter 17
Echocardiography and Embolic Sources in the Heart
K.-A.JoHANNESSEN

The incidence of cerebral vascular accidents in patients with thrombo-embolic dis-

ease in the heart may be as high as 20% -30%. Myocardial infarction, atrial fibrilla-
tion, and mitral valve disease are the lesions most commonly associated with cerebral
embolism.
Two-dimensional echocardiography is a noninvasive technique which gives cross-
sectional visualization of both myocardium, heart valves, and intracavitary struc-
tures. It is of no discomfort to the patient and may be performed even in critically ill
patients.
As many as one out of six ischemic strokes may be caused by emboli from the
heart [6]. Patients with acute myocardial infarction have a 2%-5% incidence of
stroke, increasing to 9%-12% when considering only patients with anterior infarc-
tions. These emboli originate from left ventricular thrombi, which are found by
cross-sectional echocardiography in 30% -45% of patients with anterior wall infarc-
tion [2, 4, 11, 12, 20, 23, 38, 41]. These thrombi may cause embolization in 15%-
30% of the patients [16, 20, 39, 41].
Patients with atrial fibrillation have a 5-7 fold increased risk of stroke [8,18,43].
The annual incidence of cerebral vascular accidents in patients with nonvalvular
atrial fibrillation is 5%-6% [6,8], and the overall risk may be as high as 20%-25%
[13,33]. The etiology of these strokes is assumed to be left atrial thrombi. Also, sev-
eral abnormalities of the heart valves may be associated with peripheral emboliza-
tion. Thus, there are many important clinical situations in which cross-sectional
echocardiography may disclose the etiology of cerebral embolism.

Left Ventricular Thrombi

Left ventricular thrombi detected with cross-sectional echocardiography occur in

30%-45% of patients with anterior infarction, but in only 0%-5% of patients with
inferior infarction [2, 4,11,12,20,23,38,41]. Most thrombi develop during the first
3-10 days after the acute infarction, but new thrombi have been detected as late as
3 months after the acute infarction [38]. Standard anticoagulation therapy does not
prevent ventricular thrombi [4, 41], but high dose intravenous heparin started early
after the infarction may reduce thrombus incidence [28]. However, short-term treat-
ment (10 days) had no effect on the overall incidence of thrombi during the first
6 months after the acute infarction [22]. No treatment which totally prevents thrombi
has yet been described.
Echocardiography and Embolic Sources in the Heart 231

Fig.!. Left ventricular thrombus in the apical region 7 days after acute anterior infarction. The
thrombus has been present for 4 days, and has started to develop central echolucency. The patient
suffered stroke 5 days later, and the thrombus then disappeared

Left ventricular thrombi are mainly located in the left ventricular apex (Fig. 1);
apical akinesia or dyskinesia is a prerequisite for thrombus development. Patients re-
gaining apical wall motion do not form apical thrombi [22].
The following cross-sectional criteria for the diagnosis of left ventricular thrombi
have been widely accepted [2, 4, 11, 12,20,23,38,41]:
1. Definite thrombus margin
2. Clear delineation of the endocardium at the site of the thrombus
3. Thrombus visible with two different transducer positions
Structures with shadowy margins or unidentifiable borders against the ventricular
lumen may represent artifacts (reverberations, sidelobes, blood stases). A structure
which is only seen with one echo view and which disappears with a small tilting of the
transducer should not be diagnosed as a thrombus. These are the major criteria. Ad-
ditional ones have been used, including independent movement of an intracavitary
structure, apical location, and changes of intracavitary structure during serial exami-
nations. The latter may be useful in patients where the first echo scan is equivocal,
whereas independent movement and apical location probably should not be required
in order to diagnose a thrombus.
It is important to examine the apical region with multiple transducer positions
when looking for apical thrombi. Many clots will be missed if the apex is not examined
with axillary transducer positions, cross sectioning the apical region in small slices
(Fig. 2). With a standard apical four-chamber or two-chamber view, the true apex is
often missed, and small thrombi escape the echo beam.
232 K.-A. Johannessen

Fig. 2. This left ventricular thrombus could not be seen in a standard apical two- or four-chamber
view. However, a lateral and oblique transducer position revealed this apical thrombus

The largest risk of embolization is during the first 3 weeks after the infarction
(75%-80%), and most thrombi which cause embolization do so within 3 months
after the acute infarction [4,20,39,41]. Approximately 80% of the diagnosed emboli
occur in the cerebral circulation. Thus, the patients with left ventricular thrombi
represent a group of patients with high risk of stroke.
In several reports, anticoagulation therapy has failed to prevent embolization
from left ventricular thrombi [1, 2, 19-21]. It therefore seems that prevention of
thrombi will be superior to treatment of thrombi with anticoagulants after a clot has
developed. However, it should be emphasized that no sufficiently large randomized
trial dealing with this question has been carried out yet.
It has been suggested that left ventricular thrombi develop only in transmural in-
farctions with endothelial necrosis. However, the fact that ventricular thrombi occur
in 10% -50% of patients with nonischemic cardiomyopathy [14] does not support the
hypothesis that acute ischemic endocardial necrosis is necessary for thrombus forma-
tion.
The following cross-sectional characteristics of left ventricular thrombi are asso-
ciated with embolization [16, 21, 39]:
1. Protrusion of thrombus into the ventricular lumen
2. Independent movement of thrombus during cardiac contractions
3. Central echolucency
Serial examinations of thrombi may be useful, particularly in patients where ini-
tially flat, nonmobile thrombi develop central echolucency and vigorous indepen-
dent movements during the first 2-3 weeks after the acute infarction [19].
Whereas acute thrombi have been associated with a high risk of embolization,
little is known about the embolic risk of chronic thrombi. However, some reports
Echocardiography and Embolic Sources in the Heart 233

have indicated that embolization may occur from ventricular thrombi as late as 2 years
after an acute myocardial infarction.
The disappearance of a previously identified left ventricular thrombus after a pa-
tient has suffered cerebral or peripheral arterial occlusion may be demonstrated with
cross-sectional echocardiography [1,19]. This is strong evidence for the thrombus as
an embolic source. Persistent apical akinesis in such patients implies a substantial
risk for reformation of thrombus and reembolization. The presence of a thrombus
remaining after embolization may also be associated with a high risk of new emboli
despite anticoagulation treatment [19]. Therefore, before starting with anticoagulants,
the risk of reembolization should be carefully balanced against the risk of increased
secondary hemorrhage into an embolic cerebral infarct. Thus, knowledge about the
embolic source in these patients is of great importance.

Left Atrial Thrombi

The clinical importance of atrial thrombi is evident in patients who are candidates for
electroconversion, but atrial fibrillation per se has also a high incidence of cerebral
embolism. Atrial fibrillation has been found in as many as 30% of patients with cere-
bral embolic infarctions, and 10% -30% of patients with atrial fibrillation and stroke
may have recurrent embolization during the first 2-3 years after the initial cerebral
attack [8, 10, 13, 18, 33, 43]. In a study of 140 patients, Sage and coworkers reported
new cerebral vascular accidents in 20% of patients not on anticoagulation treatment
during a follow-up of 9 years [32].
The true incidence of atrial thrombi detectable with cross-sectional echocardiog-
raphy is unknown. Although several studies with echocardiography have shown a
low detection rate of atrial thrombi in patients with atrial fibrillation, the technique
has unique potentials for this purpose [26, 34, 35, 37, 40]. The previously reported
low rates of atrial thrombus detection by cross-sectional echocardiography may pri-
marily be a consequence of the technique of examination.
The study of Herzog et al. [17] demonstrated an improved technique for visualiz-
ing left atrial appendage thrombi. They demonstrated the importance of the correct
beam plane for imaging the greatest tomographic area of the atrial appendage. Late
systole should be used for analysis, because maximal dilatation of the left atrial ap-
pendage occurs at this time. The transducer should be oriented for a standard para-
sternal short-axis view at the aortic level, and the beam plane angulated superiorly
with lateral tilting of the transducer, so that aortic valves, but not the tricuspid valves,
are imaged [17].
Absence of coordinated contractility, as in atrial fibrillation, and dilatation of the
atrium, as in mitral valve disease, are assumed to be prerequisites for atrial thrombus
formation. In a study using cross-sectional echocardiography [35], 33 atrial thrombi
were identified in 293 patients with rheumatic mitral valve disease. Twenty-one ad-
ditional patients had thrombus in the left atrium at operation. Eleven of these un-
detected thrombi were located in the left atrial appendage. Schweizer et al. [34] iden-
tified only five patients with atrial thrombi among 92 patients having open heart
surgery, whereas an additional eight had an atrial thrombus located in the atrial ap-
pendage.
234 K.-A. Johannessen

As with ventricular thrombi, trabeculation and reverberations are important dif-

ferential diagnoses to atrial thrombi. In the immediate future the detection rate of
left atrial thrombi may be vastly improved by the use of transesophageal echocar-
diography.

Endocarditis

Three percent of all cerebral emboli arise from valvular infections, and major cere-
bral emboli have been reported in 6% -30% of patients with endocarditis at autopsy.
Such emboli are more often seen with mitral than with aortic valve endocarditis, and
staphylococci are responsible for approximately 70% [25,31,42].
Valve endocarditis may be present even when vegetations are not demonstrated
on echocardiography, and serial cross-sectional and M-mode echocardiographic ex-
aminations are often necessary to verify the diagnosis. Echocardiographic character-
istics are shaggy echoes on the leaflets and irregular densities from the valves into
either the left ventricular outflow tract in systole (aortic valve vegetation) or the left
ventricle during diastole (mitral vegetation, Fig. 3). Globular, polypoid masses with
elongated lesions having turbulent movements during cardiac contractions may also
be characteristic.

Prosthetic Valves

Heart valve prostheses may be the sites of thrombi detected by echocardiography,

but additional examinations often have to be done in these patients to verify or rule
out embolic lesions. Although clots may be confused with reverberations from the
prosthesis itself, the findings of echo densities in the vicinity of a prosthetic valve
may be suggestive of thrombosis of the valve, especially if these findings represent
changes from previous examinations of the patient. As distinct thrombus masses are
rather rare, one will have to rely on the indirect evidence of a malfunctioning pros-
thesis [9].

Mitral Valve

Embolization from the mitral valve may occur in patients with valve calcification,
endocarditis, and mitral valve prolapse. Real time imaging and serial examinations
may be necessary to diagnose vegetations which may show vigorous movements dur-
ing valve closure or opening. The diagnosis of thrombotic or infective lesions may be
difficult in patients with calcifications in the valvular annulus, but irregular densities
moving between the left atrium and ventricle through the valve opening, globular
polypoid masses with or without elongated lesions, and independent movements are
usually diagnostic.
 Echocardiography and Embolic Sources in the Heart 235

B
Fig. 3 A, B. Apical four-chamber views in a 50-year-old man with mitral valve prolapse (anterior
leaflet) and a vegetation on the lateral leaflet. The vegetation showed vigourous movements be-
tween the left ventricle and left atrium (A, B), and was markedly reduced in size after the patient
had suffered a cerebral infarction, believed to be embolism from the vegetation
236 K.-A.Johannessen

Mitral Valve Prolapse

Mitral valve prolapse, which is inherited in an autosomal dominant mode with a

varying prevalence both between sexes and races [7, 36], has caused much interest in
echocardiography. Its importance as a cause of cerebral ischemia remains to be satis-
factorily documented, as few systematic antemortem data are available. In studies
with necropsied patients, 1%-2% have been described as having mitral valve pro-
lapse.
With stringently used M-mode criteria, the prevalence of the condition may be
about 1%-4% [5,24]. Less stringent criteria giving higher prevalences may reflect
lower specificity rather than higher sensitivity. The reported incidence of cerebral
ischemia in young patients with mitral valve prolapse has varied from 2% to 20%.
Several major diagnostic criteria for mitral valve prolapse have been suggested:
(a) Marked displacement of mitral leaflets with coaptation point at the level of the
annulus or on its left atrial side
(b) Moderate systolic displacement of the leaflets with at least moderate regurgita-
tion, chordal rupture, or annular dilatation
(c) Marked (> 3mm) late systolic buckling posterior to the C-D line [29].
Mild prolapse signifies that the bellies of either or both leaflets are barely detectable
on the left atrial side of the annulus, moderate prolapse that they are readily seen,
and marked prolapse that recognition is quite obvious. The clinical relevance of such
classifications has not been extensively documented.
There are more female than male patients with mitral valve prolapse suffering
cerebral vascular accidents, as might be expected from the higher incidence of
women having mitral valve prolapse. Mitral valve prolapse also tends to be more fre-
quent in stroke patients below 50 years of age, a fact that may reflect the higher inci-
dence of atherosclerotic disease in the elderly [15]. Barnett et al. found mitral valve
prolapse to be 4.5 times more frequent in young patients with cerebral ischemia com-
pared to control subjects [3]. Among patients with mitral valve prolapse and cerebral
ischemic disease, approximately 30% of the patients suffer TIA alone, whereas
stroke occurs in 60%. Approximately 20% sustain recurrent episodes.
The most likely mechanism of cerebral embolism is a noninfected thrombus at
the prolapsing mitral leaflet [30]. One study found left atrial masses in eight out of
15 patients with mitral valve prolapse and cerebral ischemia [27]. In the same study,
no atrial masses were found in 30 patients with mitral valve prolapse and no neuro-
logical symptoms.

Conclusions

Modem echocardiography has opened vast possibilities for studying the heart as an
embolic source. Although the clinical usefulness of the method in all conditions men-
tioned above has been amply documented, a substantial amount of work remains to
be done. The method should be extended to larger populations and the data treated
Echocardiography and Embolic Sources in the Heart 237

epidemiologically. Furthermore, therapeutic trials on emboli prevention in the heart

should be conducted, using echocardiography as one important source of informa-
tion. Further refinement in echo techniques as well as the combined use of two-
dimensional echo and doppler techniques may in the near future improve both
specificity and sensitivity of the method.

References

1. Arvan S (1985) Persistent intracardiac thrombi and systemic embolization despite anticoagulant
therapy. Am Heart J 109: 178-181
2. Asinger RW, Mikell FL, Elsperger J, Hodges M (1981) Incidence of left ventricular thrombosis
after acute transmural myocardial infarction. N Engl J Med 305: 297-302
3. Barnett HJM, Boughner DR, Taylor DW, Cooper PE, Kostuk WJ, Nichol PM (1980) Further
evidence relating mitral valve prolapse to cerebral ischemic events. N Engl J Med 302: 139-144
4. Benichou M, Aubry J, Larbi MB (1983) Detection of left ventricular thrombi in the acute phase
of myocardial infarction by two-dimensional echocardiography: a series of 103 cases (in French).
Arch Mal Coeur 76: 1012-1019
5. Bryhn M, Persson S (1984) The prevalence of mitral valve prolapse in healthy men and women
in Sweden. Acta Med Scand 215: 157-160
6. Cerebral Embolism Task Force (1986) Cardiogenic brain embolism. Arch Neurol 43: 71-84
7. Devereux RB, Brown T, Kramer-Fox R, Sachs H (1982) Inheritance of mitral valve prolapse:
effect of age and sex on gene expression. Ann Intern Med 97: 826-832
8. Fairfax AJ, Lambert CD, Leatham A (1976) Systemic embolism in chronic sino-atrial disorder.
N Engl J Med 295 : 190-192
9. Feigenbaum H (1986) Echocardiography, 4th edn. Lea and Febiger, Philadelphia
10. Fischer CM (1979) Reducing risks of cerebral embolism. Geriatrics 34: 59-66
11. Foale R, Smith G, Nihoyannopoulo P, Maseri A (1984) Left ventricular thrombus in acute myo-
cardial infarction: best left alone? Circulation [Suppl] 70:11-310
12. Friedman M, Carlson K, Marcus FI, Woolfenden JM (1982) Clinical correlations in patients
with acute myocardial infarction and left ventricular thrombus detected by two-dimensional
echocardiography. Am J Med 72: 894-898
13. Godtfredsen J (1975) Etiology, course and prognosis. A follow up study of 1212 cases. Munks-
gaard, Copenhagen
14. Gottdiener JS, Gay JA, Van Voorhees L, DiBianco R, Fletcher RD (1983) Frequency and
embolic potential of left ventricular thrombus in dilated cardiomyopathy: assessment by 2-
dimensional echocardiography. Am J Cardiol 52: 1281-1285
15. Hart RG, Easton JD (1982) Mitral valve prolapse and cerebral infarction. Stroke 13:429-430
16. Haugland JM, Asinger RW, Mikell FL, Elsperger KJ, Hodges M (1984) Embolic potential of
left ventricular thrombus detected by two-dimensional echocardiography. Circulation 70: 588-
598
17. Herzog CA, Bass D, Kane M, Asinger R (1984) Two-dimensional echocardiography imaging of
left atrial appendage thrombi. J Am Coll Cardiol 3 : 1340-1344
18. Hinton RC, Kistler JP, Fallon JT, Friedlich AL, Fischer CM (1977) Influence of etiology of
atrial fibrillation on incidence of systemic embolism. Am J Cardiol 40: 509-513
19. Johannessen K-A (1987) Peripheral emboli from left ventricular thrombi of different echocar-
diographic appearance in acute myocardial infarction. Arch Intern Med 147:641-644
20. Johannessen K-A, Nordrehaug JE, von der Lippe G (1984) Left ventricular thrombosis and
cerebral vascular accident in acute myocardial infarction. Br Heart J 51: 553-556
21. Johannessen K-A, von der Lippe G, Vollset SE (1986) Risk of peripheral emboli from left ven-
tricular thrombi in acute myocardial infarction. Xth World Congress in Cardiology abstract,
p558
22. Johannessen K-A, Nordrehaug JE, von der Lippe G (1987) Left ventricular thrombi after short
time high dose anticoagulants in acute myocardial infarction. Eur Heart J 8: 975-980
238 K.-A.Johannessen: Echocardiography and Embolic Sources in the Heart

23. Keating EC, Gross SA, Schlamowitz RA, Glassmann J, Mazur JH, Pitt WA, Miller D (1983)
Mural thrombi in myocardial infarctions. Am J Med 74:989-995
24. Levy D, Savage D (1987) Prevalence and clinical features of mitral valve prolapse. Am Heart J
113: 1281-1290
25. McDevitt E (1965) Treatment of cerebral embolism. Mod Treat 2: 52-63
26. Mikell FL, Asinger RW, Rowke T, Hodges M, Sharma B, Francis GS (1979) Two-dimensional
echocardiographic demonstration of left atrial thrombi in patients with prosthetic mitral valves.
Circulation 60: 1183-1190
27. Nichol P, Kertesz A (1979) Two-dimensional echocardiographic detection of atrial thrombus in
patients with mitral valve prolapse and strokes. Circulation [Suppl] 60: II-18
28. Nordrehaug JE, Johannessen KA, von der Lippe G (1985) Usefulness of high-dose anticoagulants
in preventing left ventricular thrombus in acute myocardial infarction. Am J Cardiol 55: 1491-
1493
29. Perloff JK, Child JS, Edwards JE (1986) New guidelines for the clinical diagnosis of mitral valve
prolapse. Am J Cardiol 57: 1124-1129
30. Pomerance A (1969) Ballooning deformity (mucoid degeneration) of atrioventricular valves.
Br Heart J 31: 343-351
31. Pruitt AA, Rubin RM, Karchmer AW, Duncan GW (1987) Neurologic complications of bac-
terial endocarditis. Medicine 57: 329-343
32. Sage n, Van Uitert RL (1983) Risk of recurrent stroke in patients with atrial fibrillation and
non-valvular heart disease. Stroke 14: 537-540
33. Sawyer CG, Bolin LB, Stevens EL, Daniel LB, O'Neill NC, Hayes DM (1958) Atrial fibrilla-
tion: its etiology, treatment and association with embolization. South Med J 51: 84-93
34. Schweizer P, Bardos P, Erbel R, Meyer J, Merx W, Messmer BJ (1981) Detection of left atrial
thrombi by echocardiography. Br Heart J 45: 148-156
35. Shrestha NA, Moreno FL, Narciso FL, Narciso FV, Torres L, Calleja HB (1983) Two-dimen-
sional echocardiographic diagnosis of left atrial thrombus in rheumatic heart disease. A clinico-
pathological study. Circulation 67: 341-347
36. Strahan NV, Murphy EA, Fortuin NJ, Come PC, Humpheries JO (1983) Inheritance of mitral
valve prolapse syndrome: discussion of a three-dimensional penetrance model. Am J Med 74:
967-972
37. Sunagawa K, Onta Y, Tanaka S, Kikuchi Y, Nakamura M, Hirata T (1980) Left atrial ball
thrombus diagnosed by two-dimensional echocardiography. Am Heart J 100: 89-94
38. Visser CA, Kan G, Lie KI, Durrer D (1983) Left ventricular thrombus following acute myo-
cardial infarction: a prospective serial echocardiographic study of 96 patients. Eur Heart J 4:
333-337
39. Visser CA, Kan G, Meltzer RS, Dunning AJ, Roelandt J, Van Corler M, de Koning H (1985)
Embolic potential of left ventricular thrombus after myocardial infarction: a two-dimensional
echocardiographic study of 119 patients. J Am Coli Cardiol5: 1276-1280
40. Warda M, Garcia J, Pechacek LW, Massumkhani A, Hall RJ (1985) Auscultatory and echocar-
diographic features of mobile left atrial thrombus. J Am Coli Cardiol5: 379-382
41. Weinreich DJ, Burke JF, Pauletto FJ (1984) Left ventricular mural thrombi complicating acute
myocardial infarction. Ann Intern Med 100: 789-794
42. Weinstein L, Schlesinger JJ (1974) Pathoanatomic, pathophysiologic and clinical correlation in
endocarditis. N Engl J Med 291: 832-837
43. Wolf PA, Dawler TR, Thomas HE, Kannel WB (1978) Epidemiologic assessment of chronic
atrial fibrillation and risk of stroke: the Framingham study. Neurology 28: 973-977
Chapter 18

31P Nuclear Magnetic Resonance Spectroscopy

of Cerebral and Cardiac Ischemia
M. M. COHEN, S. J. Kopp, J. W. PETIEGREW, and T. GLONEK

Introduction

Despite major differences in structure and function between heart and brain, energy
metabolism is qualitatively similar in the two organs. The energy required for physio-
logical activity, for preservation of structural integrity, for maintenance of mem-
branes, and for growth and reproduction is provided through the degradation of
adenosine triphosphate (ATP) to the diphosphate (ADP). The energy available in
phosphocreatine (PC) can be utilized through transfer of the high-energy phosphate
to ADP to form additional ATP.
Under circumstances, such as ischemia, hypoxia, or anoxia, when high-energy
phosphate utilization exceeds supply, levels of energy-rich phosphates have been
demonstrated to be altered, using traditional biochemical techniques. PC concentra-
tions are depleted rapidly, but ATP levels are maintained temporarily. If oxygen
deprivation persists, however, those levels fall as well [6].
Under resting circumstances the mature brain, although comprising but 2% of
total body weight, utilizes 20% of the oxygen consumed. As a consequence of this
disproportionate energy requirement, the brain is far more vulnerable to interrup-
tion of oxygen supply than the heart and cannot tolerate 10 min of anoxia without
sustaining irreversible damage.
Chemical analysis of high-energy and other phosphates in biological tissue has
been extremely limited by the need to rapidly fix, then destroy, the tissue to be
analyzed. Additionally, techniques for qualitative and quantitative analysis are lack-
ing for a large number of these compounds in the concentrations in which they exist
in mammalian organs and tissues. Nevertheless, analysis of these materials has been
helpful in evaluation of potentially beneficial pharmacological agents.
31p nuclear magnetic resonance (NMR) spectroscopy more elegantly serves the
same analytic functions as biochemical analysis. Sensitivity is markedly enhanced,
and compounds not previously reported as being present in brain have been iden-
tified [10]. In addition, the analysis can be carried out in tissue extracts, in a perfused
organ and in an intact animal or human in vivo. Determination of the chemical shift
also allows calculation of pH in either the perfused or in situ organ. As a conse-
quence of these three factors, NMR spectroscopy has a great potential for pharmaco-
logical studies.
Although many values for phosphorylated compounds are essentially similar to
those obtained by biochemical methods, there are areas of difference. 31p NMR de-
tects less than 10% of both inorganic phosphate (Pi) and ADP as determined by
chemical analysis [2]. This suggests that classic tissue extraction methods are not
240 M. M. Cohen et al.

representative. ADP bound to actin-myosin complexes or other pools could be re-

leased, as could Pi during the analytic process.

Heart

31p NMR spectroscopy has been employed in both the isolated perfused mammalian
heart and in experimental animals and humans in situ. In 1980, Grove et al. [11] re-
ported their findings on rat heart, both perfused and in vivo. In the latter experi-
ments, a surface coil was placed directly on the heart through abdominal and trans-
thoracic incisions. The two spectra obtained were sufficiently similar to allow assign-
ments to be made to eight peaks related to their chemical shifts: (1) f3-ATP, (2) a-
ATP + a-ADP, (3) y-ATP + f3-ADP, (4) PC, (5) Pi + 2-phosphate of2,3-diphospho-
glycerate, (6) 3-phosphate of 2,3 diphosphoglycerate, (7) intracellular Pi, and (8) Pi
from perfusion buffer.
During ventilatory arrest PC decreased to almost undetectable levels over 10 min ,
and ATP concentration then decreased until no longer detectable at 17 min. Tissue
pH fell to 6.4 during the anoxic period.
Brooks and Willis [4] measured the rate of decline and resynthesis of PC in per-
fused guinea pig heart during 12 min of global ischemia produced by stopping perfu-
sion. At the termination of the ischemic period, PC had declined to 22% of its con-
trol value accompanied by a more than fourfold rise in Pi. A TP values were un-
changed, and the pH fell from 7.00 to 6.68. Resynthesis of PC proceeded at a rate 14
times more rapid than did breakdown and tended to overshoot the original concen-
tration.
In a subsequent modification, high-resolution NMR of the myocardium has been
obtained by passing an eliptical 31p receiver coil through the peripheral blood vessels
into either the right or left ventricle of a dog [13]. In this manner a spectrum could
be obtained from a defined area of myocardium in less than 7 min.
The human heart has been studied by placing a flat coil outside the chest wall.
With depth-resolved spectroscopy, Bottomley [3] noted a large decrease in the PC!
ATP ratio between the shallow and deep slices. Changes in peak amplitude occurred
as the transition was made from predominantly skeletal to cardiac muscle.
The potential benefits of several pharmacological agents have been studied with
31p NMR in the perfused heart by Nunnally and Bottomley [16]. Regional ischemia
was produced in rabbit heart through ligature of the left anterior descending coro-
nary artery; 30% -40% of the left ventricular wall was thus rendered ischemic. The
addition of 1 mg!l of verapamil to the perfusate at least 5 min prior to ligation main-
tained PC and ATP at nearly normal concentrations. Chlorpromazine produced a
qualitatively similar, but less dramatic effect. When verapamil was administered
after ligation, at a time when PC levels were already decreasing, a 220% increase in
PC concentration was observed.
The efficacy of propranolol in cardiac ischemia has also been studied with 31p
NMR in perfused guinea pig hearts. The addition of 1 mg!l of DL-propranolol hydro-
chloride prior to clamping of the perfusion tube reduced the magnitude of PC loss re-
sulting from a 35-min period of ischemia, but had no effect on the diminution of con-
31p NMR of Heart and Brain 241

centrations of ATP. Significantly, propranolol reduced the level of acidosis with

ischemia of as long as 60 min duration [IS].
The effects of propranolol have been studied with ischemia of shorter duration in
the perfused guinea pig heart [15]. Cardiac pH declined from 7.41 + 0.04 to 6.61 +
0.04 within 12 min and returned to the preischemic value within 3 min. Both PC and
ATP levels fell with the ischemia as Pi concentrations rose. Values for both PC and
Pi returned to normal with cessation of the ischemia, but A TP recovery was poor.
Both L- and D-propranolol suppressed the fall in pH and attentuated the changes in
high-energy phosphates.

Brain

As with the heart, 31p NMR studies have been carried out in extracts of cerebral tis-
sue, as well as in perfused brain, and in both experimental animals and humans in
vivo. Extracts of whole brain [9] and of tissue slices fixed after metabolizing in vitro
[10] have been particularly successful in increasing the sensitivity of the method.
Thirty-two separate peaks have been identified in whole brain, and among them are
substances that have yet to be characterized chemically. At least one of these is
metabolically active: at the resonance position of 0.S5 sigma. Unlike preparations
from whole brain obtained after in situ freezing, tissue slice extracts contained a
group of as yet unidentified phosphodiesters at -0.73 sigma.
The phosphomonoester signal at 3.S4 sigma is prominent and characteristic of
neural tissue. This peak constituted 9.57% of the acid-soluble phosphates of whole
guinea pig brain, and 7% -9% of those in incubated cerebral cortex slices. The peak
has now been identified as phosphoethanolamine [7]. This compound is present in
particularly large amounts in infant brain and has been observed in placenta and in
certain neoplasms [S].
When cortical slices were prepared from guinea pig or gerbil brain incubated in
an oxygen atmosphere for 15 min, then incubated for another 15 min in an atmo-
sphere of 100% nitrogen, concentrations of A TP and PC fell and those of Pi rose, as
has been reported employing traditional biochemical analyses. Control values were
achieved once again after 15 min of restoration of oxygen. Inosine monophosphate
(IMP) and a-glycerol phosphate also more than doubled in concentration during the
anoxic period and were restored to normal levels in an oxygen atmosphere. The acid
labile phosphate at 0.S5 sigma also diminished in response to anoxia and recovered
when oxygen was restored.
Following the demonstration that surface coils may be effective in studying
energy metabolism in the intact animal [13], 31p NMR was employed to determine
concentrations of PC, A TP, and Pi and the pH following unilateral carotid artery oc-
clusion in the gerbil [19]. The resulting homolateral ischemia led to changes in meta-
bolites in 50% of the animals studied. PC and A TP fell in unison in the anterior por-
tion of the homolateral hemisphere until the signal disappeared. The Pi signal was
intense and pH fell to 6.61 + 0.05. An increased Pi signal and a drop in pH were ob-
served in both the cerebellum and the contralateral cerebral hemisphere. There was,
however, no concomitant change in PC or ATP. Those animals exhibiting the greatest
242 M. M. Cohen et al.

degree of histologic change also exhibited the most prominent depletion of high-
energy phosphates.
In contrast to the gerbil studies, those of canine brain demonstrated PC levels to
decline by over 60% before those of ATP began to fall [13]. These findings are in ac-
cord with earlier biochemical analyses [6].
Anoxia was found to produce more profound effects on cerebral high-energy
phosphate levels in the aged rat than in the young mature animal [7]. When the rats
were exposed to a stream of nitrogen in a closed container, concentrations of PC fell
by half in mature animals and those of ATP by one-third. In the aged (26 months)
animals PC fell to less than one-third of the control value and A TP to one-half. The
concentrations of Pi and IMP also rose to a greater extent in the aged anoxic animal
than in the young. Fifteen minutes after the restoration of oxygenation, PC concen-
trations were reconstituted to a level above that of the control state. A TP, Pi, and
IMP levels returned to normal.
The effects of oxygen deprivation in the isolated perfused rat brain were similar
whether the perfusate was administered by constant pressure or constant flow [14].
Concentrations of high-energy phosphates and Pi were identical to those observed in
the nonperfused brain and were consistent with those obtained with other preparations.
The first 31p spectrum of human brain was reported by Cady and his associates [5]
in 1983. They studied seven premature infants of 33-40 weeks gestational age from
44 h to 17 days after birth. Those infants exhibiting severe birth asphyxia showed PC/
Pi ratios that were decreased below those of normal infants. Those ratios increased
as clinical improvement progressed. The administration of mannitol to two infants
suffering birth asphyxia resulted in a rapid increase in the ratio. The tissue pH ex-
hibited no systematic change.
In their study of infants with neonatal seizures, Younkin and his associates [20]
directed their attention to a circumstance where energy utilization was increased and
obtained findings similar to those observed in ischemia when energy supply was lack-
ing. They found a 50% decrease in the PC/Pi ratio during seizure episodes. The post-
ictal ratios were increased, such as has been noted following periods of anoxia. They
attributed this increase to postictal inhibition.
Those infants exhibiting focal seizures exhibited unilateral diminution of the
ratio, while the changes were bilateral in those with generalized seizures. The success-
ful treatment of seizures with phenobarbital in one patient during the study period
resulted in an immediate increase in the PC/Pi ratio from 0.7 to 1.2. The five infants
exhibiting ratios of less than 0.8 during the seizures subsequently demonstrated long-
standing neurological damage.

Summary and Conclusions

31p NMR spectroscopy is a powerful new tool for studying the phosphorus metabo-
lism of both heart and brain. This technique has served to provide dynamic informa-
tion about the energy state of both organs. It also may be useful as an indicator of
tissue damage and of prognosis following ischemia. The perfused organ has already
yielded valuable information concerning the potential benefits of varius pharmaco-
logical agents and thus indicates the availability of effective systems for drug testing.
31p NMR of Heart and Brain 243

Future studies are also expected to provide new information on the chemical com-
position of heart and brain, as well as indications of the physiological significance of
a number of newly recognized compounds.

References

1. Ackerman JJH, Grove TH, Wong GG, Gadian DG, Radda GK (1980) Mapping of metabolites
in whole animals by 31p NMR using surface coils. Nature 283: 167-170
2. Balaban RS (1984) The application of nuclear magnetic resonance to the study of cellular physi-
ology. Am I Physiol 246: C 10-19
3. Bottomley PA (1985) Non-invasive study of high energy phosphate metabolism in human heart
by depth resolved 31p NMR spectroscopy. Science 229:769-772
4. Brooks WM, Willis RJ (1983) 31p NMR study of the recovery characteristics of high energy
phosphates and intracellular pH after global ischaemia in the perfused guinea pig heart. I Mol
Cell Cardiol 15: 495-502
5. Cady ED, Costello AMdeL, Dawson MJ, Delpy DT, Hope PL, Reynolds EOR, ToftsPS, Wilkie
DR (1983) Non invasive investigation of cerebral metabolism in newborn infants by phosphorus
nuclear resonance spectroscopy. Lancet 1: 1059-1062
6. Cohen MM (1962) Effect of anoxia on the chemistry and morphology of cerebral cortex slices in
vitro. I Neurochem 9: 337 -344
7. Cohen MM, Kopp SI, Pettegrew IW, Glonek T (1984) 31p Nuclear magnetic resonance studies
of anoxia in aged rat brain. Eur Neurol23: 141-143
8. Cohen MM, Lin S (1962) Acid soluble phosphates in the developing rabbit brain. I Neurochem
9:345-352
9. Cohen MM, Pettegrew IW, Kopp SI, Minshew N, Glonek T (1984) P-31 nuclear magnetic reso-
nance analysis of brain: Normoxic and anoxic brain slices. Neurochem Res 9: 785-801
10. Glonek T, Kopp SI, Kot E, Pettegrew IW, Harrison WH, Cohen MM (1982) P-31 nuclear
magnetic resonance analysis of brain. The perchloric acid extract spectrum. I Neurochem 39:
1210-1219
11. Grove TH, Ackerman JJH, Radda GK, Bore PI (1980) Analysis of rat heart in vivo by phos-
phorus nuclear magnetic resonance. Proc Nat! Acad Sci USA 77: 299-302
12. Hilberman M, Subramanian VH, Haselgrove I, Cone JB, Egan IW, Gyulai L, Chance B (1984)
In vivo time resolved brain phosphorus nuclear magnetic resonance. I Cereb Blood Flow Metab
4:334-342
13. Kantor HL, Briggs RW, Balaban RS (1984) In vivo 31p nuclear magnetic resonance measure-
ment in canine heart using a catheter coil. Circ Res 55: 261
14. Kopp SI, Krieglstein I, Friedank A, Rachman A, Seibert A, Cohen MM (1984) P-31 nuclear
magnetic resonance analysis of brain: II. Effects of oxygen deprivation on isolated perfused and
non perfused rat brain. I Neurochem 43: 1716-1731
15. Nakazawa M, Katano Y, Imai S, Matsushita K, Okuchi M (1982) Effects of 1- and D-propranolol
on the ischemic myocardial metabolism of the isolated guinea pig heart as studied by 31p NMR.
I Cardiovasc Physiol 4: 700-704
16. Nunnally RL, Bottomley PA (1981) Assessment of pharmacological treatment of myocardial in-
farction by phosphorus-31 NMR with surface coils. Science 211: 177-180
17. Pettegrew IW, Kopp SI, Dadok I, Minshew NI, Feliksik IM, Glonek T, Cohen MM (1986)
Chemical characterization of a prominent phosphomonoester resonance from mammalian brain.
31p and lH NMR analysis at 4.7 and 14.1 Tesla. I Magnetic Resonance 67: 443-450
18. Pieper GM, Todd GL, Shao TW, Salhany IM, Clayton EC, Eliot RS (1980) Attenuation of
myocardial acidosis by propranolol during ischaemic arrest and reperfusion: evidence with 31p
nuclear magnetic resonance. Cardiovasc Res 14: 646-653
19. Thulborn KR, du Boulay GH, Duchen LW, Radda G (1982) A 31p nuclear magnetic resonance
in vivo study of cerebral ischaemia in the gerbil. I Cereb Blood Flow Metab 2: 299-306
20. Younkin DP, Delivoria-Papadopoulos M, Maris I, Donlon E, Clancy R, Chance B (1986) Cere-
bral metabolic effects of neonatal seizures measured with in vivo 31p NMR spectroscopy. Ann
NeuroI20:513-519
Chapter 19
Oxygen Radicals in Heart and Brain Tissue Injury
K. YTREHUS and O. D. MJ0S

Oxygen Radicals

Free radicals, well-known species in chemistry, have recently received attention in

medicine because of new knowledge about their potential toxic effects. Free radicals
are by definition molecular species which contain an odd number of electrons. This
gives them the properties of being highly chemically reactive and of inducing chain
reactions. In medicine the focus has been on oxygen radicals because there are a
number of biological oxidations, both enzymatical and spontaneous, which generate
oxygen radicals [16, 18,47,66].
The best known oxygen radicals are products of univalent reduction of oxygen to
water [22] (Fig. 1). One electron reduction yields a superoxide anion. The protonated
form of two electron reduction is hydrogen peroxide, not strictly a radical species,
but a strong oxidant. Further reduction of oxygen with three electrons gives, in its
protonated form, the hydroxyl radical. This is an extremely reactive compound with
short life time, and is nonspecific and indiscriminate in reacting with other com-
pounds. It is important to note that transition metals, especially iron, are effective
catalysts for hydroxyl radical generation [3, 26] (Fig. 2). Among the possible sources
of oxygen radicals in living tissue during normal and pathological conditions, the
autooxidation of different compounds such as hemoglobin, myoglobin, cytochrome
C, and catecholamines may be important. Organelles such as mitochondria, micro-

THE UNIVALENT PATWAY OF

OXYGEN REDUCTION

Fig. I. The univalent pathway of oxygen reduction.

The different species of reduced oxygen are shown in
protonated or nonprotonated form, depending on the
form dominating at pH 7.4
Oxygen Radicals in Heart and Brain Tissue Injury 245

02+Men+1---+. Me"+°2

Men+H.f>2 - - - + . Men+ 1+OH'+OH-

Fig. 2. The iron-catalyzed Haber-Weiss reaction leads
to hydroxyl radical formation

ATP

~
AMP

~
ADENOSINE

+
HYPOXANTHINE

protease

XANTHINE XANTHINE
Fig. 3. Part of the metabolic pathway

f
~thlne
oxidase
xanthine NAD+t:
dehydrogenase
of purine-nucleotide degradation is
demonstrated. The enzyme xanthine
oxidoreductase exists both as a dehydro-
OZ'1tA NA~H
+H genase and an oxidase. Xanthine oxi-
dase generates oxygen radicals by re-
URIC ACID URIC ACID ducing oxygen

somes, and nuclei have been shown to generate superoxide, and this is proposed to
be due to autooxidation of reduced components of electron transport assemblies
[10,63].
The existence of an oxygen radical-producing enzyme system, an NADPH
oxidase in the cell membrane of phagocytic cells, is well known, and oxygen radical
production is thought to be part of inflammation and antimicrobial defence [1, 4, 30,
67]. Several other enzymes also produce oxygen radicals; an important example is
xanthine oxidase [17, 70] (Fig. 3).

Protection Against Oxygen Radicals

To protect against the toxicity of oxygen radicals, all aerobic cells have a defence sys-
tem consisting of specific enzymes and antioxidants often called scavengers. The
most important of these scavengers is probably superoxide dismutase (SOD), which
catalyzes the dismutation of superoxide to hydrogen peroxide [7, 41] (Fig.4). Catalase,
mainly a peroxisomal enzyme, catalyzes the conversion of hydrogen peroxide into
water and oxygen [10] (Fig. 4). The cytoplasmatic selenium-containing enzyme
glutathione peroxidase reduces both lipid peroxides and hydrogen peroxide [10, 49,
246 K.Ytrehus and O.D.Mj!1ls

SOD
42+ °2 7 1t.P2+02
2H+

catalase
2H:P2 ~+2H:P

GSH-peroxldase.
H:P2+2GSH 2HtHGSSG
Fig. 4. The enzymatic reactions protecting
cells against superoxide, hydrogen peroxide,
ROOH+2GSH GSH-peroxldase. ROH W-Ll GSSG
+·7+ and lipid peroxides. SOD, superoxide dis-
mutase; GSH, reduced glutathione; GSSG,
oxidized glutathione; NADPH, nicotin-
GSSG+NADPH GSSG-reductase. 2GSH+NADP+ amide adenine dinucleotide phosphate

73] (Fig. 4). The function of vitamin E as a lipid-soluble antioxidant has been studied
for several years [60]. Other endogenous or exogenous substances able to act as anti-
oxidants or scavengers are glutathione, methionine, vitamin C, urea, dimethylsul-
phoxide, mannitol, and different phenols.

Oxygen Radicals in Disease

The toxicity of oxygen radicals is most likely due to oxidation of important com-
pounds in the cell. Polyunsaturated lipids (membrane phospholipids) and sulph-
hydryl-containing enzyme proteins are compounds known to react easily with oxygen
radicals. Oxygen radicals may also damage DNA, depolymerize hyaluronic acid, and
modulate nucleotide-cyclase activities and the action and synthesis of prostaglandins
and lipoxygenase products.
There are a number of pathological processes in which oxygen radicals may be in-
volved, e.g., pulmonary oxygen toxicity [55], toxicity of chemicals such as paraquate
and carbontetrachloride, cardiomyopathy after doxorubicin treatment [42], and car-
diomyopathy in selenium deficiency [45] and inflammatory states [11].

Oxygen Radicals in Brain Injury

With respect to the brain, acute hypertension and stroke have been investigated with
the aim of determining whether oxygen radicals playa role in these conditions [65].
Arteriolar dilatation due to experimental acute severe hypertension occurred in
parallel with an increased level of superoxide in the tissue, measured as superoxide
dismutase inhibitable reduction of nitroblue tetrazolium [65]. These results are in
agreement with our own experiments demonstrating a vasodilatation caused by oxy-
gen radicals in isolated rat hearts [70]. Increased levels of lipid peroxide breakdown
products (thiobarbituric acid reactive substances) in brain and serum have also been
demonstrated in stroke prone spontaneously hypertensive rats [61]. Cerebral contu-
Oxygen Radicals in Heart and Brain Tissue Injury 247

ischae.ia
inflal8lllation reperfusion toxic substances hyperoxia

IAX--/
leucocyte
activation
arachidonic
acid
\~
xanthine
oxidase
\
redox changes auto-
in cellular ~oxidations
J

~ l_oxygen
1/'/
radicals\.
-
irradiation of
water

/
NADPH t SH-containing
~oxidation of membrane
NADP I enzymes - cellular ~ polyunsaturated
compounds lipids

~l //
depressed energy metabolism
loss of contractility
loss of calcium control
loss of volume control
electrophysiological changes
increased mutagenicity

l
Icellular necrosisl

Fig. 5. Pathophysiological conditions proposed to lead to increased levels of oxygen radicals. This
may influence important cell functions and also result in cell death

sion, cortical laceration, intracerebral hematoma formation, and hemorrhagic corti-

cal infarction may all cause extravasation of blood, and this may lead to iron-induced
lipid peroxidation in the brain [3, 26]. This may be a significant form of tissue dam-
age in the brain since 50% of the dry weight of this organ is lipid [36]. In vivo lipid
peroxidation in rat brain followed intracortical Fe2 + injection and was also associated
with the appearance of seizures [62].
Some years ago special interest was focused on the possibility of oxygen radical
production during ischemia and reperfusion [13, 23]. With respect to ischemia of the
brain, this question has been examined by several researchers [8, 13,36,53]. A bene-
ficial effect of vitamin E on lipid peroxidation of reoxygenated homogenates from
ischemic rat brains has been reported [69], however, not all attemps to remove oxy-
gen radicals in cerebral ischemia have been successful. Forsman et al. [15] were not
able to detect any significant improvement by superoxide dismutase (SOD) and cata-
lase in neurological outcome 48 h after a lO-min episode of complete cerebral isch-
emia. Cerchiari et al. [8] describe the beneficial effects on cerebral blood flow and
on recovery of brain somatosensory-evoked potentials by adding SOD and the iron-
binding substance deferoxamine to dogs subjected to cardiac arrest. The role of oxy-
gen radicals in reperfusion damage of the spinal cord has also been examined [39].
248 K.Ytrehus and O.D.Mjj1js

In this case recombinant SOD administered at reperfusion reduced the incidence of

postischemic paraplegia and paraparesis in dogs after aortic cross clamping.

Oxygen Radicals in Heart Injury

Several investigators have also proposed that oxygen radicals participate in ischemia-
induced injury in the heart . An exacerbation of such an injury in connection with
reoxygenation after prolonged hypoxia was recognized by Hearse et al. in 1977 [29].
This has led to the suggestion that there is a relation between oxygen reintroduction
and reperfusion damage . The protective effect of the enzymes SOD and catalase and
of different oxygen-radical scavengers like dimethyl sulphoxide, mannitol , dimethyl-
thiourea and N-2-mercaptoproprionyl has been demonstrated by several inves-
tigators [2, 6, 19,20, 34,43 , 44,52,57,68, 71] . SOD and catalase were added to a
cardioplegic solution and during reperfusion of rat hearts subjected to 21O-min hypo-
thermic, cardioplegic ischemia and 60-min reperfusion [71]. This significantly im-
proved the recovery of contractile function, coronary flow, and metabolic parameters.
In addition, there was a decrease in the extent of intracellular edema in the myo-
cardium of hearts treated with SOD and catalase during ischemia and reperfusion.
However, some negative studies also exist [59, 64] and a proper explanation of the
difference in results among different research groups has not been found . This is
mainly due to restricted knowledge of both the influence of oxygen radicals and the
relative importance of the different sources of oxygen radicals in tissue .
In order to study the mechanisms by which oxygen radicals may function as me-
diators of cellular injury, we generated oxygen radicals in an isolated rat heart prepa-
ration. Contractile function [70] and ultrastructure [72] were examined. Isolated rat
hearts were cannulated via the aorta and retrogradely perfused [38] with a Krebs-
Henseleit bicarbonate buffer. Oxygen radicals were generated by adding hypoxan-
thine (0.96mM) and xanthine oxidase (0.025U/ml) to the buffer [12,17]. As illus-
trated in Fig. 6, perfusion with oxygen radicals resulted in a rapid decrease in con-
tractility and subsequent contracture. At the end of lO-min perfusion with oxygen
radicals and a 35-min recovery period the hearts were perfusion-fixed for ultrastruc-
tural examination using a glutaraldehyde-containing fixative. Ultrastructural alter-
ations were not detected before the pressure dropped, but after lO-min perfusion

Fig. 6. Left ventricular developed pressure in isolated, perfused rat hearts was measured with a latex
balloon connected to a pressure transducer. This tracing illustrates the pressure drop and the subse-
quent contracture during perfusion with enzymatically generated oxygen radicals
Oxygen Radicals in Heart and Brain Tissue Injury 249

Fig. 7. This micrograph illustrates the ultrastructural findings in the myocardium after lO-min perfu-
sion with enzymatically generated oxygen radicals and 35-min recovery with ordinary buffer perfu-
sion . Mitochondria appear swollen and contain granula, the myofilaments are disruptured, there are
empty spaces between the myofilaments (edema), and the cell membrane is partly missing

with oxygen radicals there was a prominent alteration in ultrastructure. As demon-

strated in Fig. 7, intracellular edema, mitochondrial alteration, and sarcolemmal
damage were common findings after perfusion with oxygen radicals [72]. Disruption
of myofilaments and contraction bands were also detected in some micrographs. In
order to prove that these effects were due to oxygen radicals, the specific enzymes
SOD and catalase were added to the buffer in separate experiments. In these experi-
ments the addition of xanthine oxidase and hypoxanthine failed to produce a de-
crease in contractility or ultrastructural damage (Fig. 8). Our experiments support
the hypothesis that oxygen radicals are toxic and able to produce both structural and
functional alterations in myocardial tissue.
Oxygen radicals are known to inhibit mitochondrial function [14, 25 , 27, 32] , to
reduce the activity of Na/K-ATPase [37] and Ca 2+-ATPase [28, 31, 51], and also to
reduce the ability of mitochondria to retain calcium [27, 48]. These findings could be
in agreement with the possibility that damage produced during reperfusion may be
partly due to the development of oxygen radicals in the heart, brain, and other organs
[5, 21, 46]. However, knowledge of the site of oxygen radical production as well as
direct evidence of their production under such circumstances are still lacking.
250 K. Ytrehus and O . D. Mjj1js

Fig. 8. Micrograph from hearts receiving the protective enzymes superoxide dismutase and catalase
together with oxygen radicals. No ultrastructural alterations can be seen

Possible Sources of Oxygen Radicals in Ischemic Tissue Damage

Special interest has been focused on the demonstration of the presence of xanthine
dehydrogenase in various tissues, mainly in the endothelium [33, 58]. This enzyme is
shown to convert into an oxygen radical-producing xanthine oxidase during ischemia
[9, 50]. Due to breakdown of A TP during ischemia there are increased levels of the
purine-degradation products xanthine and hypoxanthine. During reperfusion xan-
thine oxidase is able to produce superoxide and hydrogen peroxide due to the supply
of oxygen, the other substrate for the enzyme (Fig. 3).
Some investigators have demonstrated a decrease in cell defence against oxygen
radicals after hypoxia or ischemia, and this will render the cell more vulnerable to
oxygen radicals from mitochondrial, nuclear, and microsomal electron transfer [24,
35] . In the intact organism migration of activated white blood cells during reperfu-
sion into the previous ischemic area may damage viable tissue by producing oxygen
radicals [40].
Oxygen Radicals in Heart and Brain Tissue Injury 251

Summary and Conclusions

In our studies we demonstrated that oxygen radicals produced myocardial injury,

and according to other studies oxygen radicals could be a reason for exacerbation of
injury after resolution of a transient ischemic event. The clinical implication of this
may be the development of therapy with oxygen radical scavengers and antioxidants.
This may prove a useful addition to current techniques of cardioplegia, coronary
trombolysis, and coronary angioplasty as well as a useful tool in renal transplanta-
tion, vascular surgery, surgery after major trauma, and cardiopulmonary resuscita-
tion.

References

1. Allen RC, Stjernholm RL, Steele RH (1972) Evidence for the generation of an electronic excita-
tion state(s) in human polymorphonuclear leucocytes and its participation in bactericidal activ-
ity. Biochem Biophys Res Commun 47:679-684
2. Ambrosio G, Weisfeldt ML, Jacobus WE, Flaherty JT (1987) Evidence for a reversible oxygen
radical-mediated component of reperfusion injury: reduction by recombinant human superoxide
dismutase administration at the time of reflow. Circulation 75: 282-291
3. Aust SD, Morehouse LA, Thomas CE (1985) Role of metals in oxygen radical reactions. Free
Radical Bioi Med 1: 3-25
4. Babior BM, Kipnes RS, Curnutte IT (1973) Biological defence mechanisms. The production by
leucocytes of superoxide, a potential bacterial agent. J Clin Invest 52: 741-744
5. Baker GL, Corry RJ, Autor AP (1985) Oxygen free radical induced damage in kidneys sub-
jected to warm ischemia and reperfusion. Ann Surg 202 : 628-641
6. Bolli R, Zhu WX, Hartley CJ, Michael LH, Repine JE, Hess ML, Kukreja RC, Roberts R
(1987) Attenuation of dysfunction in the postischemic "stunned" myocardium by dimethyl-
thiourea. Circulation 76: 458-468
7. Brawn K, Fridovich I (1980) Superoxide radical and superoxide dismutase: threat and defence.
Acta Physiol Scand [Suppl] 492:9-18
8. Cerchiari EL, Hoel TM, Safar P, Sclabassi RJ (1987) Protective effects of combined superoxide
dismutase and deferoxamine on recovery of cerebral blood flow and function after cardiac arrest
in dogs. Stroke 18: 869-878
9. Chambers DE, Parks DA, Patterson G, Roy R, McCord JM, Yoshida S, Parmley LF, Downey
JM (1985) Xanthine oxidase as a source of free radical damage in myocardial ischemia. J Mol
Cell Cardiol 5 : 395-407
10. Chance B, Sies YA, Boveris A (1979) Hydroperoxide metabolism in mammalian organs. Physiol
Rev 59:527-605
11. DelMaestro RF, Thaw HR, Bjfllrk J, Planker M, Arfors K-E (1980) Free radicals as mediators
of tissue injury. Acta Physiol Scand [Suppl] 492 : 43-57
12. DelMaestro RF, Bjfllrk J, Arfors K-E (1981) Increase in microvascular permeability induced by
enzymatically generated free radicals. Microvasc Res 22: 239-254
13. Demophoulos HB, Flam ES, Seligman ML, Jorgensen E, Ransohoff J (1977) Antioxidant ef-
fects of barbiturates in model membranes undergoing free radical damage. In: Ingvar DH, Las-
sen NA (eds) Cerebral function, metabolism and circulation. Acta Neurol Scand [Suppl64] 56:
152-154
14. Flamigni F, Guarneri C, Toni R, Caldarera CM (1982) Effect of oxygen radicals on heart mito-
chondrial function in a-tocopherol deficient rabbits. Int J Vitam Nutr Res 52: 402-406
15. Forsman M, Fleischer JE, Milde JH, Steen PA, Michenfelder JD (1988) Superoxide dismutase
and catalase failed to improve neurologic outcome after complete cerebral ischemia in the dog.
Acta Anaesthesiol Scand 32: 152-155
252 K.Ytrehus and O.D.Mjj1ls

16. Freeman BA, Crapo JD (1982) Free radicals and tissue injury. J Lab Invest 47: 412-426
17. Fridovich I (1970) Quantitative aspects of the production of superoxide anion radical by milk
xanthine oxidase. J Bioi Chern 245 : 4053-4057
18. Fridovich I (1983) Superoxide radical: an endogenous toxicant. Annu Rev Pharmacol Toxicol
23:239-257
19. Ganote CE, Sims M, Safavi S (1982) Effects of dimethylsulfoxide (DMSO) on the oxygen
paradox in perfused rat hearts. Am J Pathol109: 207-276
20. Gaudel Y, Duvelleroy MA (1984) Role of oxygen radicals in cardiac injury due to reoxygena-
tion. J Mol Cell Cardiol 16: 459-470
21. Granger DN, Rutili G, McCord JM (1981) Role of superoxide radicals in intestinal ischemia.
Gastroenterology 81: 22-29
22. Green MJ, Hill HAO (1984) Chemistry of oxygen. Methods Enzymol105: 3-22
23. Guarnieri C, Ferrari R, Visioli 0, Caldarera CM, Nayler WG (1978) Effect of a-tocopherol on
hypoxic-perfused and reoxygenated rabbit heart muscle. J Mol Cell Cardiol 10: 893-906
24. Guarnieri C, Flamigni F, Caldarera CM (1980) Role of oxygen in the cellular damage induced
by re-oxygenation of hypoxic heart. J Mol Cell Cardiol12: 797-808
25. Guarnieri C, Muscari C, Ceconi C, Flamigni F, Caldarera CM (1983) Effects of superoxide gen-
eration on rat heart mitochondrial pyruvate utilization. J Mol Cell Cardiol15: 859-862
26. Halliwell B, Gutteridge JMC (1986) Oxygen free radicals and iron in relation to biology and
medicine: some problems and concepts. Arch Biochem Biophys 246:501-514
27. Harris EJ, Booth R, Cooper MB (1982) The effect of superoxide generation on the ability of
mitochondria to take up and retain Ca2+. FEBS Lett 146:267-272
28. Harris RN, Doroshow JH (1985) Effect of doxorubicin-enhanced hydrogen peroxide and hydroxyl
radical formation on calcium sequestration by cardiac sarcoplasmic reticulum. Biochem Biophys
Res Commun 130:739-745
29. Hearse DJ, Humphrey SM, Chain EB (1973) Abrupt reoxygenation of the anoxic potassium-
arrested perfused rat heart: a study of myocardial enzyme release. J Mol Cell Cardiol5: 395-407
30. Hertz F, Cloarec A (1984) Pharmacology of free radicals; recent views on their relation to in-
flammatory mechanisms. Life Sci 34: 713-720
31. Hess ML, Okabe E, Ash P, Kontos HA (1984) Free radical mediation of the effects of acidosis
on calcium transport by cardiac sarcoplasmic reticulum in whole heart homogenates. Cardiovasc
Res 18: 149-157
32. Hillered L, Ernster L (1983) Respiratory activity of isolated rat brain mitochondria following in
vitro exposure to oxygen radicals. J Cereb Blood Flow Metab 3: 207-214
33. Jarasch E-D, Ground C, Bruder G, Heid HW, Keenan TW, Franke WW (1981) Localization of
xanthine oxidase in mammary-gland epithelium and capillary endothelium. Cell 25: 67 -82
34. Jolly SR, Kane WJ, Bailie MB, Abrams GD, Lucchesi BR (1984) Canine myocardial reperfu-
sion injury. Its reduction by the combined administration of superoxide dismutase and catalase.
Circ Res 54:277-285
35. Julicher RHM, Tijburg LBM, Sterrenberg L, Bast A, Koomen JM, Noordhoek J (1984) De-
creased defence against free radicals in rat hearts during normal reperfusion after hypoxic, isch-
emic and calcium-free perfusion. Life Sci 35: 1281-1288
36. Kogure K, Arai H, Abe K, Nakano M (1985) Free radical damage of the brain following ischaemia.
Prog Brain Res 63: 234-259
37. Kramer HJ, Tong Mak I, Weglicki WB (1984) Differential sensitivity of canine cardiac sarco-
lemmal and microsomal enzymes to inhibition by free radical-induced lipid peroxidation. Circ
Res 55 : 120-124
38. Langendorff 0 (1895) Untersuchungen am tiberlebenden Saugthierherzen. Pfluegers Arch
Physiol61: 291-332
39. Lim KH, Connolly M, Rose D, Siegman F, Jacobowitz I, Acinapura A, Cunningham IN (1986)
Prevention of reperfusion injury of the ischemic spinal cord: use of recombinant superoxide dis-
mutase. Ann Thorac Surg 42: 282-286
40. Lucchesi BR, Romson JL, Jolly SR (1984) Do leucocytes influence infarct size? In: Hearse DJ,
Yellon DM (eds) Therapeutic approaches to myocardial infarct size limitation. Raven, New
York, pp 219-248
41. McCord JM, Fridovich I (1969) Superoxide dismutase. An enzymatic function for erythro-
cuprein (Hemocuprein). J Bioi Chern 244: 6049-6055
Oxygen Radicals in Heart and Brain Tissue Injury 253

42. Mimnaugh EG, Trush MA, Ginsburg E, Gram TE (1982) Differential effects of antracycline
drugs on rat heart and liver microsomal reduced nicotinamide adenine dinucleotide phosphate-
dependent lipid peroxidation. Cancer Res 42: 3574-3582
43. Mitsos SE, Fantone JC, Gallagher KP, Walden KM, Simpson PJ, Abrams GD, Schork MA,
Lucchesi BR (1986) Canine myocardial reperfusion injury: protection by a free radical scavenger,
N-2-mercaptoproprionyl glycine. J Cardiovasc Pharmacol 8: 978-989
44. Myers ML, Bolli R, Lekich R, Hartley CJ, Roberts R (1985) Enhancement of recovery of myo-
cardial function by oxygen free-radical scavengers after reversible regional ischemia. Circulation
72:915-921
45. Oster 0, Prellwitz W, Kasper W, Meinertz T (1983) Congestive cardiomyopathy and the selenium
content of serum. Clin Chim Acta 128: 125-132
46. Paller MS, H!1lidal JR, Ferris RF (1984) Oxygen free radicals in ischemic acute renal failure in
the rat. J Clin Invest 74: 1156-1164
47. Proctor PH, Reynolds ES (1984) Free radicals and disease in man. Physiol Chern Phys 16: 175-195
48. Richter C (1984) Hydroperoxide effects on redox state of pyridine nucleotides and Ca2+ reten-
tion by mitochondria. Methods Enzymol 105: 435-441
49. Rotruck JT, Pope AL, Ganther HE, Swanson AB, Hafeman DG, Hoekstra WG (1973) Selenium:
biochemical role as component of glutathione peroxidase. Science 179: 588-590
50. Roy RS, McCord 1M (1983) Superoxide and ischemia: conversion of xanthine dehydrogenase to
xanthine oxidase. In: Greenwald RA, Cohen G (eds) Oxy radicals and their scavenger systems.
Elsevier Biomedical, Amsterdam, pp 145-153
51. Scherer NM, Deamer DW (1986) Oxidative stress impairs the function of sarcoplasmic reticulum
by oxidation of sulfhydryl groups in the Ca2+-ATPase. Arch Biochem Biophys 246: 589-601
52. Shlafer M, Kane PF, Wiggins BS, Kirsh MM (1982) Possible role for cytotoxic oxygen metabo-
lites in the pathogenesis of cardiac ischemic injury. Circulation [Suppl 1] 66: 185-192
53. Siesjo BK, Rehncrona S, Smith D (1980) Neuronal cell damage in the brain: possible involve-
ment of oxidative mechanisms. Acta Physiol Scand [Suppl] 492 : 121-128
54. Slater TF (1984) Free-radical mechanisms in tissue injury. Biochem J 222: 1-15
55. Small A (1984) New perspectives on hyperoxic pulmonary toxicity - a review. Undersea Bio-
med Res 11: 1-4
56. Smith DS, Rehncrona S, Siesjo BK (1980) Barbiturates as protective agents in brain ischemia
and as free radical scavengers in vitro. Acta Physiol Scand [Suppl] 492: 129-134
57. Stewart JR, Blackwell WH, Crute SL, Loughlin V, Greenfield LJ, Hess ML (1983) Inhibition
of surgically induced ischemia/-reperfusion injury by oxygen free radical scavengers. J Thorac
Cardiovasc Surg 86:262-272
58. Stirpe F, Della Corte E (1969) The regulation of rat liver xanthine oxidase. J Bioi Chern 244:
3855-3863
59. Sunnergren KP, Rovetto MJ (1987) Myocyte and endothelial injury with ischemia reperfusion
in isolated rat hearts. Am J Physiol 252: HI211-HI218
60. Tappel AL (1972) Vitamin E and free radical peroxidation of lipids. Ann NY Acad Sci 203:
19-28
61. Tomita I, Sano M, Serizawa S, Ohta K, Katou M (1979) F1uctuation of lipid peroxides and re-
lated enzyme activities at time of stroke in stroke-prone spontaneously hypertensive rats. Stroke
10:323-326
62. Triggs WJ, Willmore LJ (1984) In vivo lipid peroxidation in rat brain following intracortical Fe2+
injection. J Neurochem 42: 976-980
63. Turrens JF, Boveris A (1980) Generation of superoxide anion by the NADH dehydrogenase of
bovine heart mitochondria. Biochem J 191: 421-427
64. Uraizee A, Reimer KA, Murry CE, Jennings RB (1987) Failure of superoxide dismutase to limit
size of myocardial infarction after 40 min of ischemia and 4 days of reperfusion in dogs. Circula-
tion 75: 1237-1248
65. Wei EP, Kontos HA, Christman CW, DeWitt DS, PovlishockJT (1985) Superoxide generation
and reversal of acetylcholine-induced cerebral arteriolar dilatation after acute hypertension.
Circ Res 57: 781-787
66. Weiss SJ (1986) Oxygen, ischemia and inflammation. Acta Physiol Scand [Suppl] 548:9-34
67. Weiss SJ, LoBuglio AF (1982) Phagocyte-generated oxygen metabolites and cellularinjury. Lab
Invest 47 : 5-18
254 K.Ytrehus and O.D.Mjl/ls: Oxygen Radicals in Heart and Brain Tissue Injury

68. Wems SW, Shea MJ, Driscoll EM, Cohen C, Abrams GD, Pitt B, Lucchesi BR (1985) The
independent effects of oxygen radical scavengers on canine infarct size. Circ Res 56: 895-898
69. Yoshida S, Busto R, Santiso M, Ginsberg MD (1984) Brain lipid peroxidation induced by post-
ischemic reoxygenation in vitro: effects of vitamin E. J Cereb Blood Flow Metab 4: 466-469
70. Ytrehus K, Myklebust R, Mjl/ls OD (1986) Influence of oxygen radicals generated by xanthine
oxidase in the isolated perfused rat heart. Cardiovasc Res 20:597-603
71. Ytrehus K, Gunnes S, Myklebust R, Mjl/ls OD (1987) Protection by superoxide dismutase
(SOD) and catalase in the isolated rat heart reperfused after prolonged cardioplegia: a com-
bined study of metabolic, functional and morphometric ultrastructural variables. Cardiovasc
Res 21 : 492-499
72. Ytrehus K, Myklebust R, Olsen R, Mjl/ls OD (1987) Ultrastructural changes induced in the iso-
lated rat heart by enzymatically generated oxygen radicals. J Mol Cell Cardiol19: 379-389
73. Ytrehus K, Ringstad J, Myklebust R, Norheim G, Mjl/ls OD (1988) The selenium-deficient rat
heart with special reference to tolerance against enzymatically generated oxygen radicals. Scand
J Clin Lab Invest 48: 289-295
Chapter 20
Prevention of Ischemic Brain Damage
Following Cardiac Arrest
L.MOGENSEN

Introduction

Ischemic brain damage is the most feared complication following successful treat-
ment of cardiac arrest. This paper reviews preventive and protective measures.
The single most important measure reducing or eliminating brain damage is the
earliest possible restoration of circulation. An important adjunct is effective cardio-
pulmonary resuscitation (CPR). Effective CPR was introduced about 25 years ago,
and since then many individuals have been taught this technique. At present only
CPR has the potential to postpone brain damage due to cardiac arrest. Early im-
plementation of CPR depends on the capability of cardiac arrest witnesses. In most
cases ventricular fibrillation is the underlying pathophysiological event and can usually
be converted with electrical countershocks using defibrillators. Widespread know-
ledge of CPR throughout society and appropriate allocation of defibrillators are ex-
pressions of foresight and motivation of key individuals in the society. An increased
number of patients will survive cardiac arrest, some of them with latent or overt
brain damage.
Brain damage following global ischemia is mainly determined by the ischemic im-
pact. It develops further during early reperfusion, which suggests the possibility of
some success from interventions aimed at protecting the brain during this phase. This
concept of preventing further brain damage has been analyzed in detail by Safar [19,
20], who has coined the term cardiopulmonary cerebral resuscitation. Extensive ex-
perimental studies have revealed a number of complex interacting pathophysiologi-
cal mechanisms relevant to brain damage following temporary global ischemia. Some
brain protective measures will be reviewed in the following, with emphasis on clinical
applicability. For more detailed analyses of these problems, the reader is referred to
extensive reviews [4, 9, 16, 19,20,26].

Basal Cardiopulmonary Resuscitation

The concept of basal CPR (i.e., without use of equipment) is probably a very old one
[11]. With the introduction of newer techniques, including mouth-to-mouth ventila-
tion and closed-chest cardiac massage, CPR has emerged as a clinically indispensable
form of treatment. The pathophysiology during CPR has been studied in detail [18,
19]. The results indicate that the heart acts only as a conduit during closed-chest car-
diac massage, and the thoracic cavity as the pump, both for air and blood. Retro-
256 L.Mogensen

grade blood flow during chest compression is prevented through functional valves,
mainly in the veins of the thoracic outlet. Thus, with only the mouth, two hands, and
the necessary skill, an individual may establish brain perfusion in a cardiac arrest vic-
tim and possibly save a life. Furthermore, effective CPR can be performed by most
laymen after appropriate training. Training programs have been implemented world-
wide, principally modeled after those of the American Heart Association [2]. Thus,
CPR will be attempted in more and more cases, and complications will be more com-
mon, latent or overt [8, 10, 17].
A number of studies have demonstrated the clinical value of CPR in cardiac ar-
rest of various etiologies [20]. In almost all cases CPR will be ineffective if not com-
plemented with more fundamental attempts to restore circulation, mainly by defibril-
lation, both within and outside hospitals [5, 19,21]. The prognosis is related both to
the time delay until basal CPR is initiated. and to the delay until effective circulation
is restored. Additional actions during CPR and in the early recovery phase include
physical, physiological, metabolic, or pharmacological interventions. The choice of
interventions should be related to the underlying pathological mechanisms [6-9].

Physical Interventions

A primary and easily available physical means to protect the brain is the supplemen-
tation of extra oxygen to the ventilated air. An important adjunct may be to provide
calm and quiet surroundings to decrease sensory brain stimulation.
More complicated is the induction of hypothermia. General hypothermia has
been employed for decades to increase the CNS tolerance during the critical ischemic
period in cerebral and cardiac surgery [12]. A temperature reduction of IOC will de-
crease the metabolic rate by about 7%. General hypothermia is considered the major
factor explaining why even prolonged cardiac arrest in association with near-drown-
ing in cold water does not necessarily preclude a successful outcome. During hypo-
thermia, however, a wide variety of potentially harmful effects may occur [12, 19].
They include increased blood viscosity, leftward shift of the hemoglobin oxygen dis-
sociation curve, hematologic changes including increased red blood cell stiffness,
cerebral dysfunction including apathy, drowsiness, and slow mentation as well as
signs of dysfunction in extracerebral organs. Hypothermia has been employed clini-
cally during and after CPR of cardiac arrest but with less promising results [3]. One
reason for this is the difficulty in cooling the body quickly enough to achieve suffi-
ciently advanced brain hypothermia before the onset of irreversible damage.
Hypothesizing that selective brain cooling may have potential value, we have per-
formed theoretical analyses of the rate at which the brain temperature could be re-
duced selectively, under various conditions [15]. Heat transfer was calculated using
computer simulation. We simulated effects from the use of iced water on the skull
surface, and on the skin over the two carotid arteries between the clavicles and the
skull base.
Direct skull cooling generates a cold front which penetrates into the brain, at de-
creasing speed with increasing distance. The resulting cooling of the brain starts in
Prevention of Ischemic Brain Damage Following Cardiac Arrest 257

the cortex and is calculated to reduce the temperature of the outer layer about 5°_
10°C after 10-60 min.
The cooling of the carotid arterial blood would yield an additional and more
evenly distributed cooling effect of most of the brain, vertebral territories exempt,
which initially amounts to 0.1°-0.4°C/min.
This cooling rate is probably too slow, and these measures have presumably
limited potential value for cerebral protection during the initial critical phase follow-
ing cardiac arrest. Further cooling effects from using ice-cold air for ventilation have
been calculated to be negligible. Invasive techniques such as infusion of cold and
perhaps "cerebroplegic" blood have not been evaluated.

Physiological Interventions

Physiological means to combat brain damage include early restoration of brain per-
fusion, with adequately oxygenated blood. Critical blood flow has been studied ex-
tensively [23]. Blood flow patterns differ after hypoxic, asphyxic, and ischemic brain
damage, and drug effects are complex. Time factors are important, as is the type of
ischemic impact, global or focal. During reperfusion a "non-reflow phenomenon"
has been found also for the brain, as in other organs. It may result from platelet
adhesion and aggregation, clotting, vasospasm, perivascular tissue swelling as well as
through other mechanisms. It is suggested that this can be counteracted by the use of
antiaggregatory infusions such as dextran or mannitol, usually best at normal arterial
blood pressures [7, 19].
The potential value of perfusion at moderately elevated blood pressures has been
much debated and may be useful under certain conditions [7, 19, 23]. The auto-
regulatory mechanisms of the postischemic brain are complex and vary with time.
The additional value of va so dilating drugs, notably "brain-selective" slow calcium
channel blockers, remains to be verified. More complete reviews on this subject are
available [6, 14].
Physiological measures to reduce brain damage also involve attempts to restore
extra- and intracellular acid-base balance. During the first minutes following cardiac
arrest, arterial blood gases may be normal, as the metabolic rate of whole blood is
slow. In contrast, tissue acidosis develops quickly and is both respiratory and meta-
bolic in origin. With CPR and resulting tissue perfusion the venous blood quickly
shows markedly decreased P0 2 and pH, and elevated PC0 2 , reflecting the tissue
acid-base deviations [24]. Arterial blood gases will show these changes after con-
siderable delay, due both to the slow blood flow and the reduced efficacy of alveolar
ventilation. Hyperventilation rather than the use of buffering infusions is recom-
mended during the initial 5-10 min of CPR.

Metabolic Interventions

The metabolism of the brain mainly utilizes glucose and oxygen, reflected in a respi-
ratory quotient close to 1 [22]. Energy stores are negligible and severe dysfunction
258 L.Mogensen

occurs early when the supply is stopped. Unconsciousness occurs within 15 s and flat-
tened EEG after about 30 s.
Metabolic means to preserve the integrity of the brain after cardiac arrest have
initially been focused on the problems of acidosis and nutrients. The use of buffering
solutions, in addition to hyperventilation, is now recommended only for prolonged
periods of CPR [2]. Sodium bicarbonate has long been the drug of choice, even ifre-
cently questioned [2, 19,24]. The theoretical advantages of Tris buffer, a combina-
tion of trometamol, bicarbonate, and acetate, have recently been emphasized [25].
Its buffering capacity is somewhat enhanced. The decreased production of CO 2 ,
which quickly enters the cells, will result in less intracellular acidosis due to forma-
tion of carbonic acid. In addition, Tris buffer reduces the sodium load. In many
countries this solution is available on the market as Tribonate.
It has been demonstrated that hyperglycemia at the initiation of cardiac arrest
promotes intracellular acidosis, as compared with normo- and especially hypo-
glycemia [22]. This concept has not been clinically exploited, even if it should be
noted that patients when prepared for surgery under generalized hypothermia, such
as in neurosurgery and cardiac surgery, by tradition are kept in a starved condition.
This observation may also be relevant to the final outcome after cardiac arrest in pa-
tients with less than optimally regulated diabetes mellitus, a concept which has not
yet been evaluated clinically.

Pharmacological Interventions

Pharmacological means to reduce brain damage have been studied in a large number
of experimental situations and in a few clinical trials. A thoughtful and sceptical re-
view has recently been published [6], balancing the optimism expressed by others
[19,20,26]. Some drugs studied are listed in Table 1, also indicating the major theo-
retical consideration supporting their potential use [6].
Drugs which are anticonvulsant and depress cerebral metabolism include barbitu-
rates and diphenylhydantoin. Barbiturates in high doses have been shown to depress
cerebral electrical activity and metabolism, and thus oxygen demand. The efficacy
has been convincingly demonstrated in animal experiments, prolonging neuronal
survival in ischemia. In one large clinical trial, the results were disappointing, and
barbiturates cannot be generally recommended [1]. It should be mentioned that the
use of barbiturates at this dosage level results in major hemodynamic deterioration,
which has to be counteracted. Diphenylhydantoin has experimental support but has
not been clinically evaluated.
The use of slow calcium channel blockers is theoretically attractive as they may
both affect vascular tone and combat toxic cellular effects of excessive calcium inflow
[14]. Promising laboratory studies have been performed, and a major clinical trial is
underway. The most promising agents seem to be nimodipine and lidoflazine [6, 14,
19].
Drugs which improve microcirculation include vasopressors, heparin, and dextran,
all of which have been employed in animal experiments with promising results [6].
No clinical studies have been undertaken. Prostaglandins and drugs affecting prosta-
glandin synthesis include prostacyclin, indomethacin, and acetylsalicylic acid. Again,
Prevention of Ischemic Brain Damage Following Cardiac Arrest 259

Table 1. Some drugs employed in attempts to protect the

brain after ischemia

Anticonvulsant-depressant drugs
Barbiturates
Diphenylhydantoin
Calcium channel blockers
Nimodipine
Lidoflazine
Microcirculation-promoting drugs
Heparin
Dextran
Vasopressors
Prostaglandin-related drugs
Indomethacin
Prostacyclin
Acetylsalicylic acid
Free radical scavengers
Miscellaneous
Cortisone
Naloxone

they have not been studied in controlled clinical trials of patients after cardiac arrest.
The use of free radical scavengers has strong theoretical support [13] and attracts
considerable attention; some substances are presently being employed in clinical
trials. The use of large doses of cortisone was previously common, but has now been
abandoned in routine cases, as has the use of naloxone.
The use of pharmacological agents to ameliorate brain damage has theoretical
support and much clinical attraction. So far few controlled clinical trials have been
performed. It should be recognized, however, that clinical studies of patients with
cardiac arrest encounter major difficulties, ranging from complex ethical aspects to
intricate logistic problems, i.e., in obtaining well-defined, comparable, and homo-
geneous patient groups. A number of different agents have been used tentatively.
No drug has yet emerged as an effective agent. General recommendations regarding
the use of any brain-protective drug by the clinician facing patients after cardiac
arrest remain to be established.

Summary and Conclusions

Cardiac arrest is a common event, and sudden death a likely outcome. In industrial-
ized societies, about one-fifth of all deaths are sudden and half of them witnessed
outside hospitals. However, cardiac arrest is increasingly treated by cardiopulmonary
resuscitation (CPR) saving the lives of thousands of victims annually. Some of them
unfortunately develop brain damage, either latent or overt. Serious brain damage
260 L. Mogensen: Prevention of Ischemic Brain Damage Following Cardiac Arrest

following CPR is a feared complication. A number of presumably protective mea-

sures may be employed, but the clinical value remains to be established. The early
and effective initiation of CPR followed as soon as possible by restoration of circula-
tion are the most important measures to protect the brain following cardiac arrest.

References
1. Abramson NS for the Brain Resuscitation Clinical Trial I Study Group (1986) Randomized clini-
cal study of thiopental loading in comatose survivors of cardiac arrest. N Engll Med 314: 397-403
2. American Heart Association (1986) Standards and guidelines for cardiopulmonary resuscitation
(CPR) and emergency cardiac care (ECC). lAMA 255:2841-3044
3. Baba K, Hirakawa M, Goto I, Yokota T (1971) Use of immediate hypothermia for the preven-
tion of cerebral damage following resuscitation in cardiac arrest. Masui 20: 252-258
4. Bircher NG (1985) Ischemic brain protection. Ann Emerg Med 14:784-788
5. Eisenberg MS, Copass MK, Hallstrom AP, Blake B, Bergner L, Short FA, Cobb LA (1980)
Treatment of out-of-hospital cardiac arrests with rapid defibrillation by emergency medical tech-
nicians. N Eng! 1 Med 302: 1379-1383
6. Gisvold SE, Steen PA (1985) Drug therapy in brain ischemia. Br 1 Anaesthesiol 57: 96-109
7. Graham DI (1985) The pathology of brain ischemia and possibilities for therapeutic interven-
tion. Br 1 Anaesthesiol 57: 3-17
8. lennett B (1976) Irrecoverable brain damage after resuscitation: brain death and other syn-
dromes. Resuscitation 5: 49-52
9. Kirsch lR, Dean 1M, Rogers MC (1986) Current concepts in brain resuscitation. Arch Intern
Med 146: 1413-1419
10. Krischer JP, Fine EG, Davis JR, Nagel EL (1987) Complications of cardiac resuscitation. Chest
92:287-291
11. Liss HP (1986) A history of resuscitation. Ann Emerg Med 15: 65-72
12. Maspes PE, Hughes B (eds) (1964) Hypothermia in neurosurgery. Symposium. Springer, Wien
New York
13. McCord 1M (1985) Oxygen-derived free radicals in postischemic tissue injury. N Engl 1 Med
312:159-164
14. Meldrum B, Evans M, Griffiths T, Simon R (1985) Ischaemic brain damage: The role of excit-
atory activity and of calcium entry. Br J Anaesthesiol 57: 44-46
15. Mogensen L, Mogensen E, Mogensen P (to be published) Transcutaneous brain cooling - theo-
retical considerations evaluated through computer simulation
16. Newberg LA (1984) Cerebral resuscitation: advances and controversies. Ann Emerg Med 13:
853-856
17. Nielsen lR, Gram L, Rasmussen LP, Damsgaard EM, Dalsgaard M, Richardt C, Beckmann J
(1983) Intellectual and social function of patients surviving cardiac arrest outside the hospital.
Acta Med Scand 213: 37-39
18. Rudikoff MT, Maughan WL, Effron M, Freund P, Weisfeldt ML (1980) Mechanisms of blood
flow during cardiopulmonary resuscitation. Circulation 61: 345-352
19. Safar P (1986) Cerebral resuscitation after cardiac arrest: a review. Circulation [Suppl 4] 74:
138-153
20. Safar P, Bircher N (1988) Cardiopulmonary cerebral resuscitation. An introduction to resuscita-
tion medicine, 3rd edn. WB Saunders, London
21. Sandoe E (1984) Organization of resuscitation within the hospital. Am 1 Emerg Med 2: 270-272
22. Siesjo BK, Wieloch T (1985) Cerebral metabolism in ischemia: neurochemical basis for therapy.
Br 1 AnaesthesioI57:47-62
23. Symon L (1985) Flow threshold in brain ischemia and the effects of drugs. Br 1 Anaesthesiol57:
34-43
24. Weil MR, Rachow EC, Trevino R, Gruntler W, Falk lL, Griffel MI (1986) Difference in acid-
base state between venous and arterial blood during cardiopulmonary resuscitation. N Engl 1
Med 315: 153-156
25. Wiklund L, 6qvist L, Skoog G, Tyden H, 10rfeldt L (1985) Clinical buffering of metabolic
acidosis: Problems and a solution. Resuscitation 12: 279-293
26. Yatsu FM (1986) Cardiopulmonary cerebral resuscitation. N Engl 1 Med 314:440-441
Part V
Sleep and Sleep Apnea
Chapter 21

Influence of Sleep, Activity and Circadian Rhythm

on Heart Rate, QT Interval and Cardiac Arrhythmias
C. GUILLEMINAULT and A. M. GILLIS

Sleep has a significant impact on cardiovascular physiology. A decline in heart rate

occurs, which in the cat is more pronounced during rapid-eye movement (REM)
sleep than during nonrapid-eye movement (NREM) sleep. This decline parallels
blood pressure reduction. REM sleep itself is characterized as tonic or phasic and
phasic REM sleep is associated with bursts of rapid-eye movements in humans. In
cats, these rapid-eye movements are associated with electro physiological events seen
simultaneously in the pons, in the lateral geniculate (bundle), and in the occipital
lobe. These events are called ponto-geniculo-occipital (PGO) waves. In both
species, bursts of REM are accompanied by brief episodes of tachycardia generally
followed by a period of bradycardia.

The Autonomic Nervous System and Heart Rate During Sleep

The parasympathetic nervous system is largely responsible for modulating the heart
rate during sleep in adults. However, withdrawal of sympathetic influences also ap-
pears to be of importance as a reduction in heart rate during sleep, as seen in vagoto-
mized cats. A combined vagotomy and bilateral stellectomy are necessary to com-
pletely eliminate the tonic and phasic REM heart-rate changes during sleep.
In adult cats, the heart-rate changes are progressively more pronounced from
sleep onset of NREM sleep and tonic REM sleep, due to a progressive predomi-
nance of vagal tone with a decrease of sympathetic activity. During phasic REM
sleep, several changes occur in rapid succession: a phasic inhibition of parasympa-
thetic discharge with abrupt return of sympathetic tone, followed by phasic reduction
in sympathetic activity with an increase in vagal discharge. In normal human adults,
the most commonly observed change in heart rhythm during sleep is a reduction of
heart rate. This heart rate decrease is seen even in short-term experiments in adults
when the time of sleep is artificially displaced to separate the influence of sleep and
sleep states from the influence of circadian rhythm. It is important to realize that
there are several confounding elements in the heart rate decrease seen in an indi-
vidual with a normal day/night-wake/sleep cycle. There is a circadian, i.e., 24-h,
heart-rate rhythm that in adults closely follows the temperature rhythm, with a peak
in early afternoon and a trough near 2-3 a.m. In a normal individual, the activity/in-
activity cycle linked to the sleep/wake cycle also plays a role.
264 C. Guilleminault and A. M. Gillis

Experiment 1

In an investigation of 15 male heart-transplant recipients carried out during noctur-·

nal sleep with polygraphic monitoring and simultaneous Holter EeG recording, we
were unable to detect any changes in RR intervals associated with phasic events of
REM sleep despite the presence of bursts of REM. We conducted a three-night poly-
graphic monitoring offive type-l insulin-dependent diabetics, aged 27-33 years, who
had significant autonomic dysfunction. Again, although 43 REM periods were ob-
served, no RR interval changes associated with phasic events of REM sleep occur-
red. Despite surgical or pathological interruption of autonomic nerves, however, we
have noted the persistence of a day/wake-night/sleep heart rate rhythm with a peak
in the afternoon and a trough in the middle of the night in these ambulatory subjects.
We performed a further experiment on heart-transplant recipients [4]. Six pa-
tients were kept in a supine position for 20h with a normal day/night-wake/sleep
cycle. There was a severe blunting of the circadian heart rate rhythm for the group
as a whole. Two of the six patients showed no significant day/night heart rate differ-
ences when data for each individual were submitted to mathematical testing, while
the four others had a very significant blunting of their curve. The following formula
was used: y(t) = M + A cos (omega + theta), where M stands for mesor, A for am-
plitude, theta for acrophase, t for time, and omega for angular frequency, with a
least-squares digital computer curve fitting program. It thus appears that the activity/
inactivity cycle plays a major role in maintaining the circadian rhythm of the heart
rate in individuals with lesioned autonomic nervous system (ANS) control of the
heart.

QT Interval During Wake and Sleep

Sympathetic and vagal stimuli can profoundly influence ventricular arrhythmias. Be-
cause of this, changes in autonomic tone associated with sleep, and more specifically
during phasic events of REM sleep, have focused attention on the possible influence
of sleep on ventricular arrhythmias and sudden cardiac death. Because prolongation
of the QT interval has been associated with malignant ventricular arrhythmias, we
studied patients with ventricular arrhythmias to determine if the QT interval was
prolonged during sleep and to assess the importance of activity and sleep states as
variables influencing the diurnal variation of the QT interval.

Experiment 2

We studied nine patients, six men and three women, mean age 50 years, who had a
history of frequent ventricular ectopy [1]. None had evidence of breathing problems
during sleep; three had no documented heart disease; two had valvular disease; one
had mild hypertension; one had ischemic heart disease; and one had a dilatative-con-
gestive cardiomyopathy. All patients were requested to undergo a complex protocol
aimed at dissociating the relative influences of sleep/wake, circadian rhythm and
Sleep and Cardiac Arrhythmia 265

activity/inactivity. This was the first trial ever to separate these different factors that
modulate heart rate and possibly influence the QT interval. Eight of the nine pa-
tients successfully completed the protocol; the ninth refused to complete the pro-
tocol when manipulation of the sleep/wake cycle and activity was scheduled. She
was, however, included in the tabulation for part I of the investigation. The protocol
involved four successive 24h of continuous electrocardiographic monitoring. Pa-
tients remained active during days 1 and 2, following their normal day/night-wake/
sleep cycle (part I). On day 3, patients were kept awake and on continuous total bed
rest, starting at the time of their normal morning awakening, i.e., patients had been
inactive but sleeping for about 8h just prior to the start of the day 3 "bed-rest-
awake" study. On the morning of day 4, at the same time as the start of the bed rest
study a day earlier, patients were allowed to fall asleep. Their short sleep depriva-
tion, despite the 8-h time rotation and the dissociation from the normal circadian
rhythm, allowed the eight remaining patients to sleep until early afternoon. They
were ambulated upon awakening and resumed a normal sleep schedule that evening.
In association with Holter electrocardiographic monitoring, all subjects were
polygraphically monitored while in bed. Electroencephalogram (EEG) , electro-
oculogram, chin electromyogram, respiration and airflow were recorded simulta-
neously with ECG (lead II), on paper. On the polygraph, electrocardiogram (ECG)
recordings were made at 100 mm/second paper speed on supine patients during epi-
sodes of REM and NREM sleep and during inactive wake (after a minimum of 12 h
of bed rest). The QT interval, defined as the time from onset of the QRS complex to
the point where the T wave returned to baseline, was measured simultaneously with
RR intervals for a minimum of five consecutive complexes for each state of vigilance
from each record. The corrected QT (QTc) was calculated from Bazett's formula
(QTc = QT/RR).
Statistical analysis: Data were pooled into four groups: active wake, inactive
wake, non-REM sleep and REM sleep, and were analyzed statistically, using analysis
of variance to test the equality of group means. Pairwise t-test was used to analyze
difference of means between paired groups; analysis of covariance and linear regres-
sion analysis were performed with QT interval as dependent variable, RR as inde-
pendent variable, and behavioral state as group variable. RR-adjusted group means
were calculated from the analysis of covariance for each of the four behavioral states,
and equality of adjusted means was tested by two-tailed t-test.

Results

The results indicated that QT interval was significantly prolonged during REM and
NREM sleep (P < 0.01 for both states) independent of when REM and NREM sleep
occurred during the 24-h cycle. There was a clear trend toward QT interval prolonga-
tion during inactive wake (P < 0.06); and QTc calculation gave similar statistical in-
dication. Linear regression analysis indicated a very clear relationship between QT
and RR intervals (P< 0.001). Analyses of variance of regression coefficient over dif-
ferent behavioral states and slopes of regression lines were nonsignificant, but their
intercepts were significantly different.
266 c. Guilleminault and A. M. Gillis
Our data indicate that activity appears to be an important factor influencing the
QT interval during the awake state. Although the heart rate during active wake did
not differ significantly compared with the inactive wake state or either sleep state,
the QT interval decreased significantly. It may be inferred that withdrawal of sym-
pathetic activity as a result of inactivity may account for the increase observed during
inactivity as well as during sleep. QTc was also shorter during active wake. When
comparison between wake and sleep states was made, we did not find significant re-
sults, probably due to the small number of subjects. However, during sleep, a trend
in further increase in QT interval was noted that cannot be explained by inactivity
and must be related to sleep per se and its effect on the ANS.

Breathing Problems During Sleep: Heart Rate and Cardiac Arrhythmias

In the previous experiments, great efforts were made to screen subjects for respiratory
disturbances during sleep. Obstructive sleep apnea syndrome (OSAS) is, by now,
very well defined. It is a syndrome clinically defined by a constellation of symptoms,
two of which are very commonly associated: heavy snoring at night interrupted by
periods of apnea, and daytime somnolence. Polygraphic monitoring during sleep
indicates repetitive complete or partial obstruction of the upper airway associated
with persistence of diaphragmatic movements. Oxygen saturation drops of varying
severity are simultaneously noted.
We had the opportunity to perform several analyses on nocturnal or 24-h Holter
ECGs obtained simultaneously with continuous nocturnal polygraphic recordings on
400 OSAS patients with an intact ANS, 25 patients presenting OSAS and ANS le-
sions (five Shy-Drager syndrome patients, seven chronic nephropathy patients under-
going dialysis, three heart transplant recipients and ten insulin-dependent diabetes
mellitus patients), 20 OSAS patients who underwent tracheostomy and 20 OSAS pa-
tients treated by nasal continuous positive airway pressure (nasal CPAP).
We also studied 50 subjects free of autonomic neuropathy and OSAS as control
population, using the same protocol. Each Holter ECG was processed by computer
methods that provide standard arrhythmia evaluation and determine the RR interval
between individual QRS complexes. The RR intervals are also graphically displayed
in milliseconds as a function of time, with increases in heart rate shown as a decrease
in RR interval and vice versa.

Experiment 3

Bradycardia and Obstructive Apnea in Patients with and without ANS Lesions
Cyclical variation of heart rate was defined as bradycardia followed by tachycardia
recurring in phase with apnea and with the start of respiration. The pattern of brady-
tachy arrhythmia was seen in all OSAS patients with normal ANS and absence of
atrial fibrillation during the sleep-related abnormal breathing patterns but disap-
peared completely during wakefulness [3]. This pattern was not seen in the control
popUlation. When OSAS subjects were submitted to nasal CP AP, the pattern was
Sleep and Cardiac Arrhythmia 267

progressively blunted in relationship to the progressive increase in positive airway

pressure. When the positive pressure reached a value that controlled the airway ap-
propriately during sleep, avoiding even partial collapse, there was complete disap-
pearance of the cyclic variation of heart rate (CVHR). We have defined CVHR as a
rhythmic change in heart rate between at least 9 and 15 beats per minute seen in as-
sociation with a partial or complete airway obstruction. This was seen with different
nasal CPAP values depending on the patient. If the positive pressure was decreased
by 1 or 2 cm/H20 after obtaining normal ventilation, even if oxygen saturation was
maintained above 90%, as indicated by ear oximetry (Biox), there was reappearance
of CVHR. In 15 cases of OSAS, positive pressures were reduced and increased sev-
eral times during sleep to appreciate the reappearance of CVHR. In each instance,
CVHR reappeared with little oxygen desaturation, i.e., there was presence of mild
bradycardia in association with modest decrease in airway patency and much before
reappearance of complete apnea and significant hypoxia.
Conversely, none of the patients with ANS lesions and OSAS presented a
marked CVHR, despite significant hypoxia. Mean lowest measured oxygen satura-
tion for the total group (n = 25) during sleep was 78 ± 8.6 torr. All patients presented
a regular sinus tachycardia despite significant OSAS. The mean respiratory distur-
bance index (RDI) for the group was 38 ± 14. (RDI is obtained by the following for-
mula: number of (apneas + hypopneas) X total sleep time in minutes, divided by
60.) No variation was noted in eight cases with significant autonomic lesions; all
others had a severe blunting of CVHR compared with age-matched, RDI-matched
OSAS patients without ANS lesions. It thus appears from this first investigation that
the bradycardia noted in OSAS patients is related not only to the hypoxia but also to
the reduction in size of the upper airway.

Bradycardia, Obstructive Apnea, and Pharmacological Manipulation

A second experiment was performed with 20 OSAS patients who had no evidence of
ANS lesions. These patients received intravenous injections of atropine sulfate and
edrophonium HCI. Atropine sulfate was given at 1-3mg dosage, and edrophonium
at lOmg dosage. There was a clear dose response curve with higher dosage of atropine
sulfate: progressively more important blunting of the cyclical variation of heart rate
was noted in association with OSA. The effect of atropine was not related to an im-
pact on ventilation per se. Nocturnal polygraphy indicated persistence of apnea and
hypopnea similar in duration and importance of oxygen desaturation to those seen
just prior to drug injection. The effects of edrophonium injection were judged best
after blunting of CVHR by atropine. Edrophonium per se did not induce CVHR in
seven normal control subjects monitored during sleep, but it counteracted the effect
of atropine injection and led to the reappearance of CVHR in OSAS patients with-
out autonomic nervous system lesions who had been pretreated by atropine sulfate
and presented continuous occlusive sleep apnea. It appears that the bradycardia as-
sociated with OSAS is mediated by the ANS.

Bradycardia, Obstructive Apnea, and High-flow Oxygen

Once atropine and edrophonium had been investigated, a new protocol was im-
plemented with administration of high-flow oxygen during sleep in 15 OSAS pa-
268 C. Guilleminault and A. M. Gillis

tients. During NREM sleep, 100% oxygen was administered by nasal prongs for a
mean of 20 min. Despite the persistence of OSAS and maintenance of a high-oxygen
saturation there was presence of CVHR with each OSA. There was thus a direct but
limited effect of oxygen saturation on CVHR. Oxygen saturation blunted the brady-
cardia seen with sleep apnea and the secondary hypoxia. The high-flow oxygen did
not block the cyclical variation of breathing and apnea still occurred; however, the
end of the apnea was seen with an oxygen tension around 90-95 torr. There was a
large variation in the amount of bradycardia blunted by high-flow oxygen.
The results of the nasal CPAP investigation and hyperoxia experiment gave con-
cordant results, i.e., despite great improvement in oxygen saturation, there was per-
sistence of bradycardia in association with upper airway reduction. Similarly, atropine
sulfate was unable to completely eliminate the CVHR at the given injection dosage.
The combined administration of high-flow oxygen and atropine sulfate (i.v.) suc-
ceeded in completely eliminating the bradycardia linked to hypoxia and OSA. Zwil-
lich et al. [6] had already noted that in association with oxygen administration, there
was a blunting of bradycardia associated with OSA. Martin et al. [5] reported a simi-
lar observation. But our data indicate that hypoxia seems to account for only a part
of the bradycardia. The reduction in upper-airway size per se is responsible for a part
of the bradycardia seen, and reflexes other than those triggered by hypoxia must
be involved in the bradycardia noted. Considering the important effect of atropine
on CVHR, it is possible that the parasympathetic system plays a major role on this
variable, and the bradycardia associated with apnea and hypopnea. It would be mod-
ulated not only by hypoxic influences but also by other elements, possibly speed of
flow in the upper airway or variation of pressure (possibly transpharyngeal pressure)
due to a discrete reduction in size of the upper airway.
Investigation of CVHR in association with obstructive apnea and hypopnea indi-
cated that there was always a more pronounced bradycardia during REM sleep than
during NREM sleep. The fact that apneas are often longer and are associated with
more oxygen desaturation undoubtedly plays a role. However when, in the same
subjects, a systematic investigation of 40 NREM and REM sleep apnea of similar
duration and hypoxia level was performed, the bradycardia was more marked during
REM sleep. We interpreted this as indicative of the dominance of vagal tone during
this sleep state.
In summary, it appears that the bradycardia associated with OSA reflects the
degree of hypoxia, size of the airway, the sleep state, and ANS lesions as well, as
atropine with hyperoxia blunts or eliminates the bradycardia. OSA is also associated
with conduction disturbances or cardiac arrhythmias.

Experiment 4

In a previous report, we have indicated that OSAS is frequently associated with car-
diac arrhythmias. In a review of 400 patients monitored polygraphically during one
night with associated Holter ECG, we found that conduction disturbances or cardiac
arrhythmia were present in 48% [2]. Sinus arrests ranging form 2.5 to 13 s were seen
in about 11% of the population and were generally associated with oxygen desatura-
tion below 70%. In a review of different studies on cardiac arrhythmias associated
Sleep and Cardiac Arrhythmia 269

with OSAS, Sheppard reported that others found similar results, even in studies with
much smaller (20-30 patients) patient populations. Sinus arrest was noted in 9%-
10% of the population investigated. We thus decided to specifically investigate our
patient population presenting a combination of surgical or pathogenic lesions of their
ANS together with OSAS. As already indicated, 25 patients were investigated; 18 of
them presented oxygen saturation drops below 60% (Biox ear oximeter). All had
oxygen saturation drops to at least 78% during their monitored night and, with the
exception of premature ventricular complexes (PVC), no other arrhythmia was
noted; in particular, no sinus arrest. Considering the frequency at which cardiac ar-
rhythmias have previously been reported in this population, we would have expected
to note some statistically significant increase. Also, with the possible exception of the
heart transplant recipients, coronary artery vessel lesions may also have been omitted
from the medical history. However, none were seen. It must be emphasized that five
of the above patients died within 2 years of the monitoring, and in four out of five
the terminal event was dearly linked to OSA, with death during sleep and significant
worsening of apnea in patients without tracheostomy. One interpretation of the find-
ing is that the ANS lesions did not allow the observation of the cardiac arrhythmia
expected with severe OSAS in intact individuals.

Experiment 5

In a further examination of the possible factors involved in the appearance of cardiac

arrhythmias in OSAS, we investigated the QT interval in 12 male OSAS patients
during NREM and tonic REM sleep apneic episodes. Each patient had documented
OSAS based upon clinical history and nocturnal polygraphic recording. The mean
age was 54 ± 10 years. The mean RDI was 45 ± 9, and mean lowest oxygen satura-
tion was 71 ± 6%. All patients were studied following their regular wake/sleep sched-
ule, and no manipulation of their total sleep time or sleep period during the 24 h was
performed. ECG recordings on the polygraph were made at 25 mm/second paper
speed. QT was defined as in experiment 2. RR interval was measured simultaneously.
Great effort was made to exactly define the arousal response associated with the end
of each apnea, and successive QT and RR were measured from onset of each apneic
event until the EEG definition of the arousal response and the complexes associated
with the arousal response were also calculated. Oxygen saturation drops and dura-
tion of apnea were simultaneously calculated. The corrected QT (QTc) was also cal-
culated from Bazett's formula (QTc - QT/RR). A comparison was made between
the first two QT intervals at the beginning of apnea, the last two QT intervals just
preceding the arousal response, and the first two QT intervals associated with arousal.
Statistical analysis was performed using paired t-test.

Results

There were a minimum of ten and maximum of 35 QT intervals recorded during one
obstructive apneic event. A total of 70 apneic events were analyzed. Independent of
the apnea duration, a similar pattern was noted. There was a progressive lengthening
270 c. Guilleminault and A. M. Gillis
of the RR interval with each obstructive apneic event. The QT interval was longer at
the end of apnea than at the beginning, but the QTc decreased significantly from be-
ginning to end of an apnea. The mean QT interval at the beginning of apnea was
486 ± 48 ms, with a mean RR interval of 1019 ± 131 ms. At the end of apnea, before
the arousal response, mean QT interval was 528 ± 64 ms, with a mean RR interval of
1499 ± 128ms. The mean QTc was 482 ± 34ms at the beginning of apnea; it was
435 ± 34ms at the end of apnea and 477 ± 37ms at arousal. There was a statistically
significant difference between "beginning of apnea" and "end of apnea" (mean QT
interval P:::;O.OOO1, mean RR interval P:::;O.OOOI and mean QTc P:::;O.OOOI). The
mean QTc at the end of apnea was also significantly different from mean QTc arousal
(P:::; 0.0001), but there was no statistical difference between mean QTc arousal and
mean QTc "beginning" (P:::; 0.35, T = 0.942). It must be noted that no significant
sleep stage changes were seen during one specific apnea.
Some apneas not selected for the above analysis were associated with significant
bradycardia and asystole up to 6 s duration in the above patients. One of these apneas
was monitored during phasic REM sleep - a different situation compared with the
above results - and presented a 6-s sinus arrest. The mean QT interval at the begin-
ning was 502 ms, the mean RR interval was 1054 ms, the mean QTc was 490 ms; just
before sinus arrest the mean QT was 549 ms, with a mean RR of 1613 ms and a mean
QTc of 435 ms.
These results indicate that, simultaneous with a significant parasympathetic dis-
charge, a sympathetic element is present during obstructive apnea and a sympathetic-
parasympathetic interaction is continuously occurring, modulating the EeG changes
seen in association with apnea.

Summary and Conclusions

Our series of experiments has shown the interaction between circadian rhythm, sleep
and sleep states, activity/nonactivity cycle and respiration problems during sleep on
heart rate and QT segment. It emphasized the multiple factors that must be investi-
gated when one desires to understand increased frequency of cardiac arrhythmias
during sleep and the different sleep states. Undoubtedly, most of the modulations
exercised through the ANS will be influenced by the passage from wake to sleep; and
the resetting of sympathetic and parasympathetic balances in association with sleep
states may increase the risk of ventricular arrhythmia in certain patients during sleep.
However, the understanding of the type of risk that will be enhanced, the type of pa-
tient that will be subject to this risk increase, and the mechanisms by which increased
risk will occur will require the step-by-step approaches partially outlined in the above
experiments. Very little, if any, attention has heretofore been paid to the impact of
the changes of alertness in the development or worsening of cardiac arrhythmia dur-
ing the night. In association with the investigation of nocturnal hemodynamics, the
systematic exploration of the cardiovascular phenomena linked to the appearance of
the different sleep states will help us to more appropriately appreciate risk factors
presented by patients with end-organ lesions during the night and to better adapt the
treatment. Finally, our last experiment demonstrates that even if the influence of
Sleep and Cardiac Arrhythmia 271

one arm of the ANS may be obvious (the parasympathetic tone) the influence of the
other arm (the sympathetic tone) may be very significant. The ANS may modulate
the conduction system of the heart at many levels, and depending on the node under
consideration, parasympathetic or sympathetic activity may be more dominant.

Acknowledgments. This work was supported by General Clinical Research grant 00070 funded by
the National Institutes of Health and by grant AG 07772 from the National Institute of Aging. We
thank Boyd Hayes and David Cobasko for their technical help and Alison Grant for her editorial as-
sistance.

References

1. Gillis AM, MacLean KE, Guilleminault C (1985) The effect of sleep on the QT interval in pa-
tients with ventricular arrhythmia. Sleep Res 14: 234
2. Guilleminault C, Connolly S, Winkle R (1983) Cardiac arrhythmia during sleep in 400 patients
with sleep apnea syndrome. Am J Cardiol 52 : 490-494
3. Guilleminault C, Connolly S, Winkle R, Melvin K, Tilkian A (1984) Cyclical variation of the
heart rate in sleep apnea syndrome. Lancet 1: 126-131
4. Guilleminault C, Winkle R, Coburn S (1982) 24-hour Holter ECG and sleep recording in heart
transplant patients. In: Koella W (ed) Sleep 1982. Karger, Basel
5. Martin RJ, Sanders MH, Gray BA (1982) Acute and long-term ventilatory effects of hyperoxia in
the adult sleep apnea syndrome. Am Rev Respir Dis 125: 175-180
6. Zwillich C, Devlin T, While D, Douglas N, Wei! J, Martin R (1982) Bradycardia during sleep
apnea: characteristics and mechanism. J Clin Invest 69: 1286-1292
Chapter 22

Pulmonary Hemodynamics
in Obstructive Sleep Apnea Syndromes
J.KRIEGER, E. WEITZENBLUM, B.REITZER, and D.KuRTZ

Introduction

From a historical point of view, sleep apnea syndromes first caught the attention of
the medical community because of their cardiopulmonary sequelae. As a matter of
fact, the picture described by Burwell et al. [4] as the pickwickian syndrome was
dominated by signs of right heart failure, in association with alveolar hypoventila-
tion, polycythemia, and daytime sleepiness. Even though Burwell et al. [4] did not
recognize sleep apneas as the main characteristic of the syndrome they described, it
seems most likely that their report was the first complete medical description of the
sleep apnea syndrome. The recognition by Gastaut et al. [8] almost 10 years later of
a specific respiratory pattern during sleep characterized by repeated respiratory ar-
rests provided a new understanding of the pickwickian syndrome and led to the con-
cept of sleep apnea syndrome [9]. Indeed, it appeared that repeated sleep apneas
could occur in patients who do not present with the picture initially described by Bur-
well et al. [4]. Furthermore, the present understanding of the disease implies that the
respiratory disturbance during sleep may be present for years, expressing itself only
by noisy snoring, but being silent from a clinical point of view. Only under the effect
of precipitating factors, which are not completely understood, do clinical manifesta-
tions appear, the ultimate form of which constitutes the full-blown pickwickian syn-
drome. Therefore, in this chapter, we will first briefly describe the main clinical fea-
tures of obstructive sleep apnea (OSA) syndromes before further analyzing pulmo-
nary hemodynamics in OSA patients from two points of view: immediate pulmonary
hemodynamic changes related to the occurrence of obstructive apneas during sleep
and long-term pulmonary hemodynamic sequelae affecting the daytime cardiopul-
monary status of OSA patients.

Clinical Features of Obstructive Sleep Apnea Syndromes

OSA syndromes express themselves by nocturnal manifestations, which most often

lead spouses or bed partners to complain, and by daytime manifestations, which may
seriously compromise the patients' socioprofessional status [10].

Daytime Manifestations are dominated by somnolence. Daytime somnolence may be

of various degrees: in some patients it is limited to an increased facility for dozing off
Pulmonary Hemodynamics in Obstructive Sleep Apnea Syndromes 273

in a boring situation, whereas it makes any kind of social or professional activity im-
possible in the most severely affected patients. Intellectual deterioration and atten-
tion and memory impairment further contribute to the socioprofessional difficulties
experienced by these patients. Decreased libido and sexual impotence are frequent.
Intolerance to exertion, dyspnea and ankle edema may be signs of right heart failure,
which is not present always.

Nocturnal Manifestations are dominated by snoring. Snoring generally precedes day-

time manifestations by years or decades, but most often becomes heavier when clini-
cal manifestations appear. Typically, these patients have restless sleep, with re-
peated arousals, jerking, groaning, and occasional sleepwalking. Nocturnal polyuria
is a quasiconstant feature. Surprisingly, the patients are remarkably unaware of all
the trouble they cause and only seldom complain of poor sleep quality.

The Clinical Examination only exhibits nonspecific alterations: obesity is frequent,

but not constant; hypertension is often present. Erythrocyanosis, ankle edema, and
breathlessness appear only in patients with cardiopulmonary disturbances.
Thus, the diagnosis relies mainly upon the medical history, which is rarely pro-
vided by the patient himself, making an interview with the spouse or relatives im-
portant.
The results of ear, nose, and throat examination and pulmonary function evalua-
tions, which have major therapeutic implications, will not be further described here.
The diagnosis, suspected on the basis of the medical history, can only be con-
firmed by a polysomnographic recording. Studies of pulmonary hemodynamics dur-
ing sleep can be combined with polysomnographic recordings.

Polysomnographic Features of Obstructive Sleep Apnea Syndromes

Sleep Apneas

Type, Duration, and Distribution

Since the initial description of sleep apneas by Gastaut et al. [8] three types of apneas
have been described: (1) central apneas are characterized by an interruption of respi-
ratory efforts during the apneas; (2) in contrast, respiratory efforts with a progres-
sively increasing amplitude persist during obstructive apneas; and (3) mixed apneas
are made up by a generally short initial central component, followed by a more pro-
longed obstructive episode. OSA syndromes are defined by the occurrence of mainly
obstructive and mixed apneas, even though a limited number of central apneas are
often present. The proportion of various types of apneas may vary from patient to
patient but is relatively constant in a given patient. The duration of sleep apneas
varies over a wide range, from 10 s to 2 min or more. At sleep onset, apneas are gen-
erally short in duration, or restricted to a decrease in breathing amplitude without re-
spiratory arrests (hypopneas); as sleep progresses, the duration of the apneas in-
creases, the apneas are generally longer in rapid eye movement (REM) sleep than in
non-REM sleep.
274 J. Krieger et al.

Polysomnographic Concomitants of Sleep Apneas

Changes in the state of vigilance closely parallel the changes in breathing: the respi-
ratory arrests occur when sleep sets in or deepens and breathing resumption is ac-
companied by an arousal that, although not perceived by the patients, is clearly iden-
tifiable in simultaneous EEG recordings.
Transcutaneous evaluation of blood gases by ear oximetry or transcutaneous P0 2
and PC02 measurement have made it possible to analyze the consequences of apneas
on blood gases. Understandably, each apnea is accompanied by a rapid decrease in
blood oxygenation, with hypoxemia being more severe during prolonged apneas.

150-

30-
100-

60-

20'

20'

o·

Fig.t. Polysomnographic recording during obstructive sleep apneas.toHeart rate (beats/min- I);
SaO], oxyhemoglobin saturation (%); V, ventilatory flow (l/S-I); PAP, transpulmonary arterial
pressure (mmHg); Po, esophageal pressure (mmHg); PAP tm, transmural pulmonary arterial pres-
sure (mmHg)
Pulmonary Hemodynamics in Obstructive Sleep Apnea Syndromes 275

The time relationship between apneas and blood oxygenation cannot be precisely
analyzed because of the delay between the actual change and the oximeter reading
due to circulation time and sensor response time. Nevertheless, it seems clear that
the lowest oxygenation values are reached during the first breaths upon breathing re-
sumption (Fig. 1). Because body stores of CO2 are higher than those of O 2 , changes
in blood CO2 content are more progressive, leading to a progressive CO2 retention
with the repetition of apneas.
The heart rate exhibits cyclic variations which parallel the occurrence of repeti-
tive apneas, due to a decreased heart rate during apneas and an increased heart rate
during breathing resumption (Fig. 1).
Measurements of intrathoracic pressure during sleep by means of an esophageal
balloon, which accurately reflects the pleural pressure, have shown that during
obstructive apneas the ineffective respiratory efforts against an occluded airway gen-
erate negative (subatmospheric) pressure swings, the amplitude of which increases
progressively toward the end of the apneas, reaching values of - 60 to - 80 cm H 20.

Pulmonary Hemodynamics During Obstructive Sleep Apneas

It has long been recognized that sleep in OSA patients was accompanied by an in-
crease in pulmonary arterial pressure (PAP) [5,6,12,16], this increase being parallel
with the severity of the hypoxemia and hypercapnia which was simultaneously re-
corded [5]. The increase in PAP was more severe with the deepening of sleep from
stage 1 to 4, the highest values being recorded in REM sleep [5]. The increases in
pulmonary systolic pressures were greater than those in diastolic pressure, indicating
an increased pulse pressure [5]. Later studies by the Stanford group [17,19] reported
similar data and showed, in addition, that in those patients in whom they could be
recorded, pulmonary artery wedge pressures were also increased, although to a les-
ser degree [3].
However, the understanding of the described changes in pulmonary arterial pres-
sure is made difficult by the fact that in these studies transpulmonary arterial pres-
sure was measured. This pressure is also called intravascular pulmonary arterial pres-
sure (i.e., the pressure difference between the arterial lumen and the atmosphere).
Transpulmonary arterial pressure is the algebraic sum of two components: trans-
mural pressure (i.e., the pressure difference between the arterial lumen and the
thoracic cavity) and thoracic pressure (i.e., the pressure difference between the
thoracic cavity and atmosphere). Transpulmonary arterial pressure measurements
are therefore directly influenced by the changes in thoracic pressure which occur dur-
ing obstructive apneas. Because of negative intrathoracic pressures during ineffec-
tive inspiratory efforts, transpulmonary arterial pressure decreases to negative val-
ues (less than atmospheric pressure) [18]. Since both the heart and the pulmonary
arteries are submitted to thoracic pressure changes, transpulmonary pressure mea-
surements are not a relevant hemodynamic parameter: the transmural pressure
(transpulmonary pressure - intrathoracic pressure) represents the pressure against
which the right ventricle must work.
276 J. Krieger et al.

We therefore tried to approach the transmural pulmonary arterial pressure (PAP

tm) more directly by simultaneously measuring the transpulmonary arterial pressure
by means of a Swan-Ganz catheter inserted through a subclavian vein connected to
a Statham pressure transducer and the esophageal pressure by means of a latex bal-
loon placed in the esophagus at about 35 cm from the nares and connected to a Vali-
dyne pressure transducer. After adequate calibration, the esophageal pressure was
electronically substracted from the transpulmonary pressure by means of a differen-
tial amplifier, calibrated, and tested with known pressures applied to each transducer
before their connection to the subject. This procedure was used in five subjects. The
following preliminary data (obtained from the first patient in whom data analysis is
at present complete) are based on the analysis of 52 apneas.
Figure 1 shows a typical recording during an obstructive apnea. It demonstrates
that the changes in PAP tm during ineffective respiratory efforts occur in a direction
opposite to the changes in transpulmonary arterial pressure and in esophageal pres-
sure (Po). PAP tm increases when esophageal and transpulmonary arterial pressures
decrease. This finding suggests that during inspiratory effort, the decrease in intra-
thoracic pressure produces an increase in venous return and thus an increase in right
ventricular output and an increase in PAP tm.
The evolution of PAP tm over time during the apneas was analyzed by averaging
consecutive peak inspiratory and expiratory values of PAP tm. In all 52 apneas
analyzed, a minimal value of PAP tm occurred at a given time of the apnea. The
apneas were thus analyzed in three successive segments: an initial apneic segment
from the beginning of the apnea to the minimal PAP tm, a final apneic segment from
the minimal PAP tm to the end of the apnea, and a post apneic segment beginning
with the first effective breath until the postapneic maximum was reached, lasting
8±2s. The mean duration of the 52 apneas was 44±3s. The minimal PAP tm oc-
curred at 18 ± 7 s from the beginning of the apnea. During the initial apneic segment,
PAP tm decreased with a slope of -0.3 ± 0.2mmHg· S-1. During the final apneic
segment, PAP tm increased with a slope of 0.2 ± O.lmmHg· S-1 and during the post-
apneic segment, PAP tm increased with a slope of 1.2 ± O.lmmHg· S-1.
Peak inspiratory and expiratory PAP tm pressure readings were correlated posi-
tively with heart rate during the initial apneic segment (r = 0.59; P < 0.001) and
negatively with SaOz during the final apneic segment (r = -0.78; P < 0.001). During
the postapneic segment, PAP tm was correlated with heart rate and with SaOz (r =
0.68 and -0.57, respectively).
We therefore suggest that PAP tm decreases during the initial part of an obstruc-
tive apnea, due to the decrease in cardiac output and in heart rate, and increases dur-
ing the final part due to hypoxic vasoconstriction.
It is likely that on breathing resumption, an increased heart rate and persistent
hypoxemia concur to further increase PAP tm by the conjunction of an increase in
cardiac output and persistent hypoxemic vasoconstriction.
However, these factors are obviously not the only ones that may playa role. In-
creased pulmonary artery wedge pressure has been demonstrated during obstructive
apneas [3], even though to a lesser degree than pulmonary arterial pressure, which
suggests that a decreased left ventricular output could playa role in the increase in
pulmonary arterial pressure [20]. Actually, indirect measurement of left ventricular
stroke volume (LVSV) has shown a decrease in L VSV which correlated with the de-
Pulmonary Hemodynamics in Obstructive Sleep Apnea Syndromes 277

crease in diastolic pleural pressure [20]. The suggested explanation for this observa-
tion was that through ventricular interdependence mechanisms the increased venous
return to the right ventricle could reduce left ventricle preload [20], and thus reduce
left ventricular output.
Thus, it seems that the changes in pulmonary arterial pressure during obstructive
apneas are the consequences of the interaction of multiple factors, including changes
in intrathoracic pressure, hypoxic vasoconstriction, and changes in heart rate. The
intimate mechanisms of changes in heart rate are discussed in the previous chapter
of this book (Chapt.21). Whatever their mechanism, these changes are abolished
when sleep apneas are eliminated, either by tracheostomy [15] or by nasal continu-
ous positive airway pressure [14].

Long-term Daytime Pulmonary Hemodynamic Sequelae of OSA

Right heart failure is considered to be a complication of the OSA syndrome. How-

ever, its occurrence is not well documented in large series of patients, nor have the
factors involved been well established. For instance, it has been suggested that tran-
sient increases in pulmonary arterial pressure related to apnea could result in perma-
nent pulmonary arterial hypertension, by mechanisms which remain to be estab-
lished. Early studies from the Stanford group [17] have indicated a high incidence
(59%) of awake resting pulmonary hypertension defined as a mean pulmonary arte-
rial pressure greater than 20mmHg. However, this figure was obtained in a small
group (n = 22) of OSA patients selected from a larger population. In 50 consecutive
patients with OSA, Bradley et al. [2] found 6 (12%) with right heart failure diag-
nosed on clinical and ECG criteria. These patients were not different from OSA pa-
tients without right heart failure in terms of the number or duration of sleep apneas,
but were more hypoxemic both awake and asleep and were more hypercapnic while
awake. Since they also had higher residual volumes and lower forced expiratory vol-
umes in 1 s, Bradley et al. [2] suggested that diffuse airway obstruction associated
with OSA plays a primary role in the development of sustained hypoxia and/or
hypercapnia and consecutive right heart failure. This suggestion is contradicted by
the data of Leech et al. [11] who showed that in 27 OSA patients with obstructive air-
way disease as well as in 37 OSA patients without obstructive airway disease, the oc-
currence of signs of right ventricular dysfunction was linked to the severity of sleep-
induced respiratory disorders as assessed by an index combining the frequency and
the degree of desaturation. However, clinical and ECG signs of "cor pulmonale"
may appear later than pulmonary hypertension assessed by right heart catheteriza-
tion.
In 46 consecutive OSA patients who underwent polysomnography, pulmonary
function evaluation, and right heart catheterization, we found that 9 (20%) had rest-
ing pulmonary hypertension (> 20 mm Hg). Seven of these nine hypertensive pa-
tients had significant daytime hypoxemia (Pa O 2 < 65 mm Hg). Our results were in
keeping with those of Bradley et al. [2] in that pulmonary hypertensive patients were
not different from nonpulmonary hypertensive ones in terms of apnea index or mini-
mal Sa O 2 during sleep, but were more hypoxemic and hypercapnic during daytime;
278 J. Krieger et al.

in addition, they were older and more overweight. These data suggest that daytime
hypoxemia plays a major role in the development of permanent pulmonary hyper-
tension and that chronic airway obstruction is an important factor in the develop-
ment of daytime hypoxemia in OSA patients. However, only five of nine pulmonary
hypertensive patients could be considered to have permanent chronic airway
obstruction on the basis of pulmonary function tests (forced expiratory volume in 1 s
over forced vital capacity < 0.60). Other recent data [16] showed a similar incidence
(20%) of resting pulmonary hypertension in OSA patients, some of them without
any cardiac or pulmonary abnormality, thus supporting the hypothesis that repeated
episodes of pulmonary hypertension during sleep may by themselves result in sus-
tained pulmonary hypertension. This is further supported by the effects of sleep
apnea elimination by tracheostomy [1, 13], which results in the normalization of pul-
monary arterial pressure, and by the fact that in patients with chronic obstructive
pulmonary disease associated with OSA, the pulmonary artery pressure decreases
only in those patients in whom the apneas have been eliminated by tracheostomy [7].
Thus, the development of pulmonary arterial hypertension in OSA patients is
clearly multifactorial, being linked to daytime hypoxemia, which itself results from a
complex interaction between sleep apneas, airway obstruction, and obesity.

References

1. Aubert-Tulkens G, Willems B, Veriter C, Coche E, Stanescu DC (1980) Increase in ventilatory

response to CO 2 following tracheostomy in obstructive sleep apnea. Bull Eur Physiopathol
Respir 16: 587-593
2. Bradley TD, Rutherford R, Grossman RF, Lue F, Zamel N, Moldofsky H, Phillipson EA
(1985) Role of daytime hypoxemia in the pathogenesis of right heart failure in the obstructive
sleep apnea syndrome. Am Rev Respir Dis 131: 835-839
3. Buda AJ, Schroeder JS, Guilleminault C (1981) Abnormalities of pulmonary artery wedge pres-
sure in sleep-induced apnea. Int J Cardiol 1: 67-74
4. Burwell CS, Robin ED, Whaley RD, Bickelmann AG (1956) Extreme obesity associated with
alveolar hypoventilation. A Pickwickian syndrome. Am J Med 21: 811-818
5. Coccagna G, Mantovani M, Brignani F, Parchi C, Lugaresi E (1972) Continuous recording of
the pulmonary and systemic arterial pressure during sleep in syndromes of hypersomnia with
periodic breathing. Bull Eur Physiopathol Respir 8: 1159-1172
6. Doll E, Kuhlo W, Steim H, Keul J (1968) Zur Genese des Cor pulmonale beim Pickwick-Syn-
drom. Deutsch Med Wochenschr 93:2361-2365
7. Fletcher EC, Schaaf JW, Miller J, Fletcher JG (1987) Long-term cardiopulmonary sequelae in
patients with sleep apnea and chronic lung disease. Am Rev Respir Dis 135: 525-533
8. Gastaut H, Tassinari CA, Duron B (1965) Etude polygraphique des manifestations episodiques
(hypniques et respiratoires) du syndrome de Pickwick. Rev Neurol112: 568-579
9. Guilleminault C, Tilkian A, Dement WC (1976) The sleep apnea syndromes. Annu Rev Med
27:465-484
10. Krieger J (1986) Les syndromes d'apnees du sommeil de l'adulte. Bull Eur Physiopathol Respir
22:147-189
11. Leech JA, Onal E, Givan V, Gallestegui J, Lopata H (1985) Right ventricular dysfunction re-
lates to nocturnal hypoxemia in patients with sleep apnea syndrome. Am Rev Respir Dis 131:
A 104
12. Lonsdorfer J, Meunier-Carus J, Lampert-Benignus E, Kurtz D, Bapst-Reiter J, Fletto R,
Micheletti G (1972) Aspects Mmodynamiques et respiratoires du syndrome pickwickien. Bull
Physiopathol Respir 8: 1181-1192
Pulmonary Hemodynamics in Obstructive Sleep Apnea Syndromes 279

13. Lugaresi E, Coccagna G, Mantovani M, Brignani F (1973) Effects of tracheostomy in two cases
of hypersomnia with periodic breathing. J Neurol 36: 15-26
14. Marrone 0, Milone F, Ferrara G, Oddo S, Coppola P, Cibella F (1986) Continuous positive air-
way pressure decreases pulmonary arterial pressure in obstructive sleep apneas. Am Rev Respir
Dis 133: A343
15. Motta J, Guilleminault C, Schroeder JS, Dement WC (1978) Tracheostomy and hemodynamic
changes in sleep induced apneas. Ann Intern Med 89: 454-458
16. Podszus T, Bauer W, Mayer J, Penzel T, Peter JH, Wichert P von (1986) Sleep apnea and pul-
monary hypertension. Klin Wochenschr 64: 131-134
17. Schroeder JS, Motta J, Guilleminault C (1978) Hemodynamic studies in sleep apnea. In: Guil-
leminault C, Dement WC (eds) Sleep apnea syndromes. Liss, New York, pp 177-196
18. Shepard JW (1986) Hemodynamics in obstructive sleep apnea. In: Fletcher EC (ed) Abnor-
malities of respiration during sleep. Grune and Stratton, Orlando, pp 39-61
19. Tilkian AG, Guilleminault C, Schroeder JS, Lehrman KL, Simmons FB, Dement WC (1976)
Hemodynamics in sleep-induced apnea. Ann Intern Med 85: 714-719
20. Tolle FA, Judy WV, Yu PL, Markand ON (1983) Reduced stroke volume related to pleural
pressure in obstructive sleep apnea. J Appl Physiol 55 : 1718-1724
Chapter 23
Sleep Apnea Syndrome as an Occupational Disease
P. MONSTAD, I. A. SULG, A. K. ROM, T. NISSEN, and S.1. MELLGREN

Occupational exposure to organic solvents is known to be associated with unspecific

"neurasthenic" complaints like impaired short term memory, reduced concentration,
headache, irritability, and depressive symptoms [5, 15, 18]. Scandinavian studies
showing lowered scores in standard neuropsychiatric tests in persons chronically ex-
posed to organic solvents have given rise to the concept of a specific syndrome, the
organic solvent encephalopathy. The existence of such a syndrome has been
doubted, partly because of British studies failing to replicate the findings of such a
syndrome [1].
Clinical neurophysiological examinations (EEG, EMG etc.) generally do not
give any abnormal findings in patients referred for possible organic solvent encepha-
lopathy, except for some minor evidence of polyneuropathy in some cases. Most pa-
tients referred to our department for a diagnostic work-up of possible organic solvent
encephalopathy complained of tiredness and increased need of sleep, an observation
also made by others [18]. This prompted us to initiate a sleep study in these patients.
The results in a pilot study indicate a high prevalence of sleep apnea in such cases: 7
out of 15 patients had a pathological sleep apnea index (more than 5 apneas per hour
of sleep) [12]. In another study of eight trichlorethane-exposed workers a significant
increase in apneas compared to unexposed controls was found. The present report
gives the results of our further studies, indicating that sleep apnea might be a com-
mon occupational disorder among solvent-exposed workers.

Case Material

Eighteen male house painters were invited to participate in the study. Inclusion
criteria were as follows: (a) more than 10 years of solvent exposure as a painter, and
(b) active working in ordinary day routines at the time of the study. Sixteen pro bands
agreed to participate (mean age 47.8 years, range 31-63 years). An age matched con-
trol group of 17 males (mean age 46.4 years), mainly patients remitted for orthopedic,
rheumatic, and ophthalmologic symptoms, selected without knowledge of their pos-
sible sleep complaints, were studied during their stay in hospital.

Methods

Polysomnography was performed with an ambulatory cassette system (Oxford Medi-

log 9000). Electroencephalogram (EEG), eye movements (EOG), respiration (by
Sleep Apnea Syndrome 281

means of the impedance plethysmography and a thermistor), and electrocardiogram

(ECG) were recorded. Also a noninvasive pulse oximetry was included in most of
the examinations. The polysomnographic recording was performed during a night
after ordinary working hours. In the second part of the study 10 solvent-exposed
workers who showed abnormal sleep apnea index during polysomnography in the
1st-3rd nights after exposure were retested at least 10 days later (13-100 days after
exposure). In order to avoid error due to regression to the mean, four of these work-
ers were reexamined during the first night after reexposure.

Results

Mean total sleep time, median number of apneas and apnea index in the solvent-
exposed group of controls are shown in Table 1. Total sleep time and amount of
stage 3 and 4 sleep (delta sleep, DSW) were essentially equal in the two groups. The
apnea index was significantly (P < 0.05, Wilcoxon's unpaired rank test) higher in the
exposed group. Five of the 16 painters had more than 5 apneas per hour of sleep,
satisfying the commonly used criteria for the sleep apnea syndrome. No serious
hypoxia was found by pulse oximetry in any of the subjects; the lowest oxygen satura-
tion recorded was 78%, in the subject who also had the highest apnea index. The
single control case showing abnormal apnea index turned out to be a heavy user of
benzodiazepines, drugs known to be associated with sleep apnea [3]. The relation-
ship between sleep apnea and age in the two groups is shown in Fig. I.
In the second part of the study a significant drop in apnea number (P < 0.05) was
seen after an exposure-free interval (Fig. 2). In the four patients available for a sec-
ond test during work involving solvent exposure a small rise in apnea periods was
seen (mean apneas/night: 93 in first test, 99 in second test), indicating that the ob-
served drop is not a regression to the mean phenomenon.
Four of the 16 painters had complained about impaired memory and increasing
tiredness. Three of these showed pathological sleep apnea. Neither blood pressure
nor smoking habits were significantly different between the two groups. None of the
tested persons could be considered obese. Interviews did not indicate excessive
alcohol intake in any of the subjects. However, follow-up gave indications of alcohol
abuse in one subject.

Table 1. Polysomnographic data in solvent-exposed painters and controls

n Age TST DSW AI Cases where

(mean) (min) (min) (median) AI>S

House painters 16 47.8 377 SO 2.9S' S

Controls 17 46.4 379 49 0.46 1

, P<O.OS.
TST, total sleep time; DSW, delta slow wave (stage 3-4) sleep; AI, apnea index = mean number of
apneas (episodes > 10 s) per hour of sleep
282 P. Monstad et al.

APNEAS

20t
10
50
4
40
3

20
.. Fig. I. Total number of recorded

... . .. .
apneas plotted versus age in con-
10
5 o 0 o trols (x) and in solvent-exposed
.0 0 house painters (0)
30 40 50 60 70
AGE

APNEAS

200

150

100

50
Fig. 2. Apneas per night recorded during the first
to the third night after solvent exposure (test 1)
compared to reexamination results 13-100 days
after exposure (test 2) in solvent-exposed sleep
apnoics
o~------+-------~------
TEST 1 TEST 2

Discussion

The results presented indicate a high prevalence of sleep apnea in house painters ex-
posed to organic solvents. Our results from previous studies, combined with the ob-
served reduction of sleep apnea after an exposure-free interval, indicate that organic
solvents can cause sleep apnea as a partially reversible disorder. The assumption that
organic-solvent exposure can cause sleep apnea has earlier been made by Wiese after
observing a single case of serious central sleep apnea after trichlorethane exposure
[17].
In our study, most of the observed apneas were of the obstructive type. The
pathogenetic mechanisms by which organic solvents cause mental impairment are
Sleep Apnea Syndrome 283

Table 2. Relative mortality in 1970-1980 for males 20-69 years of age and economically active in
1970, by occupational class

Class Total Cardio- Cancer Accidents! deaths

mortality vascular from unnatural
causes causes
Woodwork 87 87 90 95
Painting and paperhanging 106 107 103 111
Construction work 99 97 95 125
(not elsewhere classified)

All economically active = 100

not well understood. An excess of sleep apneas, as found in our study, might be an
important factor in producing the neuropsychiatric complaints of these workers. The
present study provides no clue to the mechanism by which organic solvents cause
sleep apnea. The most commonly used "organic solvent" is ethyl alcohol, known to
cause sleep apnea after oral ingestion [4]. However, also mucosal irritation of upper
airways can cause sleep apnea, e.g., after occupational exposure to guam dust [9].
In recent years studies have been performed which indicate that sleep apnea is an
independent risk factor for cardiovascular as well as for cerebrovascular morbidity
[8, 10, 13]. Possible etiological mechanisms have been described by Krieger and
Guilleminault elsewhere in this volume [2, 7]. Sleep apnea may cause elevation of
systemic as well as pulmonary blood pressure and also exert an arrhythmogenic ef-
fect [2, 6, 14, 16].
In view of the high prevalence of sleep apnea found among solvent-exposed
workers in this study, it is tempting to speculate on whether they are also a high-risk
group for cardiovascular mortality. A recent study [11] indicates that organic sol-
vents (at least trichlorethane) may exert a cardiotoxic, arrhythmogenic effect. Thus
sleep apnea might be a particularly strong risk factor under these circumstances.
Table 2 shows the age adjusted mortality rates among Norwegian males in different
trades [8]. A higher mortality in house painters, than in other construction workers
is seen. This increased mortality is mainly due to cardiovascular diseases. Although
cardiovascular mortality is strongly influenced by the life style and socioeconomic
conditions, we believe that these figures indicate that the possibility of organic sol-
vents being a risk factor for cardiovascular disease deserves further study.

References

1. Cherry N, Venables H, Waldron HA (1984) British studies on the neurophysiological effects of

solvent exposure. Scand J Environ Health 1(10): 10-12
2. Guilleminault C, Gillis AM (1988) Influence of sleep, activity and circadian rhythm on heart
rate, QT interval and cardiac arrhythmias. This volume, chapter 21
3. Guilleminault C, Silvestri R, Mondini S, Coburn S (1984) Aging and sleep apnea: action of
benzodiazepine. J Gerontol 39: 655-661
4. Issa FG, Sullivan CE (1982) Alcohol, snoring and sleep apnea. J Neurol Neurosurg Psychiatr
45:353-359
284 P. Monstad et al.: Sleep Apnea Syndrome

5. Juntunen J (1982) Organic solvent intoxications in occupational neurology. Acta Neurol Scand
[Suppl 98] 69: 105-120
6. Koskenvuo M, Kaprio J, Telakivi T, Partinen M (1987) Snoring as a risk factor for ischemic
heart disease and stroke in men. Br Med J 294: 16-19
7. Krieger J, Weitzenblum E, Reitzer B, Kurtz D (1988) Pulmonary hemodynamics in obstructive
sleep apnea syndromes. This volume, chapter 22
8. Kristoffersen LB (1986) Social differences in the 1970's. Central Bureau of Statistics, Oslo
9. Leznoff A, Haight JS, Hoffstein V (1986) Reversible obstructive sleep apnea caused by occupa-
tional exposure to guar gum dust. Am Rev Respir Dis 133: 935-936
10. Lugaresi E, Cirignotta F, Coccagna G, Piana C (1980) Some epidemiological data on snoring
and cardiocircuJatory disturbances. Sleep 3: 221-224
11. McLeod AA, Marjot R, Monaghan MJ, Hugh-Jones P, Jackson G (1987) Chronic cardiac
toxicity after inhalation of 1,1,1-trichiorethane. Br Med J 294:727-729
12. Monstad P, Nissen T, Sulg lA, Mellgren SI (1987) Sleep apnea and organic solvent exposure. J
NeuroI234:152-154
13. Partinen M, Alihanka J, Lang H, Kalliomaki L (1983) Myocardial infarctions in relation to sleep
apneas. Sleep Res 12: 272
14. Peter JH (1986) Hat jeder dritte Patient mit essentieller Hypertonie undiagnostiziertes Schlaf-
apnoe-Syndrom? Dtsch Med Wochenschr 111 :556-559
15. Sepplil!iinen AM, Lindstrom K, Martelin T (1980) Neurophysiological and psychological picture
of solvent poisoning. Am J Ind Med 1: 31-42
16. Tilkian AG, Motta J (1978) Cardiac arrhythmias in sleep apnea. In: Guilleminault G, Dement
WC (eds) Sleep apnea syndromes. Liss, New York, pp 197-210
17. Wise MG, Fisher JG, de la Pena AM (1983) Trichlorethane and central sleep apnea: a case
study. J Toxicol Environ Health 11: 101-104
18. 0rbrek P, Risberg J, Rosen I (1985) Effect of long term exposure to solvents in the paint indus-
try. Scand J Work Environ Health 11: 1-28
Part VI
Disease in the Cardiovascular
and Nervous System
Chapter 24
Concomitant Manifestations of Disease
in the Cardiovascular and Nervous System: An Overview
K. RASMUSSEN

A purpose of this book is to present mechanisms by which the cardiovascular and

nervous systems may interact with each other in the pathogenesis of disease. The
nervous system may cause or influence manifestations of heart disease, for instance
by creating arrhythmias or through the modification of the response to ischemia. The
heart may influence the brain, for instance by releasing embolic material of various
origins into the blood stream, thus giving rise to cerebral vascular accidents. On a
few occasions the target organ may even retaliate by returning the pathogenetic in-
fluence as some kind of echo. Nervous influence may for instance cause an arrhythmia
which in due course can induce embolism in the brain.
From a clinical point of view, however, these two situations embody only a frac-
tion of all the clinical situations in which symptoms and signs from both the heart and
the brain dominate the picture. Much more frequently, both these organs are targets
of the same general disease process. The purpose of this chapter is to offer a classifi-
cation of such conditions, emphasizing the most important and most commonly over-
looked ones, with no pretence of completeness.

A Classification System

Table 1 presents a classification system of diseases that concomitantly affect both the
cardiovascular and the nervous systems. In the cardiovascular system, diseases may
involve the endocardium, the myocardium, the pericardium, the conduction system,
and the arteries, including the coronary arteries. In the nervous system, the central
nervous system (eNS), the meninges, the peripheral nerves (PNS) or the muscles
may be the target. Like most systems for classifying diseases, the present one is
based on several principles, the main one being that of disease etiology.

Some Examples

Infections

As regards infectious diseases (group 1, Table 1) it is common knowledge that chronic

specific infections such as tuberculosis, syphilis, and sarcoidosis may affect both
organ systems. It is perhaps easier to forget that conventional bacterial or viral infec-
288 K. Rasmussen

Table 1. A classification system

Cardiovascular system Nervous system

1. Infections
Syphilis Aortitis, gumma CNS, meningitis
Tuberculosis Pericarditis CNS, meningitis
Rheumatic fever Carditis Chorea minor
Poliomyelitis Myocarditis Motor neuron
Diphteria Carditis Peripheral nerves (PNS)
Conduction system
Meningococci Myopericarditis Meningitis
Chaga Myocardial Meningoencephalopathy
Sarcoidosis Myocardial CNS, PNS, meningitis
Rubella Congenital lesions Retardation, vision, hearing

2. Hereditary-congenital diseases
Down Congenital lesions Retardation
Noonan Congenital lesions Retardation
Keams-Sayre A-V conduction Retinitis pigmentosa
Degenerative
Muscular dystrophies Myocardial Muscles
Dystrophia myotonica Myocardial Retardation, muscles
Heredoataxias Myocardial CNS,ophtalmic
Glycogen storage Myocardium CNS
Mycopolysaccaridosis Myocardial CNS
Tuberous sclerosis Rhabdomyoma CNS-lesions
Fetal alcohol Congenital lesions Retardation

3. Immunological diseases
Systemic lupus Endo-, myo-, pericardium Vessels
Arteries, conduction
Periarteritis nodosa Vessels Vessels
Temporal arteritis Arteries Arteries
Takayasus arteritis Arteries Arteries
Rheumatoid arteritis Arteries, valves Arteries
Amyloidosis Myocardial PNS

4. Other arterial diseases

Atherosclerosis Arteries, valves Aortic arch, cerebral arteries
Hypertension Arteries Arteries
Aortic medionecrosis Ostial lesions Arch, cerebral arteries
Aortic valve
Dissecting aneurysm Ostial lesions Arch, cerebral arteries
Aortic valve
Marfans syndrome Aortic valve Arch, cerebral arteries
Mitral valve
Concomitant Manifestations of Disease 289

Table 1 (continued)

Cardiovascular system Nervous system

5. Metabolic diseases
Hypotension-shock Hypoperfusion Hypoperfusion
Hypothermia Arrhythmias Functional reduction
Reduced performance
Diabetes Atherosclerosis Atherosclerosis
Hyponatremia Arrhythmias Edema
Beri-beri Myocardial Neuritis
Hyper-hypothyroidism Myocardial Hyper-hypofunction

6. Drugs
Digitalis Arrhythmias CNS, PNS
Beta-adrenergic blockers Depressant Dreams, depressant
Tricyclic antidepressants Bradyarrhythmias Depressant
Tachyarrhythmias
Phenothiazines Pro arrhythmic effects Stimulant, depressant
Alcohol Depressant Stimulant, depressant

7. Malignancy

tions like meningococcus infections and poliomyelitis commonly affect the heart.
Diphtheria causes acute disease in the heart and nerves, and probably induces chronic
changes in the conduction system of the heart [5]. The congenital rubella syndrome
seen after rubella in pregnancy represents a condition which also links the two organ
systems. The syndrome consists of, among other items, various congenital heart dis-
eases including patent ductus arteriosus and septal defects in conjunction with poor
mental development and specific neurological lesions leading to hearing and vision
disturbances.

Hereditary and Congenital Diseases

In group 2 (Table 1) hereditary and congenital diseases include chromosomal aberra-

tions, conventional and mitochondrial hereditary diseases, and lesions acquired dur-
ing pregnancy. The Kearns Sayres syndrome and the mitochondrial diseases are pre-
sented more specifically in the following chapters of this book (Chaps. 25 and 26).
Myocardial involvement in various muscular dystrophies and in heredoataxias is well
known, although commonly underestimated. Most of these patients should be fol-
lowed regularly, not only from a neurological, but also from a cardiological point of
view.
The most frequent disease in this group, which also should be most easy to pre-
vent, is the fetal alcohol syndrome. As seen in Table 2, both the heart and the brain
are often affected in the advanced form of the syndrome. The nagging question is,
however, how frequent the less advanced forms are and whether there is a threshold
290 K. Rasmussen

Table 2. Fetal alcohol syndrome (n = 41)

Postnatal growth deficiency 97%

Microcephaly 93%
Developmental delay; mental deficiency 89%
Fine-motor dysfunction 80%
Cardiac defects 49%

for the influence of alcohol on fetuses. In addressing such questions both the cardio-
logic and the neurologic frames of reference become too small, and a community ap-
proach is the only one appropriate.

Immunological Diseases

A number of systemic or immunological diseases affect both organ systems (group 3,

Table 1). Most of these conditions attack the vessels, in particular the arteries, and
may thus induce vascular lesions in both brain and heart.
Cerebral affections in systemic lupus erythematosus are dealt with in a separate
chapter (Chap. 27). This disease may attack all layers of the heart including the
endocardium, the myocardium, the coronary arteries, and in particular the peri-
cardium. An interesting feature is its ability to attack across generations, in giving
congenital atrioventricular block to babies of mothers with systemic lupus, due to
lack of fusion of the atrial and ventricular conduction system [2].
The syndrome of arteritis in patients with polymyalgia or temporal arteritis may
be one of the conditions that most frequently attacks both brain and heart. This giant
cell arteritis may cause angina pectoris and myocardial infarction as well as blindness
and a number of cerebrovascular symptoms. Especially when giant cell arteritis takes
the form of Takayashus arteritis, it may give rise to bruits from any side branch of
the aorta. This group of conditions is only one of the good reasons for not omitting
use of the stethoscope, even for neurologists. It is important to auscultate over any
artery, particularly in the axilla, in the posterior supraclavicular fossa, and over the
carotid artery.

Other Arterial Diseases

The most important reason for using the stethoscope more frequently is, however,
arterial disease of a more degenerative kind, like arteriosclerosis and medial necrosis.
All kinds of arterial disease in the aortic arch may certainly strike both the heart and
the brain. Particularly dramatic is of course the dissecting aortic aneurysm, causing
occlusions of several major branches of the aorta. This disease commonly involves
the coronary arteries and the aortic valve; however, cerebral symptoms are fairly
rare (Table 3) [3].
Concomitant Manifestations of Disease 291

Table 3. Cerebral symptoms in dissecting

aortic aneurysm [3]

Total series 527 patients

Syncope 31 patients (6% )
Paraplegia 10 patients (2%)

Metabolic Diseases

Among the many metabolic disturbances (group 5, Table 1) that may affect both
brain and heart, hyponatremia [1] is an important example. This syndrome is not un-
commonly observed in medical intensive care units, often as part of terminal heart
disease and congestive heart failure. Due to the dominance of the cerebral symptoms
the patients may also appear in neurology departments or psychiatric wards. With
progressive hyponatremia brain water increases, especially in the acute form of the
condition. Thus, the symptoms are dependent not only on the degree of hyponatremia,
but also on the speed by which it develops and its cause. The low sodium concentra-
tion may also have some secondary effects on the heart, but these effects tend to be
latent.

Drugs

A number of drugs may affect both the brain and the heart. Virtually all types of
drugs used for heart disease have side-effects on the brain and vice versa. Anti-
epileptic drugs may cause bradycardia. Phenothiazines and in particular thioridazine
may give rise to malignant ventricular arrhythmias, usually ofthe torsade des pointes
type. Tricyclic antidepressants are particularly frightening in their ability to decrease
the conduction capacity of the heart, thus inducing blocks at several levels as well as
tachyarrhythmias.
Another classical example is digitalis intoxication. Most accounts on this condi-
tion focus on the many depressant and stimulant effects on the heart itself. However,

Table 4. Symptoms of 179 patients with

digitoxin intoxication [4]

Symptom %

Fatigue 95
Visual complaints 95
Muscular weakness 82
Psychic complaints 65
Dizziness 59
Dreams 54
Headache 45
292 K. Rasmussen: Concomitant Manifestations of Disease

the drug also has both depressing and stimulant actions upon the brain. This was par-
ticularly well demonstrated by the almost classical experiment on digitalis intoxica-
tion on a community level induced by a Dutch pharmacist manufacturing digoxin
tablets [4]. Instead of digoxin the tablets were filled with equal amounts of digitoxin,
thus increasing the average dose in the patients about 2.5 times. Table 4 indicates the
main symptoms in 179 patients. Remarkably enough the symptoms were to a large
extent mental and neurological, including visual complaints, partly caused by retro-
bulbar neuritis, nausea, vomiting, and psychic disturbances. Twelve patients became
psychotic. Thus, this classification of heart-brain disorders started with one great
imitator, syphilis, and ended with another, digitalis.

Conclnsions

The study of disease patterns as presented above may have fundamental consequences
for the planning of our health care system. Superspecialization, in which we are all
engaged, and narrow career patterns including only one sub speciality may have some
advantages, in particular for specialized research. However, superspecialization has
its serious limits. It may actually prohibit the establishment of new knowledge across
conventional borders. In particular, the service to our patients may improve more
from broadening our clinical view in education and cooperation rather than from dig-
ging ourselves further down in separate trenches. Although this message may be
true, it is not easily delivered, neither to the public nor to the medical community. In
the particular field of "heart and brain" one may suggest the following principles:
1. Cardiologists and neurologists should have a good basic training in each others'
fields.
2. Postgraduate education should aim at maintaining a high level of understanding
of both fields.
3. Practical routines should be implemented to secure clinical cooperation across
department borders.
4. Across-boundary research should be strengthened at the expense of within-bound-
ary research.

References

1. Arieff AJ (1985) Effects of water, acid-base, and electrolyte disorders on the central nervous
system. In: Arieff AJ, DeFronzo RA (eds) Fluid, electrolyte and acid-base disorders. Churchill
Livingstone, Edinburgh, pp 969-1040
2. Chameides L, Truex RC, Vetter V, Rashkind WJ, Galioto FM, Noonan JA (1977) Association
of maternal systemic lupus erythematosus with congenital heart block. N Engl J Med 297: 1204-
1207
3. De Bakey ME, McCollum CM, Crawford ES, Morris GC, Howell J, Noon GD, Lawrie G (1982)
Dissection and dissecting aneurysms of the aorta: twenty-year follow-up of five hundred twenty-
seven patients treated surgically. Surgery 92: 1118-1134
4. Lely AM, Van Enter CMJ (1970) Large-scale digitoxin intoxication. Br Med J 3: 737-740
5. Rasmussen K (1971) Chronic sinoatrial heart block. Am Heart J 81: 38-47
Chapter 25

Cardiac Involvement in Kearns-Sayre Syndrome

B. SCHWARTZKOPFF, G. BREITHARDT, M. BORGGREFE, B. LbsSE, K.-y' TOYKA,
and H.FRENzEL

Introduction

Progressive external ophthalmoplegia (PEO) is accompanied by a progressive weak-

ness of the external muscles of the eyes. It was first described by Albrecht von
Graefe in 1866 [20]. First believed to be of neurogenic origin, it was not until 1951
that Kiloh and Nevin [30] suggested that the ocular signs were due to a myopathy as
they in one case found histological abnormalities in the medial rectus muscle.
PEO may be combined with a wide variety of features including neuromuscular,
central nervous, metabolic, and cardiac symptoms [4, 11, 15,21,36,45,48,52]. The
first case with cardiac involvement was reported by Sandifer in 1946 [53]. He de-
scribed a 29-year-old man with histologically proven ocular myopathy who addition-
ally presented an apical systolic murmur and a bradycardia of 20 beats/min. His
EeG revealed wide QRS complexes, suggestive of bundle-branch block, and no P
waves.
In 1958, Kearns and Sayre [28] reported the clinical triad of PE~, retinal pigmen-
tation, and complete A V block in two young men, one of whom died suddenly at the
age of 17.
The main hints for the pathogenesis of PEO and associated diseases were given
by Olson and co-workers [45] in 1972. They observed abnormal accumulations of
mitochondria as purplish blotches, the so-called ragged-red fibers in the modified
Gomori trichome stain of skeletal muscles of seven patients with PEO and additional
clinical features. Electron-microscopic investigations of the skeletal muscle revealed
structurally abnormal mitochondria. In some of these mitochondria, paracrystalline
inclusions were seen. These findings lead to the diagnosis of a mitochondrial myop-
athy, the pathogenesis of which is still unknown.
Already in 1962, the more general concept of mitochondrial myopathies was in-
troduced by Luft et al. [34]. They described a young woman with hypermetabolism.
The diagnosis of mitochondrial myopathy was based on three main observations: (1)
morphological abnormalities of muscular mitochondria, (2) a biochemical defect
consisting of a disturbed coupling of oxidation and phosphorylation in isolated muscle
mitochondria, and (3) the clinical manifestation that could be explained by mito-
chondrial dysfunction.
Several mitochondrial myopathies were described in the following years, mostly
based on morphological abnormalities of the mitochondria [15, 19, 25, 36, 45, 51,
57]. Up to now, the biochemical defect could not be disclosed in most cases. Of the
294 B. Schwartzkopff et al.

Table 1. Mitochondrial myopathies that are combined with the occurrence of ragged-red fibers in
skeletal muscle biopsies

MELAS: Mitochondrial myopathy, encephalopathy, lactic acidosis, stroke-like episodes

MERRF: Myoclonus epilepsy associated with ragged-red fibers
KSS: Kearns-Sayre syndrome: ophthalmoplegia, pigmentary degeneration of the retina,
cardiomyopathy

numerous reports, three distinct syndromes are apparent, which are all combined
with the occurrence of ragged-red fibers in the skeletal muscle biopsy [36, 51] (Table 1).
The Kearns-Sayre Syndrome (KSS) is a rare disease, which has been reported in
the literature in less than 100 cases [52]. It is characterized by PEa and atypical pig-
mentary degeneration of the retina before the age of 20; heart block may be delayed
for many years. It is therefore regarded by some authors as a facultative symptom
[28, 57]. Incomplete forms as well as progression to complete forms have been de-
scribed [5, 18,49]. KSS is regarded to be sporadic [35, 51]. There has only been one
report of siblings who were affected with the complete syndrome [54].
The second syndrome, myoclonus epilepsy with ragged-red fibers (MERRF) was
described by Fukuhara et al. in 1980 [19]. The main clinical symptoms are myo-
clonus, ataxia, weakness, and seizures. All three cardinal symptoms of KSS are mis-
sing. The frequent familial occurrence of MERRF suggests non-mendelian maternal
inheritance [19, 52].
Mitochondrial myopathy, encephalopathy, lactic acidosis, and stroke-like epi-
sodes represent the third syndrome, called MELAS. It is observed in children
beyond the 1st year of life. The strokelike events and the episodic vomiting distin-
guishes MELAS from KSS and MERRF. PEa, retinal degeneration, and heart
block are typically not observed in MELAS [36, 46]. It is suspected that this syn-
drome is hereditary and may be maternally transmitted [46].
As cardiac involvement is one of the main features of patients with KSS (but not
of MERRF and MELAS) and as prognosis is mainly determined by the insidious oc-
currence of potentially fatal total A V block, cardiologic investigations seem to be im-
portant.

Clinical Presentation of Keams-Sayre Syndrome (KSS)

Berenberg et al. [5], reviewing the literature in 1977, found 30 cases of KSS and re-
ported on five patients of their own. They stated that 63% of these patients had seri-
ous cardiac involvement such as Stokes-Adams attacks, cardiac arrest, heart failure,
or pacemakers. Both sexes are equally affected [5, 51].
In a recent review by Rowland et al. [51],70 patients including those of Beren-
berg et al. [5] were described meeting the criteria of PEa, atypical pigmentary de-
generation of the retina, and onset of disease before the age of 20. Heart block was
not included in this review because it may be delayed for many years. Thus, many
Cardiac Involvement in Kearns-Sayre Syndrome 295

Kearns - Satre Stndrome

n=7 pts.
Age
Patient • at onset
• • • at study

2 • •
• •
• •
...
4

5
• •
6

• •
I I
10 20 30 40 50 Age
Fig.t. Age of seven patients with Keams-Sayre syndrome at the time of onset of the disease and at
the time of the study

cases represented incomplete forms regarding cardiac involvement. Besides the

typical clinical triad of KSS, several other symptoms were reported. Main additional
criteria that support the diagnosis are, besides the onset of the disease before the age
of 20, a cerebellar syndrome and increased cerebrospinal fluid proteins (> 100mgl
dl). In 95% of patients, ragged-red fibers were observed in peripheral muscles [51].
Increased cerebrospinal fluid proteins (> 100 mg/dl) were present in 88%, a cerebel-
lar syndrome in 80%, a short stature in 67%, a loss of hearing in 61 %, weakness of
the limbs in 60%, and mental retardation in 49% [5,51]. Hypocalcemia, hyperphos-
phatemia, and increased alkaline phosphatase activity because of associated hypo-
parathyroidism have been reported in KSS [22, 47].
We have reported an additional six patients [56], and observed another case very
recently. Family history did not reveal any neuromuscular disease in other family
members of these seven patients. The mean age of the four men and three women
was 34 years. The onset of disease was at the mean age of 17 years. All patients pre-
sented first with PEa (Fig. 1). Biopsy of the limb muscles disclosed ragged-red fibers
in all patients. Only three patients had a retinal pigmentation, which was first diag-
nosed at a mean age of 17 years. According to Rowland et al. [51], the complete ex-
pression of the syndrome has to be assumed in these three patients. Berenberg et al.
[5] pointed out that 5-20 years after the occurrence of PEa, the manifestation of
atypical pigmentation of the retina was seen in some patients. Therefore, incomplete
forms of KSS are probably present in the other four patients. Limb muscle weakness
occurred in six patients, cerebellar ataxia in three, hearing loss in two, and defective
vision in two. One patient was mentally retarded. Only one of our patients had com-
plete A V block and reported syncopes, but there were no clinical signs of heart fail-
ure in any patient.
From the data of the literature and our own findings (Table 2), it is apparent that
a heterogeneous clinical picture exists in patients with KSS. Cardiac manifestation
has still received little attention, and only little insight has been gained up to now.
296 B. Schwartzkopff et al.

Table 2. Clinical presentation of seven of our own patients with Kearns-Sayre syndrome

Symptoms Patient
1 2 3 4 5 6 7
Ptosis x x x x x x x
CPEO x x x x x x x
Retinitis pigmentosa x x x
Card. conduct. defects x x x x x
Prox. myopathy x x x x x x
Weakness of vision x x x
Cerebellar ataxia x x x
Deafness x x
Short stature x x

Cardiac Involvement in KSS

Electrocardiographic Findings (Table 3)

Atrioventricular (AV) and intraventricular conduction defects are the typical find-
ings in KSS. Electrocardiographic changes were noticed in most patients 9 years after
the first symptoms of KSS [5].
Mter reviewing 51 patients from the literature and reporting one additional case,
Fauchier et al. [18] reported that 36 patients (71%) had total A V block. In 25 of
these cases, intraventricular conduction defects were present before the occurrence
of complete heart block: 3 patients showed left anterior fascicular block, 5 showed
complete right bundle-branch block, and 17 patients a combination of both forms of
block (bifascicular block). In half of the 16 patients without complete AV block,
right bundle-branch block combined with left anterior fascicular block was present,
whereas a first- or second-degree A V block was present in only two other patients.
In the remaining patients, a right or left bundle-branch block was observed. In our
seven patients, complete A V block occurred in one patient who subsequently re-
ceived a permanent pacemaker. Left anterior fascicular block combined with right
bundle-branch block was present in another one and incomplete right bundle-branch
block in three of seven patients.
Up to now, it is unknown which patient with KSS is at risk of developing a poten-
tially fatal A V block. The progression of intraventricular conduction defects to com-
plete A V block appears to be more frequent and to occur at a younger age in patients
with KSS than in patients with similar electrocardiographic defects of other
etiologies. In the latter group of patients with combined left axis deviation and right
bundle-branch block, the risk of progression to complete A V block is less than 6%
per year [32]. Therefore, in these patients a bifascicular block with left axis deviation
is not sufficient indication for permanent prophylactic pacing.
In contrast, the risk in KSS is markedly higher. After reviewing the literature,
Fauchier et al. [18] emphazised that 10 of 29 mainly young patients with KSS without
Cardiac Involvement in Keams-Sayre Syndrome 297

Table 3. Electrocardiographic findings in seven of our own patients with

the Keams-Sayre syndrome

QRSaxis LAFB + RBBB (n = 1) AV IIJO (n = 1)

incompl. RBBB (n = 3)
normal axis (n = 2)
Mean Abnormal

QRS 88± 18ms 117

AH 85 ± 16ms 0/6
HV 47± Urns 2/6 + AV IIIo (distal)
Block distal H(180Imin) (n = 1)
Wenckebach 183 ± 14/min 0/6
ERP(A) 190 ± 18ms 0/6
ERP(AVN) 252±73ms 0/6

ERP (A), effective refractory period of the atrium; ERP(AVN), effective

refractory period of the AV node; Wenckebach, rate at which Wencke-
bach-like block occurred in the A V node; AH, intranodal conduction
time during direct catheter recording from the A V conduction system;
HV, His-Purkinje system to ventricular muscle conduction time

pacemaker died suddenly, presumably of complete heart block, whereas of 23 pa-

tients treated with a permanent pacemaker, only 3 died (one from cachexia, one
from gastrointestinal bleeding, and another from cardiac failure). Thus, prophylactic
pacing may improve prognosis of patients with KSS. As pointed out by Berenberg et
al. [5], the progression to complete AV block may take only a few days up to several
years.
Therefore, regular long-term EeG recordings may be helpful to disclose inter-
mittently occurring AV block. Rheuban et al. [49] reported such a case of a 19-year-
old white female in whom episodes of near-fatal complete A V block were observed
during continuous long-term Holter EeG recording.
In our patients, no A V block and no malignant ventricular arrhythmias occurred
in several 24-b long-term EeG recordings. Exercise EeG did not disclose A V con-
duction defects in any of our patients.
These findings underscore the need for further investigations to establish the
indication and the timing for prophylactic pacemaker implantation. Therefore, elec-
trophysiological studies seem to be mandatory to evaluate the potential role of more
sensitive means to detect conduction defects in KSS.

Electrophysiological Findings

Already in 1972, Morris et al. [40] reported a prolonged HV interval during His
bundle electrography in a 16-year-old girl with second-degree AV block in KSS.
Our group has performed electrophysiological studies in six patients (three men,
three women) with KSS [56]. Sinus node function, the effective refractory periods of
the atrium (190 ± 18ms) and the AV node (252±73ms) as well as atrionodal con-
29S B. Schwartzkopff et al.

1000
I I I

a Kearns - Sayre Syndrome 901 k

t I I J II j ~ 'I' 1 sf1000
I , I I 1 1 I, i't.I 1 I 1 I

I
~ ' r
I.jlll
-t I· f
~
II"

J
I , • ,
II :\,l ~\ ~ ~'--I". I
1 TV ' lJ I lj
V1 1 i 1 : . -f' 00L-l"~
II
I
l'll
~ I , ~, . . . '-1

RVA I
~'I
: ! J .;I I 1 ~ ~V~h"-"1:.I:L
, r'
I I, 1\ I
I I , :1 , I' , IT '

H~~~~~~~t
HRA2
{ I I
I ,.
. I I ,
I · I
I I
~
I

51-51 =330ms H1246

Kearns - Sayre Syndrome
. . __ .. -- --1 - - - " . . _._ . .. -
900k

Fig. 2a, b. Original recordings in a case with Keams-Sayre syndrome showing, from top to bottom,
time lead (t), leads I, II, and Vi and intracardiac recordings RVA (electrogram from the apex of the
right ventricle) , HBE (electrogram from the His bundle region), HRA), and HRA2 (two electro-
grams from the high right atrium) . a At an atrial pacing cycle length of 120/min (Sl-Sl = 500ms), an
alternating pattern of intraventricular conduction was observed with QRS duration changing be-
tween 110 and 140ms. b Increasing the rate of atrial pacing to about ISO/min (Sl-Sl = 330ms), a
block distal to the His bundle spike occurred in an alternating fashion
Cardiac Involvement in Keams-Sayre Syndrome 299

Kearns - Sayre Syndrome

n=6!7 pts.

Response to ajmaline
HV
(ms)

80

70
60

50

40

30

20
Fig. 3. Response of HV time after
10
intravenous administration of ajmaline in
six of our seven patients
control ajmaline
(lmg/kg iv)

0--0 block distal His at 180 and

160/min (resp.)

duction time (A-H intervals; 85 ± 16ms) were normal in all patients. These findings
are in agreement with the few reports in the literature [18, 50].
Mean QRS duration was 88 ± 18ms; it was abnormal (120ms) in one case. Mean
HV time was 47 ± 11 ms. It was abnormal (> 55 ms) in two patients of whom one had
right bundle-branch block combined with left anterior fascicular block, the other one
incomplete right bundle-branch block.
During incremental atrial pacing, Wenckebach-type block occurred in the AV
node at a rate of 183 ± 14 beats/min in five of our six patients. In one case with
preexisting left anterior fascicular block and right bundle-branch block, a 2: 1 block
distal to the His bundle occurred at an atrial pacing rate of 180 beats/min (Fig. 2).
After administration of ajmaline (Img/kg body weight i.v.), there was a mean in-
crease in the HV interval of 44% in all our patients (Fig. 3). In the case in whom
there was a block distal to the His bundle at a pacing rate of 180 beats/min during
control, it occurred already at 160 beats/min after administration of ajmaline. This
patient received a pacemaker [56].
Fauchier et al. [18] reported the results of electrophysiological investigations in
nine patients, eight cases from the literature and one observation of their own. In
seven patients, a pathologically prolonged HV interval was reported (65-120ms).
Six of these had right bundle-branch block combined with left anterior fascicular
block and one had complete left bundle-branch block. In one patient with complete
A V block, the defect was located distal to the His bundle. In only one case was the
HV interval normal; this patient also had a left anterior fascicular block and right
bundle-branch block.
The greater preponderance of abnormal conduction distal to the His bundle in
the review by Fauchier et al. [18] is probably due to some selection factor as most
300 B. Schwartzkopff et al.

studies were single case reports [9, 38, 40, 50] that tend to report only the markedly
abnormal cases. In contrast, the group of patients studied in our department probably
does not represent a sirniliar bias as these were consecutive patients.
The prognostic significance of a prolonged HV interval in the setting of KSS has
not yet been settled. With regard to the high incidence of total A V block in KSS, its
prognostic significance cannot be compared with other patients with intraventricular
conduction disturbances without KSS. The largest experience has been reported by
Dhingra et al. [13] who performed electrophysiological studies in 496 patients with
chronic bifascicular block. In 15 patients with various cardiac disorders, a block dis-
tal to the His bundle during atrial pacing was noted during intact intranodal conduc-
tion. During a mean follow-up of 3.4 ± 0.59 years, seven of these patients developed
a high-degree A V block. One had treadmill-provoked A V block, and two other pa-
tients died suddenly. In 6 of the 496 patients, a block distal to the His bundle occur-
red during atrial pacing-induced A V nodal Wenckebach periods. None of these pa-
tients developed A V block during follow-up.
Further information on patients with intraventricular conduction disturbances is
available fr<?m an interventional study by Dini et al. [14). 85 patients with chronic
bundle-branch blocks of different etiology were studied using electrophysiological
techniques. Of the patients 33% had organic heart disease, 22% had coronary heart
disease, and 45% had primary conduction defects. This prospective study showed a
prognostic value of the acute i.v. administration of ajmaline (ajmaline test) in those
cases in whom a second- or third-degree A V block distal to the His bundle was in-
duced during atrial pacing. The mere increase in HV interval by ajmaline was of no
prognostic significance. Fifteen patients developed a documented second- or third-
degree A V block during the following 2 years. The accuracy of the ajmaline test for
the prediction of advanced A V block was 73%, the sensitivity 59%. Whether the
ajmaline test has the same value in patients with KSS needs further evaluation.
The occurrence of ventricular tachycardia has less frequently been reported in
KSS [31,49). In one young woman, ventricular tachycardia, which required cardio-
version, followed an episode of bradycardia and hypotension. She received a pace-
maker because of A V block but died suddenly a few months later. Autopsy did not
reveal cardiac abnormalities [49]. Kleber et al. [31] reported a 26-year-old male pa-
tient with KSS who was treated with arniodarone because of nonsustained ventricu-
lar tachycardia. Additionally, he received prophylactically a pacemaker because of
first-degree A V block. In our patients, no spontaneous malignant ventricular ar-
rhythmias were seen during long-term BeG recording. Programmed right ventricu-
lar stimulation was carried out in all seven patients. One patient revealed ventricular
flutter, which was induced by two premature stimuli applied at a basic drive cycle
length of 370 ms. The significance of this finding remains obscure as this patient had
no prior documentation of ventricular tachycardia and had been asymptomatic.
Thus, presently available information suggests a potential role of electrophysio-
logical studies in patients with KSS to evaluate the severity and location of conduc-
tion defects and to elucidate a potential propensity to malignant ventricular ar-
rhythmias. Further studies are required to establish the final pathomechanisms lead-
ing to the high incidence of sudden death in these patients. As long as the results of
these studies are not available, only preliminary guidelines based on present knowl-
edge can be given. Since patients with KSS are young and the risk of advanced A V
Cardiac Involvement in Keams-Sayre Syndrome 301

block is obviously high, implantation of a pacemaker should be considered early in

any patient with KSS and intraventricular conduction disturbances. The role of in-
vasive electrophysiological testing of the conduction system with the use of, e.g.,
ajmaline to better predict the occurrence of total A V block in these patients, remains
unsettled, but prophylactic pacemaker implantation should be considered also in
those who develop a block distal to the His bundle, e.g., during atrial pacing before
or after ajmaline administration. Prophylactic pacemaker implantation may also be
considered in those with markedly prolonged HV interval. At least these patients
and those with normal electrophysiological findings should undergo electrophysio-
logical monitoring at regular intervals. Finally, there is no doubt that patients with
spontaneous second- or third-degree A V block should receive a pacemaker.

Myocardial Function

For several years, an impairment of myocardial function had not been considered as
a feature of cardiac involvement in KSS [50].
There have been only a few reports on hemodynamic studies in KSS. The first re-
port was in 1973 when Uppal [58] described a 19-year-old man with right bundle-
branch block and intermittent second-degree AV block, PE~, and retinal pigmen-
tary changes who was examined because of Stokes-Adams attacks. Cardiac catheteri-
zation revealed normal intracardiac and intravascular pressures and a normal cardiac
index of 3.61/m2 • Clark et al. [9] found normal hemodynamic parameters in a 13-
year-old boy with complete A V block. Further studies supported these findings in
other young patients with KSS [8, 38]. Berenberg et al. [5], when reviewing the liter-
ature in 1977, found reports on clinical signs of cardiac failure in 4 of 35 patients.
It was unclear whether chronic volume overload due to total A V block or cardio-
myopathy was the main pathogenetic cause.

Kearns - Sayre Syndrome

mean PA n=6/7 pts.
pressure
(50-100 W)

40

30 Fig.4. Increase in mean pulmonary

artery pressure during supine bicycle
ergometer test in six of our seven
20 patients with the Keams-Sayre syn-
drome. In two cases, mean pulmo-
nary artery pressure increased be-
yond the upper limit of normal
10 (30mmHg in our department)
which, in the absence of pulmonary
disease, is interpreted as an indica-
tion of left ventricular dysfunction
rest exercise
302 B.Schwartzkopff et al.

Since these studies had been done only at rest, we also included exercise testing
in our study protocol for patients with KSS to detect latent forms of left ventricular
impairment. All six patients had normal hemodynamic parameters at rest [56]. How-
ever, during exercise, impaired myocardial function with a pathological increase in
mean pulmonary artery pressure (> 30 mmHg) and in wedge pressure (> 20 mmHg)
became obvious in two patients (Fig. 4). In one of these patients there was a decrease
in ejection fraction as determined by radionucleotide angiography. These findings
are consistent with a latent cardiomyopathy [35].
Recently, morphological, clinical, and biochemical findings were reported in a
26-year-old patient with dilative cardiomyopathy who died of congestive heart fail-
ure 2 years after the implantation of a pacemaker [24, 31, 41]. These findings suggest
that not only the conduction tissue but also the ordinary myocardium are involved in
KSS. The extent of this involvement and the time of clinical manifestation are still
unclear. It is well possible that by prevention of total AV block (and, thus, possibly
of sudden death), left ventricular dysfunction might be the leading manifestation of
the end stage of the disease.

Morphological Findings in KSS (Table 4)

In the skeletal muscle of patients with KSS, ragged-red fibers were typically found in
the modified trichrome stain [48], disclosing abnormal mitochondria. In skeletal
muscle, 5%-20% of fibers were affected [2, 45, 48]. Electron microscopy reveals dif-
ferent changes of mitochondria. They may be normal in size and structure, but in-
creased in number; some are enlarged, appearing in clusters or being scattered among
other organelles. The cristae may be increased in number, irregularly oriented, or
they may form concentric whorls; conversely, there may be vacuolated mitochondria
with a loss of cristae. Some mitochondria contain inclusions such as paracrystalline
structures or globular bodies [7,27,45,48, 52].
Morphological findings of the myocardium in KSS are contradictory. Kearns [29]
described at autopsy a 17-year-old boy with PEO, pigmentary degeneration of the
retina, and complete heart block who died during a syncopal attack. He found left

Table 4. Morphological findings in patients with the Kearns-Sayre

syndrome

Heart Skeletal muscle

Reduction of myofibrils + +
Mitochondria
Paracrystalline inclusions 0 +
Concentric cristae + +
Loop-like cristae + +
Vacuolization + +
Dense bodies + +
Cardiac Involvement in Keams-Sayre Syndrome 303

ventricular hypertrophy and hyperchromatic enlarged heart muscle nuclei. Other

authors macroscopically described hearts of normal size and weight. Light micro-
scopically, there were no changes of the myocardium [10, 26, 49, 50). Clark et al. [9]
investigated the cardiac conduction system of a 13-year-old boy who died from sud-
den cardiopulmonary arrest. They described extensive fibrosis, mainly in the distal
portion of the His bundle in a normally sized heart. Fibrosis of the cardiac conduc-
tive tissue was also found by other authors [7].
Recently, in a 26-year-old patient with drug-refractory congestive heart failure,
autopsy revealed an enlarged and dilated left ventricle [24]. Electron microscopy of
specimens taken 3 h after death showed increased numbers of mitochondria and a re-
duction in myofibrils in some myocytes.
Up to now, there have been only a few reports about morphological findings in
endomyocardial biopsies of patients with KSS. McComish et al. [38] reported hyper-
trophied heart muscle cells and focally increased numbers of structurally normal
mitochondria in the endomyocardial biopsy of a 32-year-old man with PEa, atypical
pigmentation of the retina, and first-degree A V block. In a 3D-year-old man with
complete A V block in KSS, Charles et al. [8] described hypertrophied heart muscle
cells with proliferation of mitochondria, accumulation of glycogen, and occasional
giant mitochondria, some of them being oval or spherical with numerous long closely
packed cristae arranged in parallels. Bastiaensen et al. [4] also reported abnormal
cardiac mitochondria in a 62-year-old man with PEa, retinal pigmentation, and in-
complete right bundle-branch block, but no detailed description or pictures were
presented. Denis et al. [12] mentioned in an addendum to their paper that they
found concentric cristae mitochondriales in the endomyocardial biopsy of a 38-year-
old man with complete A V block.
We observed a mainly normal myocardium in the right ventricular endomyo-
cardial biopsies of six patients by light microscopy. There were no signs of acute
myocarditis, the endocardium was not thickened, and there was no muscle fiber
hypertrophy. In only two patients was fibrous tissue slightly increased to more than
4% (normal value in our laboratory < 4%), which could be a hint for a loss of fibers
with replacement fibrosis.
Electron microscopy revealed in our patients that the majority of myocardial
fibers were composed of normally structured organelles. Increased numbers of
mitochondria with structurally normal cristae could be observed in some myocytes of
any patient. This mitochondriosis with a variation in form and size of mitochondria
was seen in the subsarcolemma as well as between the myofibrils. In five of the six
patients studied, only 1% -2% of the myocytes were exclusively composed of mito-
chondria with structural abnormalities [55). Three main types can be characterized:
rare hugh mitochondria with concentric cristae, large ones with looplike or with
parallel cristae, and small vacuolized ones (Fig. 5) [55]. The myofibrils were reduced
in these myocytes. Paracrystalline structures were not observed.
Our findings are consistent with a mitochondrial cardiomyopathy. Similar mito-
chondrial abnormalities as observed in the heart muscle cells of our patients have
been found in the cerebellum, sweat glands, and liver in other patients with KSS,
supporting the thesis of a generalized mitochondrial cytopathy [16, 27). Whether the
conductive tissue is more afflicted than the myocardium cannot be answered by our
findings as no selective biopsy of the conduction system is possible.
304 B. Schwartzkopff et al.

Fig.5. Electron-microscopic finding of a heart muscle cell with mitochondria of various size and
shape, composed of cristae with concentric and looplike arrangement, x 20000

Mitochondrial cardiomyopathy of unknown origin was found in children [3, 6, 23,

33,37,42], presenting with the picture of hypertrophic cardiomyopathy [6, 23, 25,
33,37] or of dilative cardiomyopathy [3,42], but without symptoms of KSS.
Hug and Schubert [25] reported structural abnormalities of cardiac mitochondria,
similar to those we found, in left ventricular biopsy of a 6-month-old boy with
marked cardiomegaly and moderate hepatomegaly but without involvement of the
skeletal muscles. Many heart muscle cells were affected but there were no signs of
congestive heart failure. The picture of hypertrophic cardiomyopathy with mito-
chondrial abnormalities was also found in Leigh's disease by Lange et al. [33]. Mackay
et al. [37] reported an ll-year-old boy with proximal skeletal myopathy and hyper-
trophic cardiomyopathy in whom right ventricular biopsy showed a major increase in
mitochondrial size and number with formation of unique ring-shaped mitochondria
on electron-microscopic morphometry.
An X-linked recessive dilative cardiomyopathy was described by Neustein et al.
[42] in a 4-month-old boy in whom endomyocardial biopsy disclosed vacuolized mito-
chondria and numerous mitochondria with concentric cristae.
Askanas et al. [2] could show that similar morphological abnormalities were
found in the skeletal muscle biopsy of patients with ragged-red fibers as well as in
cultures of normal human muscle treated with 2,4-dinitrophenol. These findings ex-
plain that mitochondrial abnormalities both in number as well as in size and structure
can be found not only in KSS but also in various other pathological conditions.
Cardiac Involvement in Keams-Sayre Syndrome 305

As morphological changes of mitochondria are typical for mitochondriopathy but

unspecific for the underlying biochemical defect [2, 37, 48], it might be assumed that
different biochemical defects may lead to various clinical expressions and different
progression under the same morphological picture. As paracrystalline inclusions
were not observed in the endomyocardial biopsies of our patients but were present
in the skeletal muscle, mitochondrial heterogeneity of different organs or a different
biochemical defect may be present.
Whether the number of affected myocytes with structurally abnormal mitochondria
is of importance for myocardial function is still unresolved. The rare finding of ab-
normal fibers in our patients having normal hemodynamic parameters at rest seems
to support this hypothesis. The imparied myocardial function during exercise in two
of our patients demands more affected myocardium. Whether more left than right
ventricular myocytes are involved or whether morphologically still normal myocytes
have already impaired biochemical function during exercise is still unclear.
The observed mitochondriosis in some still structurally normal myocytes may be
a hint for a disturbed biochemical function of the individual mitochondrion, partially
being compensated by increased numbers. To clarify these questions, biochemical
investigations are necessary to evaluate the underlying defect and to assess potential
interventions.

Biochemical Findings in KSS

An important, but unspecific finding in mitochondrial myopathies is an increase in

the blood concentration of lactic acid which may vary in severity in different patients
[34, 36, 39]. Lactic acidosis is explained by impaired utilization of pyruvate in the
Krebs cycle and by increased glycogen utilization and glycolytic activity in muscle
with defective aerobic metabolism [2, 34, 36]. The increased pyruvate concentration
is reduced to lactate by lactate dehydrogenase or it is transaminated by alanine trans-
aminase. During exercise, lactic acidosis becomes evident in some patients with nor-
mal values at rest. In addition, pyruvate, lactate, as well as alanine may be increased
in the blood of patients with KSS.
Allen et al. [1] found a reduction in muscle carnitine and in mitochondrial en-
zymes in a 25-year-old woman with KSS. She had decreased plasma and cerebrospi-
nal fluid folate levels as well. There was a prompt plasma folate response in this pa-
tient but cerebrospinal fluid folate was not significantly affected by folic acid treat-
ment. Nevertheless, clinical improvement was evident, and there was a marked in-
crease in muscle free carnitine. These findings were interpreted as unspecific, but the
basic disorder of mitochondrial enzyme function of unknown cause might have been
influenced by pharmacological amounts of pterins such as folic acid.
In an ll-year-old girl, a palmitoyl-CoA synthetase deficiency was described, an
enzyme that is located in the outer mitochondrial membrane and that acts mainly on
the mitochondrial transport of long-chain fatty acids [43].
There are interesting findings regarding the mitochondrial electron transport sys-
tem in KSS. The electron transport system in the mitochondrial inner membrane is
composed of four enzyme complexes as follows: NADH - coenzyme-Q-reductase
(complex I), succinate - coenzyme-Q-reductase (complex II), coenzyme QH2 -
306 B. Schwartzkopff et al.

cytochrome-c-reductase (complex III), and cytochrome-c-oxidase (complex IV) [17].

Milller-Hocker et al. [41] found a deficiency of cytochrome-c-oxidase (complex IV)
in the myocardium as well as in the skeletal muscle of a 26-year-old man with KSS at
autopsy. The cardiac myocytes were affected in a disseminated manner. The defect
was present in the contractile and conductive myocardial fibers as well. Also struc-
turally normal myocytes showed decreased enzymes. The enzyme protein, being en-
coded on mitochondrial DNA and synthesized in the mitochondria, is an integral
part of the inner mitochondrial membrane. These findings suggest that a defect of
mitochondrial DNA may be the primary cause of the biochemical disturbance.
Decreased activity in the electron transport system of the mitochondrial inner
membrane was also found by Yorifugi et al. [59] in the skeletal muscle of five pa-
tients with KSS. Mitochondrial contents were increased, and all enzyme activity per
mitochondrial protein of the electron transport system was markedly decreased.
The authors concluded that the decreased enzyme activity of the mitochondrial elec-
tron transport system in each mitochondrion may result in a compensatory increase
in mitochondrial contents in the muscle of KSS. Morphologically observed mito-
chondriosis in our patients may be explained by these findings.
Ogasahara et al. [44] reported on biochemical studies in five patients with KSS.
The mitochondrial fraction was increased in muscles from KSS patients, and co-
enzyme 010 (CoO) was slightly decreased. Administration of 120-150mg/d of CoO
improved the abnormal metabolism of pyruvate and NADH oxidation in skeletal
muscle. CoO therapy decreased cerebrospinal fluid protein concentration and im-
proved neurological symptoms. In one patient, the PO interval shortened from 0.28
to 0.16 s within 3 months; after 1 year of therapy it was again 0.28 s. Complete left
bundle-branch block did not change after 4-7 months of therapy in two patients. In
one of three patients ST depression, seen before therapy in all of them, ameliorated.
Long-term studies should clarify if administration of coenzyme 010 is a reliable
therapeutic concept for patients with KSS. There are still many unresolved ques-
tions. The main pathogenetic cause of mitochondrial dysfunction in KSS remains still
unclear. A spontaneous defect of mitochondrial DNA or inheritance in some pa-
tients as well as a persistent viral infection are to be discussed as the origin of the
metabolic disorder in KSS.

Clinical Considerations and Conclusions

Kearns-Sayre syndrome belongs to the group of mitochondrial myopathies of which

the pathogenic mechanisms and the biochemical defects are still unknown. Morpho-
logical studies show that the myocardium is also affected by a mitochondrial cytopathy.
Besides neuromuscular symptoms, cardiac involvement is frequent, mainly evi-
denced by A V and intraventricular conduction defects. Electrophysiological studies
are helpful in obtaining more insight into the defect of the conduction system and
may also prove to be helpful in determining the risk of developing complete A V
block. With regard to the insidious occurrence of A V and intraventricular conduc-
tion defects, pacemaker implantation should be considered early during the course
of the disease as soon as there is either a documentation of second- or third-degree
Cardiac Involvement in Keams-Sayre Syndrome 307

Management of patients with Kearns-Sayre Syndrome

clinical presentation
/' \
EKG AV 11°/111° normal or IV block
I ~
HVt or block normal

l
EPS distal H
(ajmaline ?)

Biopsy
I normal
+/-
I
Therapy PM proph. PM FU
Fig. 6. Management of patients with the Kearns-Sayre syndrome depending on the presence or ab-
sence of A V block, the findings at electrophysiological study (EPS), and the presence or absence of
abnormal electron-microscopic findings with regard to pacemaker implantation (PM). FU, follow-
up visits without pacemaker implantation

A V block or even earlier in the presence of a markedly prolonged HV time during

an electrophysiological study (Fig. 5). The potential role of acute administration of
antiarrhythmic drugs during electrophysiological studies as a stress test needs further
evaluation. In addition, the course of the disease might be changed in those who re-
ceive a pacemaker by avoiding arrhythmic death but leaving time for the develop-
ment of heart failure. Endomyocardial biopsy may be of great interest to assure the
diagnosis of a more widespread cardiac involvement. Further studies have to clarify
whether the extent of affected myocytes is of prognostic value regarding myocardial
function and the potential development of cardiac failure. For early detection of car-
diac failure not only in those patients without but also in those with a pacemaker,
regular cardiological controls are advisable. The role of medical therapy, intended to
support mitochondrial metabolism, is still under investigation. Clinical and biochem-
ical studies are necessary to solve this question. Long-term follow-up of patients with
KSS seems to be necessary to get more insight into the spontaneous course of the dis-
ease as well as in the course of those patients receiving treatment.

Acknowledgement. This work was supported by a grant from the Sonderforschungsbereich 242 of the
Deutsche Forschungsgemeinschaft, Bonn-Bad Godesberg, FRG.

References

1. Allen RJ, DiMauro S, Coulter DL, Papadimitriou A, Rothenberg S (1983) Kearns-Sayre syn-
drome with reduced plasma and cerebrospinal fluid folate. Ann Neurol13: 679-682
2. Askanas V, Engel WK, Britton DE, Adornato BT, Eiben RM (1978) Reincarnation in cultured
muscle of mitochondrial abnormalities. Two patients with epilepsy and lactic acidosis. Arch
NeuroI35:801-809
3. Barth PG, Scholte HR, Berden JA, van der Kiei 1M, Luyt-Houven IEM, van't veer-Korthof
ET, van der Horten JJ, Sobotka-Plojhar MA (1983) An X-linked mitochondrial disease affect-
ing cardiac muscle, skeletal muscle and neutrophil leukocytes. J Neurol Sci 62:327-355
308 B. Schwartzkopff et al.

4. Bastiaensen LAK, Joosten EMG, Rooij JAM, Hommes OR, Stadhouders AM, Jaspar HHJ,
Veerkamp JR, Bockelman H, van Heinsbergh VWM (1978) Ophthalmoplegia-plus, a real noso-
logical entity. Acta Neurol Scand 58:9-34
5. Berenberg RA, Pellock JM, DiMauro S, Schotland DL, Bonilla E, Eastwood A, Hays A,
Viscale cr, Behrens M, Chutorian A, Rowland LP (1977) Lumping or splitting? "Ophthalmo-
plegia-plus" or Keams-Sayre syndrome? Ann Neurol1 :37-54
6. Bogousslavsky J, Perentes E, Deruaz JP, Regli F (1982) Mitochondrial myopathy and cardio-
myopathy with neurodegenerative features and multiple brain infarcts. J Neurol Sci 55: 351-357
7. Castaigne P, Lhermitte F, Escourolle R, Chain F, Fardeau M, Hauw JJ, Curet J, Flavigny C
(1977) Etude anatomo-clinique d'une observation d'''ophthalmoplegia plus" avec analyse des
lesions musculaires, nerveuses centrales, oculaires, myocardiaques et thyroidiennes. Rev Neurol
133:368-369
8. Charles R, Holt S, Kay 1M, Epstein EJ, Russell RJ (1981) Myocardial ultrastructure and the
development of atrioventricular block in Keams-Sayre syndrome. Circulation 63: 214-219
9. Clark DS, Myerburg RJ, Morales RR, Befeler B, Hernandez FA, Gelband H (1975) Heartblock
in Keams-Sayre syndrome, electrophysiologic-pathologic correlation. Chest 68:727-730
10. Coulter DC, Allen RJ (1981) Abrupt neurological deterioration in children with Keams-Sayre
syndrome. Arch NeuroI38:247-250
11. Daroff RB, Solitare GB, Pincus JH, Glaser GH (1966) Spongiform encephalopathy with chronic
progressive external ophthalmoplegia. Central ophthalmoplegia mimicking ocular myopathy.
Neurology 16: 161-170
12. Denis B, Morena H, Rossignol B, Machecourt J, Sebag M, Stoebner P, Martin-Noel P (1976)
Association d'une ophtalmoplegie externe progressive et d'une myocardiopathie avec bloc
auriculoventriculaire complet. Arch Mal Coeur 69 :747-752
13. Dhingra RC, Wyndham CR, Bauernfeind R, Swiryn S, Deedwania PC, Smith T, Denes P,
Rosen KM (1979) Significance of block distal to His bundle induced by atrial pacing in patients
with chronic bifascicular block. Circulation 60 : 1455-1464
14. Dini P, Ialongo D, Adinolfi E, Alboni P, Malavarne C, Lotto A, Finzi A, Pagnoni F, Baldi N,
Morgera T, Gronda M, Rognoni G, Occhetta E, D' Aulerio M (1982) Prognostic value of His-
ventricular conduction after Ajmalin administration. In: Masoni GA, Alboni P (eds) Cardiac
electrophysiology today. Academic, London, pp 515-522
15. Drachman DA (1982) Ophthalmoplegia plus. The neurodegenerative disorders associated with
progressive external ophthalmoplegia. Arch Neurol18: 654-681
16. Egger J, Lake BD, Wilson J (1981) Mitochondrial cytopathy. A multisystem disorder with rag-
ged red fibers on muscle biopsy. Arch Dis Child 56:741-748
17. Estrabrook RW, Pullman ME (1967) Oxidation and phosphorylation. In: Colowick SP, Kaplan
NO (eds) Methods in enzymology, vol 10. Academic, London, pp 215-250
18. Fauchier JP, Monpere C, Latour F, Neel C, Cosnay P, Brochier M (1983) Bloc auriculo-
ventriculaire aucours d'un syndrome du Keams et Sayre. Arch Mal Coeur 18: 295-303
19. Fukuhara N, Tokiguchi S, Shirakawa S, Tsubaki T (1980) Myoclonus epilepsy associated with
ragged-red fibers (mitochondrial abnormalities): disease entity or syndrome? Light and electron
microscopic studies of two cases and review of the literature. J Neurol Sci 47: 117-133
20. von Graefe A (1866) Bemerkungen iiber doppelseitige Augenmuskelliihmung basilaren Ur-
sprungs. Albrecht Graefes Arch 12: 265-277
21. Hammerstein W, Mortier W, Noack EM, Frenzel H, Liebert HJ, Toyka KV, Horstkotte D,
Bischof J, Weber U (1983) Klinische, morphologische und biochemische Befunde beim Keams-
Sayre-Syndrom. Fortschr Ophthalmol 80: 193-200
22. Horwitz SJ, Roessmann U (1978) Keams-Sayre Syndrome with hypoparathyroidism. Ann
Neurol 3: 513-518
23. Hiibner G, Grantzow R (1983) Mitochondrial cardiomyopathy with involvement of skeletal
muscles. Virchows Arch [A] 399: 115-125
24. Hiibner G, Gokel JM, Pongratz D, Johannes A, Park JW (1986) Fatal mitochondrial cardio-
myopathy in Keams-Sayre syndrome. Virchows Arch [A] 408:611-621
25. Hug G, Schubert WK (1970) Idiopathic cardiomyopathy. Mitochondrial and cytoplasmatic
alterations in heart and liver. Lab Invest 22:541-552
26. Jager BV, Fred HV, Butler RB, Carnes WH (1960) Occurrence of retinal pigmentation,
ophthalmoplegia, ataxia, deafness and heart block. Report of a case with findings at autopsy.
Am J Med 29: 888-893
Cardiac Involvement in Kearns-Sayre Syndrome 309

27. Karpati G, Carpenter S, Larbrisseau, Lafontaine R (1973) The Kearns-Shy-Syndrome. A multi-

system disease with mitochondrial abnormality demonstrated in skeletal muscle and skin. J
NeurolSci 19:133-151
28. Kearns TP, Sayre GP (1958) Retinitis pigmentosa, external ophthalmoplegia and complete
heart block. AMA Arch Ophthal 60: 280-289
29. Keams TP (1965) External ophthalmoplegia, pigmentary degeneration of the retina and cardio-
myopathy, a newly recognized syndrome. Trans Am Ophthalmol Soc 63: 559-568
30. Kiloh U, Nevin S (1951) Progressive dystrophy of the external ocular muscles (ocular myopathy).
Brain 74: 115-143
31. Kleber FX, Park JW, Hubner G, Johannes A, Pongratz D, Konig E (1987) Congestive heart
failure due to mitochondrial cardiomyopathy in Keams-Sayre Syndrome. Klin Wochenschr 65:
480-486
32. Kulbertus HE (1976) Re-evaluation of the prognosis of patients with LAP-RBBB. Am Heart J
92:665-672
33. Langes K, Frenzel H, Seitz RJ, Kluitrnann J (1985) Cardiomyopathy associated with Leigh's dis-
ease. Virchow Arch [A] 407:97-105
34. Luft R, Ikkos D, Palmieri G, Ernster L, Afzelius B (1962) A case of severe hypermetabolism of
nonthyroid origin with defect in the maintenance of mitochondrial respiratory control: a corre-
lated clinical, biochemical and morphological study. J Clin Invest 41: 1776-1804
35. LOsse B, Kuhn H (1982) Belastungsuntersuchungen bei Patienten mit Kardiomyopathie. Herz
7:91-102
36. DiMauro S, Bouilla E, Zeviani M, Nakagawa M, DeVivo DC (1985) Mitochondrial myopathies.
Ann NeuroI17:521-538
37. Mackay EH, Brown RS, Pickering D (1976) Cardiac biopsy in skeletal myopathy: Report of a
case with myocardial mitochondrial abnormalities. J Pathol120: 35-42
38. McComish M, Compston A, Jewitt D (1976) Cardiac abnormalities in chronic progressive exter-
nal ophthalmoplegia. Br Heart J 38: 526-529
39. Morgan Hughes JA, Darveniza P, Kahn SN, Landon DN, Sherratt RM, Laud JM, Clark JB
(1977) A mitochondrial myopathy characterized by a deficiency in reducible cytochrome B.
Brain 100:617-640
40. Morriss JH, Engster GS, Nora JJ, Pryor R (1972) His bundle recording in progressive external
ophthalmoplegia. J Pediatr 81: 1167-1170
41. Muller-Hocker H, Johannes A, Droste M, Kadenbach B, Pongratz D, Hubner G (1986) Fatal
mitochondrial cardiomyopathy in Keams-Sayre syndrome with deficiency of cytochrome-c-
oxidase in cardiac and skeletal muscle. Virchows Arch [B] 52: 353-367
42. Neustein HE, Lurie PR, Dahms B, Takahashi M (1979) An x-linked recessive cardiomyopathy
with abnormal mitochondria. Pediatrics 64: 24-29
43. Nicbroj-Dobosz J, Ryniewicz B, Fidzianska A, Badurska B (1985) Lipid storage myopathy in
Kearns-Sayre syndrome. Neurology 35: 1582-1586
44. Ogasahara S, Nishikawa Y, Yorifugi S, Soga F, Nakamura Y, Takahashi M, Hashimoto S, Kono
N, Tarui S (1986) Treatment of Keams-Sayre syndrome with coenzyme QI0. Neurology 36:
45-53
45. Olson W, Engle WK, Walsh GO, Einaugler R (1972) Oculocraniosomatic neuromuscular dis-
ease with ragged-red fibers. Arch Neurol26: 193-211
46. Pavlakis SG, Phillips PC, DiMauro S, DeVivo DC, Rowland LP (1984) Mitochondrial myopathy,
encephalopathy, lactic acidosis, and stroke like episodes: a distinctive clinical syndrome. Ann
NeuroI16:481-488
47. Pellock JM, Behrens M, Lewis L, Holub LD, Carter S, Rowland LP (1978) Keams-Sayre syn-
drome and hypoparathyroidism. Ann Neurol 3: 455-458
48. Pongratz D, Perwein J, Hubner G, Koppenwallner C, Toyka KV, Birnberger KL (1979) Wertig-
keit der Skelettmuskelbiopsie bei progressiver externer Ophthalmoplegie. Klin Wochenschr
57:779-788
49. Rheuban KS, Ayres NA, Sellers D, DiMarco JP (1983) Near-fatal Keams-Sayre syndrome. Clin
Pediatr (Phila) 22: 822-825
50. Roberts NI, Perloff JK, Kark RAP (1979) Cardiac conduction in the Keams-Sayre syndrome.
Am J Cardiol44: 1396-1400
51. Rowland LP, Hays AP, DiMauro S, DeVivo DC, Behrens M (1985) Diverse clinical disorders
associated with morphological abnormalities of mitochondria. In: Scarlato G, Cerri C (eds)
Mitochondrial pathology in muscle disease. Piccin Medical Books, Padova, pp 142-158
310 B. Schwartzkopff et al.: Cardiac Involvement in Keams-Sayre Syndrome

52. Rowland LP (1983) Molecular genetics, pseudogenetics and clinical neurology. Neurology 33:
1179-1195
53. Sandifer PH (1946) Chronic progressive ophthalmoplegia of myopathic origin. J Neurol Neuro-
surg Psychiatry 9: 81-83
54. Schnitzler ER, Robertson EC (1977) Familial Keams-Sayre syndrome. Neurology 29: 1172-
1174
55. Schwartzkopff B, Deckert M, Frenzel H (1987) Strukturanomalien der Mitochondrien des Herz-
muskels beim Keams-Sayre Syndrom (KSS). Z Kardiol [Suppl] 1: 18
56. Schwartzkopff B, Frenzel H, LOsse B, Borggrefe M, Toyka KV, Hammerstein W, Seitz R,
Deckert M, Breithardt G (1986) Herzbeteiligung bei progressiver extemer Ophthalmoplegie
(Keams-Sayre Syndrom). Z Kardiol 75: 161-169
57. Shy GM, Silberberg DH, Appel SH, Mishkin MM, Godfrey EH (1967) A generalized disorder
of nervous system, skeletal muscle and heart resembling Refsum's disease and Hurler's syn-
drome. Am J Med 42: 163-178
58. Uppal SC (1973) Keams syndrome, a new form of cardiomyopathy. Br Heart J 35 : 766-769
59. Yorifugi S, Oyasahara S, Takahashi M, Tarui S (1985) Decreased activities in mitochondrial
inner membrane electron transport system in muscle from patients with Keams-Sayre syn-
drome. J Neurol Sci 71: 65-75
Chapter 26

Some Neurological and Hereditary Aspects

of Progressive External Ophthalmoplegia and
Mitochondrial Encephalomyopathy
S.1. MELLGREN, T. TORBERGSEN, E.B.MATHIESEN, N.J.BRAUTAsET, and S. LINDAL

Increasing weakness of the extraocular muscles, progressive external ophthalmoplegia

(PEO), may be combined with a variety of other clinical features. It is also often seen
in the classic ocular and oculopharyngeal myopathies. For many years these have
been dealt with as variants of muscular dystrophies in neurological texts. Recogniz-
ing that underlying mitochondrial abnormalities may be present in some patients
with these syndromes contributes to making the classification very complicated.

Ocular Muscular Dystrophy

Ocular myopathy is the simplest form of the ocular syndromes but may also be
heterogeneous. It usually starts before the age of 30, ptosis is more or less promi-
nent, and there is often external ophthalmoplegia as well. Some, but not all, show
mitochondrial abnormalities in muscular biopsies. About half of the patients with
ocular myopathy have a family history of ptosis or external ophthalmoplegia consis-
tent with autosomal dominant inheritance.

Oculopharyngeal Muscular Dystrophy

Oculopharyngeal muscle dystrophy (OPMD) has been considered an autosomal dom-

inant muscle disease where the clinical features are progressive ptosis, usually after
the age of 50, followed by progressive dysphagia. Some patients may later develop
involvement of the face and extraocular muscles, and even weakness of the limbs.
However, apart from our family A to be briefly described in this paper, only few
cases of OPMD have shown distal spread with neurogenic involvement. As OPMD
is a late appearing and relatively mild disorder, rather extensive pedigrees have been
reported. Barbeau [1] was able to trace the condition through ten generations of a
French-Canadian family. Although it has been questioned whether it is justified that
OPMD deserves a separate classification, the clinical features are rather characteristic
and so are also the morphological changes in the muscles. Limb muscle biopsy shows
vacuolation of fibers, often rimmed by red stain with Gomori-trichrome staining,
resembling but not identical to the "ragged red fibers" of the oculocraniosomatic
syndrome [6]. Electron microscopy shows characteristic filamentous whorls and in-
clusions in some of the nuclei.
312 S.1. Mellgren et al.

OPMD, with onset in the thirties and rapid progression, may also be autosomal
recessive [3].

OPMD with Distal Spread and Neurogenic Involvement

Several reports have stressed oculopharyngodistal myopathy in single patients and

families [4, 7, 12]. We have recently examined members of a big family, where the
proband had 13 siblings, with a neuromuscular disease similar to the clinical picture
of the case described by Schmitt and Krause [12], in five of the seven patients exam-
ined. The picture consists of ptosis initially, and with pareses of extremities preced-
ing or appearing together with dysphagia. The pedigree is published elsewhere [2]
and indicates autosomal dominant heredity. The main clinical data are presented in
Table 1. Age at onset as well as heredity are in accordance with most cases of
OPMD. This also accounts for muscle biopsy findings, the myopathic changes in
EMG and slight elevation of CK in most cases. The muscle biopsies showed charac-
teristics in accordance with those often reported in OPMD [6]. Predominating was a
considerable variation of fiber size, including atrophic angular fibers, some of which
may be secondary to neurogenic involvement. Rimmed vacuoles were also observed.

Table 1. Summary of clinical findings in three of the seven patients exam-

ined in family A

Patient
11112 III 13 III 16

Sex F M M
Age 65 63 56
Age at onset 50 50 50
Symptom progression
p = ptosis, p--+ ex, b p--+ ex, b ex--+ p
b = bulbar weakness,
ex = paresis of extremities

Signs'
Ophthalmoplegia ++ + 0

Ptosis ++ ++ +
Facial weakness + + 0

Bulbar weakness ++ + 0

Muscle atrophy 0 + +
Proximal paresis ++ + +
Distal paresis + 0 +
Myotatic reflexes b 0 + ++
• 0 = normal, ++ = severe; b 0 = absent, ++ = normal.
Progressive External Ophthalmoplegia - Mitochondrial Encephalomyopathy 313

Fig. I. Electron micrograph showing a subsarcolemmal collection of myelin figures (M). The mito-
chondria appear normal with varying shapes (arrowhead) (x 3000)

Electron microscopy (EM) showed vacuoles and membranous whorls or myelin-like

figures (Fig. 1) which have been seen in, among other conditions, ocular myopathy
[6]. We did not observe any ragged-red fibers or abnormal mitochondria in these
cases. Patients with mitochondrial myopathies are usually younger and most cases
occur sporadically, although both autosomal dominant and autosomal recessive
inheritance occur among patients with mitochondrial abnormalities and external
ophthalmoplegia [3]. On the other hand, morphologically normal mitochondria do
not exclude mitochondrial myopathy [5], even when lactate and pyruvate in blood
and spinal fluid are normal.
All of the seven patients examined in this large family had electrodiagnostic signs
of polyneuropathy. They had decreased nerve conduction velocities, either on the
motor or sensory side or both. Six showed neurogenic EMG changes, and in five of
these there was a combination of neurogenic and myopathic characteristics. Poly-
neuropathy is very uncommon in OPMD [12] but seems to occur more often in mito-
chondrial myopathies [10, 13]. Except for sinus bradycardia in the index case and un-
specific T changes considered to be insignificant in his 65-year-old sister, there were
no cardiological manifestations.
Oculopharyngeal myopathy with no mitochondrial abnormalities seems to be
rarely combined with cardiomyopathy, although it has been described. Goto et al.
314 S.1. Mellgren et a1.

[7] report two cases, one apparently with a sporadic disorder, the other probably
familial. Both had heart enlargement. ECG of the sporadic case revealed atrial fibril-
lation, abnormal left-axis deviation, idioventricular rhythms, ventricular premature
beats and intraventricular block. The second patient, presumed to have a familial
disorder, had ventricular-premature beats, left-ventricular hypertrophy and antero-
septal-myocardial infarction. The vector electrocardiogram revealed impairment of
the intraventricular conduction. His 16-year-old son had mild atrioventricular block
and right incomplete bundle branch block.

Mitochondrial Myopathies

The mitochondrial myopathies constitute a complex and heterogeneous group of

muscular disorders with structural and functional abnormalities of the muscle mito-
chondria. However, in most patients, muscle is not the only tissue affected. Aerobi-
cally active organs with a high ATP demand, such as the brain, retina, heart, and
predominantly the type 1 fibers of skeletal muscle, are particularly vulnerable to de-
fects of this type. Sometimes cerebral symptoms are dominant, warranting the term
mitochondrial encephalomyopathy. Well-known syndromes are Kearns-Sayre syn-
drome, MELAS (mitochondrial myopathy, encephalopathy, lactic acidosis, and
stroke-like episodes), and MERRF (myoclonic epilepsy and ragged-red fibers).
However, the lack of biochemical consistency and cases with overlapping clinical
features make the justification of these distinctions controversial. Mitochondrial
myopathies and mitochondrial diseases have been excellently reviewed by DiMauro
et al. [5] and Morgan-Hughes [9], but a few features are summarized here.
Morphologically, mitochondrial abnormality may be suspected in skeletal muscle
on the light microscopical level with the Gomori-trichrome method. This method
stains large peripheral collections of abnormal mitochondria, and gives the picture of
so-called red-ragged fibers, which may also be demonstrated in frozen sections
stained for the mitochondrial enzymes succinate- and NADH dehydrogenase. The
ultrastructural changes are varied, including increased number and size of mito-
chondria. There may be paucity of cristae and various intramitochondrial inclusions
with complex paracrystalline and osmiophilic structures (Fig. 2). Similar mitochondrial
changes may be demonstrated in neurons, Schwann cells, liver and cardiac muscle
cells.
Lactic acidosis, although nonspecific, is a common abnormality in mitochondrial
disease. It is explained by impaired utilization of pyruvate in the Krebs cycle and by
increased glycogen utilization and glycolytic activity in muscles with defective
aerobic metabolism. Lactate may also be increased in the spinal fluid, particularly
in mitochondrial encephalomyopathies, possibly reflecting affection of CNS mito-
chondria.
Examples of dysfunction in substrate transport are deficiencies of the carnitine
acy1carnitine carrier system, particularly carnitine and carnitine palmityl transferase,
CPT I and II. Luft's disease, or hypermetabolic mitochondrial myopathy, is the clas-
sical example of a defect in energy conservation owing to a failure in coupling of
oxidative phosphorylation in muscle mitochondria.
Progressive External Ophthalmoplegia - Mitochondrial Encephalomyopathy 315

Fig.2. Collection of subsarcolemmal mitochondria with groups of paracrystalline inclusions (arrow-

head). Multiple electron-lucent vacuoles (V) are seen dispersed throughout the mitochondria
(x 40000)

Defects of the mitochondrial respiratory chain have been defined in many pa-
tients in later years (reviewed in more detail by DiMauro et al. [5] and Morgan-
Hughes [9]). Complex I of the respiratory chain contains NADH-ubuquinone reduc-
tase; complex II, succinate-ubiquinone reductase; complex III, ubiquinol-cyto-
chrome c reductase; while complex IV constitutes cytochrome c oxidase with the
cytochromes a and a3. Defects in these components are displayed by a range of clini-
cal features, like ataxia, dementia and myoclonus caused by deficient cytochrome b
and succinate-ubiquinone reductase. Kearns-Sayre syndrome has been connected to
NADH-ubiquinone reductase and ubiquinol-cytochrome c reductase. This accounts
also for other oculoskeletal mitochondrial myopathies which may also be caused by
deficient cytochrome c oxidase.
Although many patients with mitochondrial disease have progressive external
ophthalmoplegia (PEO) in some this is not present. On the other hand, PE~, as in
our family A, may occur without characteristics of mitochondrial myopathy. This
examplifies the wide spectrum in this group of disorders as does also familial occur-
rence. Most cases of mitochondrial disease have been described as sporadic. How-
ever, a family with mitochondrial encephalomyopathy (family B) has recently been
examined in the Tromsf/l Department of Neurology, and further studies on this family
are still in progress.
316 S. I. Mellgren et al.

Mitochondrial Encephalomyopathy with Maternal Pattern of Inheritance

The proband of this family B, a female born 1935, was first admitted in 1985 owing
to progressive ataxia during the last 5-6 years. She had had hypacusis for many years
and used hearing aids since the age of 40. Although she had a few times experienced
myoclonic jerks when falling asleep, she was not considered to have epilepsy from a
clinical point of view. Neurological examination revealed retinitis pigmentosa. Fur-
ther, there was generally slight muscular weakness particularly in distal lower limb
muscles. Vibration and joint position sense in distal lower limbs were also slightly af-
fected. The achilles reflex was absent bilaterally, and there was truncal as well as ex-
tremity ataxia. Plantar responses were normal. EMG showed a mixture of mainly
myopathic, but also neurogenic characteristics. EEG revealed intermittent short epi-
sodes with spike and wave complexes. CT scan showed calcifications in the basal
ganglia and some atrophy of cerebral and cerebellar cortices. Of the evoked re-
sponses, the brainstem evoked response was pathological. Lactate and pyruvate in
blood and spinal fluid were markedly increased. Muscle biopsy examined with light
and electron microscopy demonstrated abnormal mitochondria with paracrystalline
inclusions and other changes (Fig. 2).
Most of these features suggesting the same disease were also detected in the pro-
band's eight siblings and in her four daughters. The proband's mother, who died
many years ago, had had unsteadiness and hypacusis as well.
The pedigree (Fig. 3) indicates maternal inheritance like the family with mito-
chondrial myopathy and myoclonus described by Rosing et al. [11]. In this variety of
inheritance it is assumed that mitochondrial DNA which has undergone mutation
is passed from the mother to all her children, but is then further transmitted only
through the daughters. It has also been suggested that the severity of the symptoms
reflects the proportion of mutant mitochondrial DNA within the original fertilized
egg and the cells derived from it. Cardiological features in family B with mitochondrial
encephalomyopathy are presented by Lunde et al. [8].

:I 3 4 5 8 B 9 10

II

13 14 15

III

Fig.3. Pedigree of family B. (_ e), affected male, female; (/), dead; (0 0), unaffected male,
female; (X), examined; I-III, generation; 1-18, patient number
Progressive External Ophthalmoplegia - Mitochondrial Encephalomyopathy 317

Concluding Remarks

The most common type of mitochondrial myopathy is associated with PEO and other
signs, often fulfilling the criteria of Kearns-Sayre syndrome. The hereditary diseases
in this presentation point to some sort of extremes of progressive external ophthal-
moplegia and related disorders: one with PEO, without other characteristics of mito-
chondrial myopathy and cardiopathy, but with myopathy and neuropathy; the other
with mitochondrial encephalomyopathy without PEO, but with various different
cardiac disturbances.
Acknowledgement. We thank Margaret Srether for typing and other assistance with the manuscript.

References

1. Barbeau A (1966) The syndrome of hereditar late onset ptosis and dysphagia in French-Canada.
In: Kuhn E (ed) Progressive Muskeldystrophie. Springer, Berlin Heidelberg New York, pp
102-109
2. Brautaset NJ, Mellgren SI, Torbergsen T, Lindal S (1986) A family with progressive external
ophthalmoplegia with polyneuropathy as a predominant feature. Ups Med J Sci [Suppl] 43: 83
3. Bundey S (1985) Genetics and neurology. Churchill Livingstone, Edinburgh
4. Croft PB, Cutting JC, Jewesbury ECO, Blackwood W, Mair WGP (1977) Ocular myopathy
(progressive external ophthalmoplegia) with neuropathic complications. Acta Neurol Scand 55:
169-197
5. DiMauro S, Bonilla E, Zeviani M, Nakagawa N, DeVivo DC (1985) Mitochondrial myopathies.
Ann NeuroI17:521-538
6. Dubowitz V (1985) Muscle biopsy. A practical approach, 2nd edn. Bailliere Tindall, London
7. Goto I, Kanazava Y, Kobayashi T, Murai Y, Kuroiwa Y (1977) Oculopharyngeal myopathy
with distal and cardiomyopathy. J Neurol Neurosurg Psychiatry 40: 600-607
8. Lunde P, Torbergsen T, Rasmussen K, Mathiesen E, Mellgren SI (1987) Cardiological findings
in a family with mitochondrial encephalo-myopathy. Abstracts from heart and brain seminar,
University of TromSjlj, p 50
9. Morgan-Hughes JA (1986) Mitochondrial diseases. Trends Neurosci 9: 15-19
10. Peyronnard J-M, Charron L, Bellavance A, Marchand L (1980) Neuropathy and mitochondrial
myopathy. Ann NeuroI7:262-268
11. Rosing HS, Hopkins LC, Wallace DC, Epstein CM, Weidenheim K (1985) Maternally inherited
mitochondrial myopathy and myoclonic epilepsy. Ann Neurol17:228-237
12. Schmitt HP, Krause K-H (1981) An autopsy study of a familial oculopharyngeal muscular dys-
trophy (OPMD) with distal spread and neurogenic involvement. Muscle Nerve 4:265-305
13. Yannikas C, McLeod JG, Pollard JD, Baverstock J (1986) Peripheral neuropathy associated
with mitochondrial myopathy. Ann NeuroI20:249-257
Chapter 27
Neurological and Cardiological Findings
in Systemic Lupus Erythematosus
R.OMDAL, S. I. MELLGREN, and G.HUSBY

Introduction

Systemic lupus erythematosus (SLE) is a disease characterized by multiple immuno-

logical aberrations [6]. The yearly incidence and prevalence in Scandinavia have been
estimated to be 4.8 and 28 per 100000 people, respectively [18]. Usually, arthritis
and rashes are the most prevalent symptoms, and antinuclear antibodies (ANA) are
detected in most patients' sera. Central nervous system (CNS) and cardiac involve-
ment have also been well documented [6, 8] and are regarded as such important fea-
tures of the disease that they are included in the diagnostic criteria for SLE [23].

Neurological Manifestations of SLE

The CNS, the peripheral nervous system (PNS), and muscle can be affected in SLE.
CNS affection is reported most frequently and occurs in 40%-70% of cases [1, 8,15].
There are reasons to believe that these figures are too small as minor or subtle
neurological symptoms and signs can be easily overlooked or excluded. Further-
more, neurological manifestations may sometimes be attributed to the use of such
drugs as corticosteroids, antimalarial agents, or cytostatics.
More recent papers stress that most patients have nonfocal disorders and often
present with migraine or psychiatric symptoms such as mental depression. This may
occur intermittently and not necessarily in a stage of active disease.
Autopsy studies show that the dominant neuropathological finding is a small-
vessel vasculopathy with destructive and proliferative changes leading to microinfarcts
[15]. This may be responsible for the cerebral atrophy often observed in brain scans
in SLE patients [10]. Some claim this, however, to be related to treatment with cor-
ticosteroids [5].
The mechanisms for the various CNS manifestations in SLE are not clarified, but
may be secondary to the wide immunological aberrations occurring in this disease
[14]. Possible examples are entrapment of circulating immune complexes by Fc-
receptors in the choroid plexus, antineuronal antibody production, and antilympho-
cyte antibodies cross-reacting with brain tissue. Antiphospholipid antibodies, which
may be detected as cardiolipin antibodies, are associated with cerebral infarction in
SLE, but also with recurrent spontaneous abortions, thrombocytopenia, and venous
and arterial thrombosis anywhere in the systemic circulation [12]. These antibodies
Neurological and Cardiological Findings in Systemic Lupus Erythematosus 319

could be responsible for an alternative mechanism for endothelial cell injury leading
to the neurological phenomena in SLE.
The frequency of miscellaneous CNS manifestations of SLE varies in different
series. Most often psychiatric disturbances, organic brain syndromes, and seizures
are reported as the most prevalent disorders. Apart from these, a variety of other
symptoms and signs can appear, such as migraine and other headache syndromes,
aseptic meningitis, transverse myelopathy, multiple sclerosis, chorea, visual distur-
bances, and stroke. Not necessarily all of these manifestations are attributed to the
disease itself, but they could in some cases be coincidental or secondary to drug
therapy [16].
We studied retrospectively 30 consecutive patients with SLE admitted to the De-
partment of Rheumatology and subjected them to a clinical neurological examina-
tion. The accumulated neurological manifestations from the beginning of the disease
until the time of examination were thus collected. Twenty-five of these patients
(83%) had experienced CNS manifestations and related symptoms one or more times
during their illness (Table 1). The most frequent single finding was migraine in 12 pa-
tients (40%). This could be further divided into classic migraine (six patients), com-
mon migraine (three patients), migraine sine migraine (two patients), and migraine
accompagne (one patient). The second largest group (six patients, 20%) had un-
specific, severe, protracted headache. Six patients (20%) had vertigo of different
varieties and five patients (17%) psychiatric disturbances. Amaurosis fugax had oc-
curred in two and seizures in another three patients (10% ) during periods of high dis-
ease activity. Various other CNS manifestations appeared in small numbers.
The high frequency of CNS and related manifestations in this study is probably
due to its design as well as to long duration of the disease (mean 10.2 years; range
1-30 years). The frequent finding of migraine is noteworthy. It has, however, been
reported previously [2, 21], while others have regarded migraine as a minor or non-
existent problem in SLE [1, 3, 8]. Our findings are in agreement with the view that
the "modem SLE patient" is a reasonably well young woman with arthralgia or ten-
dinitis and intermittent migraine or depression [14]. Severe headache, either of mi-
grainous type or other, therefore seems to be an important manifestation in SLE.
Vertigo of various types was also a rather frequent complaint (20%). No figures
for this has to our knowledge been reported before, and thus no meaningful com-
parison with other materials is possible.
Psychiatric disorders occurred in periods of high disease activity. The frequency
was similiar to that found by others [9], but is dependent on the definition of psychi-
atric illness. If a broader conception is used, including organic brain syndromes and
psychoneurosis, higher figures are obtained [1].
Epilepsy is regarded as rather typical of SLE, but is mostly seen during episodes
of high disease activity or in the terminal phase where it has been reported in up to
70% of patients [20]. In most patients with moderate or low disease activity, seizures
are therefore observed less frequently and our figures for this manifestation are in
agreement with earlier works [6, 8, 15].
Peripheral neuropathies and myopathies manifest themselves in several different
ways in SLE. Reported frequencies for peripheral neuropathy vary from 6% to 21 %
in different studies [1, 8, 9, 11]. A variety of different neuropathies may occur [15],
with sensory polyneuropathy regarded as the most common form. Purely motor-type
320 R.Omdal et al.

Table 1. CNS and related manifestations in 30 patients with

SLE

n %

Migraine 12 40
Classic 6
Common 3
M. sine migraine 2
M. accompagne 1

Cephalalgia 6 20

Vertigo 6 20

Psychiatric disorders 5 17
Psychosis 2
Mental depression 2
Confusion 1

Epileptic fits 3 10

Amaurosis fugax 3 10

Tremor 2

Scotoma 2

Incoordination 2
Parkinsonism 1

Gait unsteadiness 1

Photopsia 1
Left hemisensory loss 1
Myelopathy 1
Cerebral infarction 1
Aseptic meningitis 1
Encephalopathy' 1

• Sequele after meningococcus meningoencephalitis.

manifestations resembling the Guillain-Barre syndrome are sometimes observed, as

well as a mixed type [8, 21]. The mechanism for PNS affection is probably an auto-
immune reaction leading to axonal degeneration or demyelination. Mononeuritis
multiplex is associated with vasculitis of the vasa nervorum leading to ischemia in
single nerves. Like in the eNS manifestations of SLE patients, however, neuropathy
secondary to uremia or drugs should be excluded.
Myopathy also occurs in different forms in SLE [19]. The clinical picture of mus-
cular weakness is nonspecific and can be difficult to distinguish from neuropathy.
Typical myositis confirmed by muscle biopsy is reported in frequencies from 5% to
Neurological and Cardiological Findings in Systemic Lupus Erythematosus 321

Table 2. Neuromuscular manifestations in 30 patients with

SLE

n %

Carpal tunnel syndrome 7 23"

Peripheral polyneuropathy 2
Cervical radiculopathy 1
Myositis 3 10
Myositis with vasculitis 1
Muscular weakness 7 23

" Suspected on clinical grounds in an additional three patients.

40% [8, 9, 20], but myastenia gravis, vacuolar myopathy, vasculitis, inclusion body
myositis, and myopathy induced by corticosteroids and chloroquine are occasionally
seen in SLE.
In our study neuromuscular manifestations were reported in 11 patients (37%)
(Table 2). Carpal tunnel syndrome confirmed by impaired nerve conduction studies
(seven patients) or suggested by typical clinical findings (three patients) was the most
frequent single finding (ten patients, 33%). Two patients had peripheral neuropathy
confirmed by nerve conduction studies, and one patient a cervical radiculopathy
thought to be secondary to mononeuritis multiplex. Three patients (10% ) had histo-
logically proven myositis. In addition, seven patients had experienced severe muscu-
lar weakness.

Cardiological Manifestations of SLE

Cardiac involvement in SLE has been well documented [8, 13]. It consists primarily
of four basic types: pericarditis, endocarditis, myocarditis, and coronary heart dis-
ease. It is rarely observed as the initial manifestation of SLE [8].
Pericarditis is clinically observed in 20%-30% of cases [7, 8,11] and is regarded
as the most frequent cardiac manifestation of SLE. Echocardiography demonstrates
pericardial fluid as evidence of pericarditis in up to 40% [17] while post mortem
studies find evidence of pericardial involvement in up to 80% of cases [4,21]. The
pericardial effusion secondary to pericarditis is usually small and infrequently leads
to hemodynamic complications. In some instances of highly active SLE with clinical
cardiac failure, however, the differentiation between cardiac tamponade and myo-
carditis is important. Myocardial involvement is difficult to assess antemortem, but
there is evidence from smaller invasive studies that myocarditis is not a rare phenom-
enon and can occur in SLE patients with no clinical signs of cardiac dysfunction [22].
At necropsy, myocardial involvement varies from about 20% to 40% of cases [4, 20].
Myocardial infarction caused by accelerated coronary atheromatosis or coronary
vasculitis should also be considered in the case of cardiac failure in SLE. as well as
322 R. Omdal et al.

Libman-Sacks endocarditis that, although rare, may cause valvular dysfunction.

Long-term corticosteroid treatment may have a possibly deleterious effect in some of
these cases [4], especially in hypertension and coronary atheromatosis which may ac-
celerate.

Conclusion

Systemic lupus erythematosus is a disease with multiorgan involvement. It gives rise

to symptoms and signs of varying severity. Both neurological and cardiological mani-
festations occur in a substantial number of patients with this disease and should be
remembered in the clinical evaluation of such patients.

References

1. Abel T, Gladman DD, Urowitz MB (1980) Neuropsychiatric lUpus. J Rheumatol 7: 325-333

2. Adelman DC, Saltiel E, Klinenberg JR (1986) The neuropsychiatric manifestations of systemic
lupus erythematosus: an overview. Semin Arthritis Rheum 15: 185-199
3. Brandt KD, Lessel S (1978) Migraineous phenomena in systemic lupus erythematosus. Arthritis
Rheum 21 :7-16
4. Bulkley BH, Roberts WC (1975) The heart in systemic lupus erythematosus and the changes in-
duced in it by corticosteroid therapy. Am J Med 58: 243-264
5. Carette S, Urowitz MB, Grosman H, St.Louis E (1982) Cranial computerized tomography in
systemic lupus erythematosus. J Rheumatol 9: 855-859
6. Dubois EL, Tuffanelli DL (1964) Clinical manifestations of systemic lupus erythematosus. Com-
puter analysis of 520 cases. JAMA 190: 104-111
7. Dubois EL, Wierzchowiecki M, Cox MB, Weiner JM (1974) Duration and death in systemic
lupus erythematosus. JAMA 227: 1399-1402
8. Estes D, Christian CL (1971) The natural history of systemic lupus erythematosus by prospec-
tive analysis. Medicine 50: 85-95
9. Feinglass EJ, Arnett FC, Dorsch CA, Zizic TM, Stevens MC (1976) Neuropsychiatric manifes-
tations of systemic lupus erythematosus: diagnOSis, clinical spectrum and relationship to other
features of the disease. Medicine 55 : 323-339
10. Gayliss NB, Altman RD, Ostrov S, Quencer R (1982) The selective value of computed tomogra-
phy of the brain in cerebritis due to systemic lupus erythematosus. J Rheumatol 9 : 850-854
11. Grigor R, Edmonds J, Lewkonia R, Bresnihan B, Hughes GRV (1978) Systemic lupus ery-
thematosus: a prospective analysis. Ann Rheum Dis 37: 121-128
12. Harris EN, Gharavi AE, Asherson RA, Boey ML, Hughes GRV (1984) Cerebral infarction in
systemic lupus: association with anticardiolipin antibodies. Clin Exp RheumatoI2:47-51
13. Hejmancik MR, Wright JC, Quint R, Jennings FL (1964) The cardiovascular manifestations of
systemic lupus erythematosus. Am Heart J 68 : 119-130
14. Hughes GRV (1980) Central nervous system lupus - diagnosis and treatment. J Rheumatol 7:
405-411
15. Johnson RT, Richardson EP (1968) The neurological manifestations of systemic lupus ery-
thematosus. Medicine 47:337-369
16. Kaell AT, Shetty M, Lee BCP, Lockshin MD (1986) The diversity of neurological events in sys-
temic lupus erythematosus. Arch Neurol 43: 273-276
17. Maniscalo BS, Feiner JM, McCans JL (1975) Echocardiographic abnormalities in system lupus
erythematosus. Circulation 52: 211
Neurological and Cardiological Findings in Systemic Lupus Erythematosus 323

18. Nived 0, Sturfelt G, Wollheim F (1985) Systemic lupus erythematosus in an adult population in
Southern Sweden: incidence, prevalence and validity of ARA revised classification criteria. Br
J Rheumatol 24: 147-154
19. Richardson EP Jr (1980) Systemic lupus erythematosus. In: Vinken PJ, Bruyn GW (eds) Hand-
book of clinical neurology. Elsevier/North Holland, Amsterdam, pp 273-293
20. Ropes MW (1976) Systemic lupus erythematosus. Harward University Press, Cambridge
21. Rothfield N (1981) Clinical features of systemic lupus erythematosus. In: Kelley WN, Harris ED
Jr, Shaun R, Sledge CB (eds) Textbook ofrheumatology. Saunders, Philadelphia, pp 1106-1132
22. Strauer BE, Brune I, Schenk H, Knoll D, Perings E (1976) Lupus cardiomyopathy: cardiac
mechanics, hemodynamics, and coronary blood flow in uncomplicated systemic lupus ery-
thematosus. Am Heart J 92:715-722
23. Tan EM, Cohen AS, Fries JF, Meshane DJ, Rothfield NF, Schaller JG, Talal N, Winchester RJ
(1982) The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis
Rheum 100:714-727
Part VII
Hypo- and Hyperbaric Environment
Chapter 28
High Altitude Physiology and Pathophysiology:
Medical Observations During the Norwegian
Mount Everest Expedition
K. T.STOKKE

Hypobaric medicine deals with hypoxia, and neither other gases nor pressure itself is
of importance for signs and symptoms. The relative composition of the air is constant
at all altitudes, oxygen making up 20.93% of the total. When oxygen passes from
outside air via the lungs and blood to the tissues a series of drops in pressure take
place. With increasing altitude this oxygen cascade flattens and finally determines
the absolute limit of the ascent.
The task of climbing Mt.Everest (8848m), set by Norton in 1924, took 29 years
to fulfill, and it took another 24 years for someone to reach the summit without
supplementary use of oxygen. The successful attempt of Messner and Habeler in
1978 was the final proof, ending a long discussion about whether it would be possible
to reach the summit of Mt. Everest without the use of extra oxygen. Earlier con-
siderations were based on "standard atmosphere pressure", which for Mt. Everest is
236mmHg. Near the equator the pressure at all altitudes is higher than that pre-
dicted by the International Civil Aviation Organization, possibly reaching close to
250mmHg. This difference between 236 and 250mmHg is enough to explain why
earlier theoretical considerations were wrong [62]. Nevertheless, the summit of
Mt. Everest represents the upper limit for human life on earth. This limit can be
reached without the use of supplementary oxygen - by some, when the weather is
good and the atmosphere pressure is "high" [55].
Why does Mt. Everest represent this upper limit? How can the organism with-
stand the ever-increasing hypoxia, and which are the most threatening pathophysio-
logical changes?
This review gives an outline of the physiological and pathophysiological changes
which are seen at high altitude. The author was expedition doctor on the Norwegian
Mt. Everest expedition in 1985. The expedition turned out to be the most successful
Everest expedition ever. Seventeen participants reached the summit, and - more
important - there were no accidents. Although the expedition was a climbing ex-
pedition exclusively, some scientific investigations were performed. Data on cerebral
blood flow, changes in plasma and red-cell volumes, and the rationale and outcome
of a hemodilution will be presented.
This review deals with the physiological and pathophysiological aspects of a
gradual exposure to high altitude. Acute exposure as seen by use of a low-pressure
chamber or travelling by nonpressurized aircraft will not be discussed. However, it
might be fruitful to be reminded [19] of what happens if the exposure is too abrupt
as was the case for the balloonists of the last century who, deliberately or by acci-
dent, were exposed to severe hypoxia. At an altitude of 1500m dim light vision is im-
paired, and at 3000m some higher intellectual functions (e.g., doing arithmetic) are
328 K.T.Stokke

affected. At 4500m handwriting is sloppy and thought processes are slow. At 5500m
some people collapse, and at 6000 m everyone is influenced by the altitude and more
collapse. At 7500 m unconsciousness occurs in just a few moments. Thus, acute expo-
sure threatens life even at an altitude which is well tolerated after acclimatization.

The Process of Acclimatization

Acclimatization to a high altitude involves a complex of processes [24], of which the

respiratory and hematological changes are the most conspicuous. Yet peripheral tis-
sues also show adaptation to the threat of hypoxia. About 80% of the acclimatization
is accomplished in 10-14 days, and the complete process is finished in 4-6 weeks.
Adaptation to high altitude is invariably associated with an increase in ventila-
tion, both in frequency and depth. Hyperventilation aims at keeping alveolar oxygen
pressure (P02) as close to that of the atmosphere as possible. At the same time, CO 2
is washed out, leading to pronounced respiratory alkalosis. It has been estimated [61]
that a climber who reaches the summit of Mt. Everest without the supplementary use
of oxygen may have a blood pH of 7.78. This extreme alkalosis - which if of sudden
onset is not compatible with life - in tum induces a shift of the oxyhemoglobin dis-
sociation curve to the left. Consequently, erythrocyte uptake of oxygen is enhanced,
and delivery to the tissues is hampered.
Hypoxia evokes ventilatory response and hyperventilation, whereas the concomi-
tant loss of CO2 has the opposite effect. The shift between the demand for oxygen
on the one hand and the resulting hypocapnia on the other turns respiration on and
off. This Cheyne-Stokes respiration is common at altitudes above 5000m, can be
pronounced, and may disturb sleep markedly [51].
The ventilatory response to hypoxia differs among individuals, races, and
species. In the llama (Lama glama) a weak response has been reported [17]. Instead
of hyperventilating - and thereby maintaining the highest possible P02 - oxygen
extraction from the blood is increased [17]. It has been claimed that the sherpas gen-
erally have a blunted response, and that this explains why they "suffer" less from
Cheyne-Stokes respiration at high altitude [12]. It has, however, been questioned
whether their response is low [12, 45, 52]. The possible beneficial or negative effect
of a blunted hypoxic ventilatory response is a matter of debate [45] (see below).
During acclimatization the concentration of 2,3-diphosphoglycerate (2,3-DPG)
in red blood cells increases [64]. This in tum leads to a shift of the oxyhemoglobin-
dissociation curve to the right. While hypoxia-induced hyperventilation and alkalosis
induce a shift to the left, 2,3-DPG normalizes this shift. At altitudes above 6300m
the effect of alkalosis prevails and the oxyhemoglobin-dissociation curve shifts pro-
gressively to the left [64].
Increased 2,3-DPG simplifies delivery of oxygen to the tissues, while alveolar up-
take is reduced. The possible benefit is often difficult to predict. Animals that suc-
cessfully live at high altitudes, like the yak and llamas, have no 2,3-DPG at all [17].
No increase in the concentration of erythrocyte 2,3-DPG has been observed in the
sherpas. At extreme altitudes the total positive outcome of a 2,3-DPG change in man
is doubtful [7].
High Altitude Physiology and Medicine 329

Also the hematological changes are conspicuous. Normally the hemoglobin con-
centration increases by about 30% -50%. Hypoxia induced increased secretion of
erythropoietin in turn leads to enhanced erythropoiesis [63]. Some of the increase in
hemoglobin concentration and hematocrit is due to the almost unavoidable dehydra-
tion, which is also seen during a normal process of acclimatization. Until recently any
rise in hematocrit was considered beneficial, and the more pronounced it was, the
more effective was thought to be the transportation of oxygen to the tissues. However,
the viscosity of the blood increases almost exponentially with increasing hematocrit,
and tissue perfusion and tissue extraction of oxygen may suffer [5]. The optimal
hematocrit at high altitude is not known [63]. No proof exists that hematocrit values
above 45% have any positive influence on oxygen transport, irrespective of the
altitude [43].
An interesting fact with regard to the evaluation of the potential benefit from an
increased number of red blood cells is that such changes are not found in some native
Andeans, well adapted to high altitude [8]. In some rodents and in the llama no in-
crease in the hemoglobin concentration and in hematocrit is seen [17].
In short, the hematologic response to altitude has no dramatic influence on the
positive outcome of acclimatization. The increase in hemoglobin concentration is
counteracted by increased viscosity, and the increase in 2,3-DPG seems beneficial at
some altitudes, but probably not at extreme ones. Other, till now unknown, mecha-
nisms must be responsible for the favorable oxygen transport in the sherpas.
Exposure of an individual to hypoxia leads to an increase in cardiac output [24].
This is due to an increased heart rate without concomitant decrease in stroke vol-
ume. Mter a few days the heart rate is normalized, as is resting cardiac output. With
exercise, however, maximum cardiac output is markedly reduced [41] and is more
pronounced in sojourners than in natives [15, 36]. Hence, altitude clearly limits maxi-
mal cardiovascular performance [24]. The increased hemoglobin concentration and
increased alveolar ventilation cannot fully compensate this, and maximal oxygen
consumption will be reduced.
Adaptation to the hypoxic state also takes place in the tissues. The immediate ef-
fect may be shunting of blood from less essential organs to organs of more vital im-
portance, e.g., heart and brain. With time, reserve capillaries are recruited, either by
opening closed capillaries or by the formation of new ones [24, 52]. An increase in
the myoglobin concentration of the muscles [38] facilitates the movement of oxygen.
Adaptive changes have also been demonstrated at the mitochondrial level, their
number and their concentration of cytochrome oxidase being increased [38].

Performance at High Altitude

Human performance at extreme altitude is severely limited by the low oxygen pres-
sure. In a theoretical analysis of pulmonary gas exchange, West and Wagner [59]
concluded that the maximal O 2 uptake on the summit of Mt. Everest was only slightly
above the basal requirements of the body. The predicted maximum oxygen con-
sumption (V02.",,) was exquisitely sensitive to small variations in barometric pressure
[55].
330 K.T.Stokke

The barometric pressure on mountains in the equatorial or temperate zones is

often higher than that predicted from the International Civil Aviation Organization
Standard Atmosphere [62]. For Mt.Everest the difference is about 14-17mmHg,
depending on the weather. Thus, the "physiological altitude" of Mt. Everest is lower
than the real one. V0 2mu decreases by about 1% for every 100 m of ascent. Accord-
ing to this approximation, maximal oxygen uptake on the summit of Mt. Everest
should be about 12% of that at sea level. More accurate field studies have been per-
formed by West et al. [60,61]. In two well-acclimatized subjects an O2 consumption
of about 111min was measured under conditions corresponding to those on the sum-
mit of Mt. Everest. The effect of various factors had earlier been underestimated,
and the correction of these explains why man can just reaches the highest point on
earth.
Which individual factors influence high altitude performance, and is maximal
oxygen uptake at sea level a useful predictor? In a study of world-class high altitude
climbers Oelz et al. [34] found V02mu values which, although higher than those ob-
served in untrained subjects, were similar to those of nonathletic well-trained sub-
jects and well below the values of elite long-distance runners. There was no relation-
ship between individual climbing capability and the observed V0 2mu •
The significance of hypoxic ventilatory response on performance has also been
debated [45, 46]. At sea level a lower breathing rate at high exercise levels may be
an advantage in endurance sports [45]. At high altitude a low ventilatory response
has been linked with a tendency to develop acute mountain sickness (AMS, see
below) [11, 30]. At extreme altitudes the price of the increased work of breathing
may be outweighed by a small increase in alveolar oxygen, and in this way the perfor-
mance will be improved when supplemental oxygen is not used. Although some very
successful climbers have a high ventilatory response when tested at sea level [25, 45],
conclusive data are still lacking.
At present it is not known why some people perform better than others ~t ex-
treme altitudes. Neither maximal oxygen uptake nor hypoxic ventilatory response
can fully explain the existing differences. It has been speculated that since the maxi-
mal pulmonary diffusion capacity for oxygen seems to represent a limitation [59],
greater diffusion capacity may be a peculiar characteristic of elite altitude moun-
taineers [34]. It has also been speculated whether the severe alkalosis and hypo-
capnia induced by extreme hyperventilation may be more limiting than hypoxia at
extreme altitudes [20].

Acute Mountain Sickness

Acute mountain sickness (AMS) denotes a heterogeneous complex of symptoms,

commonly occurring on travels to altitudes above 3000-3500 m. There is wide inter-
and intraindividual variation in the tendency to develop AMS and in the speed of
onset and severity [9, 29]. Usually there is a latency period lasting several hours to a
few days before symptoms are manifest. About 50% of all travellers develop symp-
toms if they make a rapid ascent to 3500m, and 80%-90% will have symptoms at
5000 m if time for acclimatization is insufficient.
High Altitude Physiology and Medicine 331

Despite the increasing interest in AMS, fairly little is known about the disease
mechanism. The relative hypoxia at increasing altitudes represents the central patho-
genetic factor. Membrane permeability may increase [1, 47], possibly due to a
weakening ofthe A'fP-dependent sodium pump [20]. In addition, changes in the ves-
sel tone (dilatation or constriction) and/or changes in blood flow accentuated by
exercise and an increased pressure, all contribute to fluid retention and edema for-
mation [10, 11, 37, 47], the central element in the various manifestations of AMS
[49].
In its mild or moderate forms most of the symptoms are "vascular" and to some
extent reminding of the clinical picture of migraine, e.g., headache, loss of appetite,
nausea, and lassitude. Cheyne-Stokes respiration during sleep and undue shortness
of breath on exertion are other common symptoms. Although usually a self-limiting
condition without serious sequelae, development may be dramatic, culminating in
the clinical picture of pulmonary or cerebral edema [6, 9, 14,29].
Retinal bleeding [19, 26] is usually considered part of the AMS complex. About
half of all persons staying above 6000m will have a hemorrhage in the retina [19].
Whereas the other elements of the AMS complex may be avoided by a slow and suc-
cessful acclimatization, hemorrhage strikes more randomly. Although more frequent
in persons suffering from severe AMS, it may also be seen in persons having no
symptoms at all.
When the macular region is not affected the bleedings are not noticed by the sub-
ject and will then represent no hindrance for further ascent. They usually disappear
within a few weeks after the return to sea level.
Retinal bleeding is also seen in about half of all newborns, and at least the con-
dition of low oxygen pressure in the fetus is similar to that encountered by high-
altitude climbers.
"Climb high, sleep low" and "hasten slowly" are rules which should be followed
and which favor acclimatization. With the exception of retinopathy, high altitude ill-
ness can be prevented by slow ascent, and descent is the only rational therapy. Those
few cases which end fatally have most often been misinterpreted.
Minor symptoms may be alleviated by symptomatic drugs, and descent may then
be neglected. Although general or local edema is often part of the more severe clini-
cal picture [37], the use of diuretics may be a two-edged sword. Dehydration predis-
poses to development of AMS, and despite presence of edema, the person may in
fact need fluid supply. Various drugs (e.g., dexamethazone, spironolactone, naproxen,
ataraxic drugs, salicylates etc.) have been tried, both with a preventive and a curative
purpose [21, 28], however with no convincing effect. The use ofpentoxifylline (Tren-
tal) by its positive influence on blood rheology (for references, see [54]) has been
suggested in the prevention of AMS [35]. More data both on the effect of pentoxifyl-
line on AMS and on the perfusion problems at extreme altitudes (see later) are
needed, however.
Acetazolamide (Diamox), although the object of debate, is the only drug which
seems to have some beneficial effect when used prophylactically [23, 27]. The mech-
anism of action in AMS is obscure but is due to the inhibiting effect on carbonic acid
anhydrase, thereby reducing the marked tendency to respiratory alkalosis caused by
hyperventilation. Acetazolamide also leads to an increased cerebral blood flow [39].
This is caused by CO 2 retention and possibly also by a direct, specific effect on the
332 K.T.Stokke

smooth muscle of the cerebral vessels [16]. The drug reduces periodic breathing dur-
ing sleep at high altitudes, and thus also sleep hypoxemia [50]. In this respect it is
superior to almitrine [13]. In a recent report [3] the positive effect of acetazolamide
on exercise performance and muscle mass at high altitudes has been described. Thus,
the beneficial effects of acetazolamide seem well documented. Its diuretic effect
must, however, be borne in mind and explains why some abandon the drug.
Although the headache of the milder forms may effectively be treated by para-
cetamol, and may represent no hindrance for proceeding to a higher altitude, neglect
of more severe symptoms is always dangerous. The only safe therapy and prevention
of progress is to descend, preferably to an altitude lower than that at which the symp-
toms started. Treatment of severe (malignant) forms follows common guidelines
(e.g., coma care, airway management, oxygen, morphine (?), dexamethazone etc.).
In a recent report [33] the use of nifedipine in the treatment of pulmonary edema is
suggested.

Chronic Mountain Sickness

Chronic mountain sickness [17, 63] (Monge's disease) was first described in 1928.
This rare disease affects people permanently living at altitudes higher than 3000 m in
the Andes but not natives living at corresponding altitudes in the Himalayas. Thus,
genetic factors seem to play a central role although the disease mechanism is un-
known. Monge's disease is characterized by extreme polycythemia, leading to severe
perfusion problems. Visual disturbances, breathlessness, paresthesia, finger club-
bing, and cyanosis are common findings.
Chronic mountain sickness has to be distinguished from pulmonary disease
which, due to hypoxia, might lead to secondary polycythemia and thus induce similar
symptoms. Phlebotomy relieves the symptoms [63], as does descent to sea level.
The clinical picture is quite different from that of high altitude deterioration, the
detrimental effect of hypoxia which anybody who stays at a high altitude for a pro-
longed time may experience [18] (see below).

Fluid and Electrolyte Balance

Proper fluid balance is important, not only in avoiding AMS during the acclimatiza-
tion period, but may represent the single most important medical factor for a suc-
cessful summit attempt at extreme altitude. The fluid loss may be extreme and often
is not noticed by the climber. Especially at high altitude the sensation of thirst is a
late and unreliable symptom of dehydration. The exhaustion combined with practical
difficulties in obtaining enough water make the climber neglect this difficult and im-
portant challenge. Numerous reports indicate that many of the deaths at extreme
altitude directly or indirectly are caused by water deprivation.
At extreme altitude most of the fluid loss is by the airways. The extreme hyper-
ventilation may alone be responsible for a water loss of 4-6 litres per day, out of a
total of 6-8litres (included loss by urine, skin and feces).
High Altitude Physiology and Medicine 333

To some extent dehydration at extreme altitude is almost unavoidable and may

be regarded as part of a normal process of acclimatization, cf. fluid retention in AMS.
However, a more severe dehydration threatens life in various ways. The tendency to
develop AMS is increased, and almost a paradox, especially the risk of developing
pulmonary edema. During acclimatization hematocrit increases and, if in addition,
plasma volume decreases red blood cells pile up with formation of rouleaux, leading
to a dramatic increase in viscosity [63]. Tissue perfusion is reduced and gas exchange
and temperature regulation suffer. The danger of thrombosis and frostbite is over-
whelming.
Earlier, high hematocrit was cited in support of an optimal acclimatization, "the
higher the hematocrit, the better". On the contrary, high hematocrit levels represent
a danger signal to health and life. Excessive increases to levels above 60% are due
not only to increased erythropoiesis but almost certainly also to dehydration.
Practically, fluid balance can be monitored in two ways, subjectively by estimat-
ing the diuresis aiming at maintaining a normal urine output or, objectively by regu-
lar measurements of hematocrit. The use of a plastic urine bottle is recommended,
also at extreme altitudes. A rough estimate of the morning portion is far better than
an "evaluation" of the "yellow cornets" in the snow. Moreover, the possible comfort
of staying within the tent also favours the use of plastic bottles. Hematocrit is easily
measured by a battery-operated minicentrifuge (e.g., Compur M 1100). The mea-
surements not only have a preventive function but also an educational one, making
the climber focus on the importance of proper fluid balance.

Hemodilution

Hemodilution at high altitude has been performed now and then during the last lO-
IS years, most often on German, Austrian and Italian climbers. The attitude has
varied from enthusiasm [65] to considerable scepticism [19, 42, 52]. Obviously, some
misunderstanding is present. First, hemodilution is not done in order to improve the
performance but to prevent danger. Second, hemodilution does not represent any
difficult and dangerous task if done in the right way. The rationale for performing
hemodilution at special occasions seems obvious as hemoconcentration is the most
important pathophysiological threat at extreme altitude. However, the possible posi-
tive or negative outcome is difficult to assess since so many other variables also are
of importance for the final result of an expedition.
Hemodilution can either be done by the mere infusion of fluid (plasma, saline or
a solution of macromolecules), by removing blood, or by combined infusion and re-
moval. Infusion alone has too short-lasting an effect to be of practical help on the
mountain. Combined drawing of blood and infusion of fluid will markedly improve
the effect. Reinfusion of one's own plasma is, however, difficult (high hematocrit,
lack of centrifuge). Colloid osmotic solutions will expand the plasma compartment
but may lead to anaphylactic reactions and should be restricted to bleeding shocks or
similar critical situations. The use of physiological saline or Ringer solutions will
have less effect than plasma expanders as the distribution volume is much larger.
Larger volumes therefore will be needed. On the other hand, an immediate and
334 K.T.Stokke

complete normalization is seldom needed and the dilution may later be completed by
peroral fluid supply. The combined removal of blood and infusion of saline, although
practically difficult "in the field" at extreme altitude, is fairly simple - and it is safe.
Hemodilution will never be a routine on expeditions as it will not improve perfor-
mance [42]. On the other hand, removal of 0.5-1 litre of blood has not been shown
to have any detrimental effect. On the contrary, on the American Medical Research
Expedition to Everest a small but significant improvement of psychological tests was
found [42]. In cases with excessive polycythemia development of thrombosis, frost-
bite, hypothermia and pulmonary edema may be prevented. The opinion that hemo-
dilution is potentially hazardous [19, 42] should not be left undisputed. When done
on the right indications as an antihemoconcentration attempt, the maneuver may be
lifesaving.

Frostbite and Hypothermia

Frostbite and hypothermia represent important threats at extreme altitude. Normal-

ly, the temperature falls by 0.65°C for every 100m of ascent. A temperature of
+15°C at sea level may thus correspond to -40°C at an altitude of 8500m. On the
other hand, the protection of the atmosphere on the heat effect of sunlight will de-
crease, leading to extreme temperature variations day and night.
At a given temperature frostbite is much more common at high altitude than in
lowland areas. Several factors contribute: the chilling effect of the wind is important,
as the wind velocity often is extreme at high altitude. On the other hand, the number
of bombarding molecules is reduced according to the altitude and this works in a
beneficial way. In addition to external factors, the polycythemia (a disadvantageous
effect of the acclimatization process) and the almost unavoidable dehydration, both
contribute to the excessive increase in viscosity and reduced perfusability of the
blood [44]. Moreover, the canalization of blood to the central, most vital organs also
contributes, making frostbite an important problem.
General hypothermia often will co-exist with frostbite. When vital organs are
threatened by hypothermia, unnecessary heat loss is prevented by reduced blood
supply to the extremities. This "physiological amputation" is beneficial by reducing
heat loss and should not be treated until the central organs can tolerate the reversal
of cold blood from the extremities.
Frostbite and hypothermia seem to be almost obligatory on expeditions to ex-
treme altitudes. In principle they should be prevented and can often be avoided.
Treatment follows the same guidelines as in lowland areas, but any attempt will be
practically more difficult. Since dehydration often coexists, fluid supply may be
necessary. At altitudes above 5500 m oxygen will also be beneficial. Pentoxifylline
(Trental) seems to have a positive influence on blood rheology [54] and might be
used in a preventive manner, not only against frostbites but to overcome the multi-
farious effects of impaired microcirculation.
High Altitude Physiology and Medicine 335

Other Pathophysiological Changes at Extreme Altitude

A stay at high/extreme altitude has a deleterious effect on the body. Although a few
people have been living at 5950 m for an extended period of time [58], this seems to
be "the exception which confirms the rule" that nowhere in the world permanent
settlement above 5300m is possible [18]. Above this "magic" altitude there will be an
unavoidable bodily deterioration. Appetite and sleep are disturbed, weight is lost
and the general body condition becomes increasingly worse.
In addition to the somatic changes which also may include AMS with its multi-
farious complex of symptoms and various airways infections which may be difficult
to treat, sensory impairments are also frequent. Night vision is impaired and gustatory
sensitivity may be changed. Mental changes are also frequent. Memory is weakened
as is the capability of decision making. Hallucinations may occur. Mood changes
with unreasonableness in arguments are also frequent. As already mentioned retinal
bleeding occurs in people staying above 6000m [19]. Since the retina may be consid-
ered a visible part of the brain, similar changes within the brain may also take place.
Direct proof is lacking, and the question of full reversibility of the psychomotor
changes has till now not been answered.
No matter, it seems clear that cerebral function is impaired while the subject is at
extreme altitude. It also seems undebatable that an impaired function may persist for
some period of time (more than a year?). It has, however, till now not been proved
that a stay at extreme altitude leads to permanent brain damage.

Medical Observations During the Norwegian Mount Everest Expedition

Although the expedition was planned to be a climbing expedition exclusively, two

scientific investigations were performed on the members.

Changes in Plasma and Red-Cell Volumes

Training at moderate altitude has been used by athletes to improve the performance
at sea level. Alternatively, blood doping has been used to increase blood volume and
thereby to achieve better results [4]. Since not all athletes seem to benefit from
altitude training, one might speculate whether their failure can be explained by an in-
sufficient response in red cell production to hypoxia. Is there an individual variation
with respect to hematologic response to altitude? Although several studies on the ef-
fect of moderate altitude on blood volume have been performed, they are not con-
clusive with respect to this question. The expedition gave an opportunity to perform
a long-term study of the blood volume response at a higher altitude.
Ten members took part in the study. Measurements were performed before
departure from Oslo (sea level) and in Everest Base Camp (5300 m) four weeks later.
Plasma volume was determined by the isotope dilution technique. Exactly calibrated
solutions of 5 pCi (185 kBq) 125I-albumin were given intravenously after 30 min bed
336 K.T. Stokke

Table 1. Age, weight and hematologic data of all participants. Measurements were performed in
Oslo (Om), and in Base Camp (5300m) four weeks later

Subject Age Weight Hematocrit Plasma Blood Red-cell

no. (years) (kg) (%) volume volume volume
Om 5300m Om 5300m Om 5300m Om 5300m

1 40 75 47 55 3000 3550 5130 6900 2130 3350

2 38 72 47 49 2145 2635 3670 4640 1525 2005
3 35 85 42 53 2950 2630 4685 4935 1735 2305
4 47 70 47 50 2895 3020 4950 5405 2055 2385
5 32 69 42 50 2655 2475 4225 4435 1570 1960
6 33 76 43 50 3575 3080 5760 5510 2185 2430
7 29 66 45 50 2595 2655 4300 4750 1705 2095
8 49 74 50 50 2265 2705 4055 4840 1790 2135
9 42 69 47 51 3030 2970 5175 5405 2145 2435
10 33 74 47 53 2770 3350 4730 6295 1960 2945

Mean 38 73 45.7 51.1 2788 2907 4668 5312 1880 2405

The changes in plasma volume are not significant (0.2 < P < 0.4), whereas the changes in blood vol-
ume (0.01 < P < 0.02) and erythrocyte volume (P < 0.001) are both significant (Wilcoxon's test for
pair comparison). Volumes in ml

rest. Blood was drawn from the opposite arm immediately before, and 15 and 30 min
after the injection. Hematocrit was determined by centrifugation. All blood samples
were then hemolyzed by the addition of 2-3 drops of Saponin and sent back to Nor-
way for counting by a conventional gamma-counter. By combining the results of
plasma volume determinations with hematocrit, and also correcting for the ratio be-
tween venous hematocrit and body hematocrit [2], total blood volume and red cell
volume were calculated. The ratio of body hematocrit to venous hematocrit (0.91) is
constant over a wide hematocrit range, and this ratio is the same at high altitude and
at sea level [31].
The results for hematocrit, plasma volume, total blood volume and red-cell vol-
ume of all participants are shown in Table 1 and Fig. 1. Red-cell volume increased in
all participants (range 11%-57%), and excessively (50%-57%) in two of them.
However, during acclimatization to high altitude plasma volume often decreases,
leading to hemoconcentration which may be misinterpreted as increased erythro-
poietic activity [22]. Plasma volume decreased in four of our subjects, making the
change in total blood volume less consistent. In one climber dehydration and re-
duced plasma volume led to a decrease in total blood volume. While a significant
correlation was found between maximal oxygen uptake (measured before departure
by monitoring mixed expired air during cycle ergometry) and red-cell volume at sea
level, the hematologic response to altitude seemed independent of physical fitness.
In conclusion, our study showed an increase in red-cell mass in all subjects and
also a considerable individual variation. Varying degrees of dehydration make data
on total blood volume less simple to interpret. A more extensive report has been pre-
sented elsewhere [48].
High Altitude Physiology and Medicine 337

Hematocrit Blood volume Plasma volume Red- cell volume

60 7.0 ~5 3.5

55 6.0 3.0 3.0

50 5.0 2.5 2.5

45 4.0
~ 2.0 .2.0

40 3.0 1.5 1.5

Om 5300m Om 5300m Om 5300m Om 5300m

Fig.t. Hematocrit, blood volume, plasma volume and erythrocyte volume (in litres) of all partici-
pants. As for hematocrit identical values are marked as one point, whereas the lines show the indi-
vidual changes

Hemodilution

In the climber with the most pronounced erythropoietic response (subject no. 1)
hematocrit gradually increased to a value of 69%. Although later analysis of data on
blood parameters (measurement of blood volume by isotope dilution) revealed an
exaggerated erythropoietic response, some of the hematocrit increase was thought to

8848 m

70 ~
-
0
0

t; 60
/
-\.
t t\.__ !,
450ml
350 ml

0 _-- Drink \
6-7litres

-
-0 _/
.,
E
/---
per day
:r:

-
50 /-

__40~'~~~~'50__L-~'60
~

L.u~.~~~~'~~~'~~
10 20 30
days

Fig. 2. Hematocrit and hemodilution data of subject no. 1. The highest hematocrit value (69%) was
observed in Advanced Base Camp at 6500m. Hemodilution was performed in Base Camp (5300m),
a total of 800 ml of blood being withdrawn. Hematocrit was 52% the day after the ascent to the
summit
338 K.T.Stokke

be due to dehydration. Despite descending to Base Camp (5300m) and despite a

fluid intake of 5-7 litres per day, hematocrit remained high (above 60% - supposed
limit of danger). Hemodilution was performed on two consecutive days, the first day
by withdrawing 450 ml of blood, the second day 350 ml. Infusions of one litre of
physiological saline and one litre of a Ringer acetate solution were given. Before
hemodilution symptoms revealed an impaired peripheral blood circulation, most
pronounced was an extreme feeling of coldness. Hemodilution completely changed
his general condition. Four days after the hemodilution he successfully reached the
summit, and the day after the ascent his hematocrit was 52% . Hematocrit and hemo-
dilution data of subject no. 1 are shown in Fig. 2.
As has already been claimed hemodilution should be considered more often, not
to improve the performance (which it does not) but in order to prevent a catastrophe
(thrombosis, frostbite, hypothermia ... ). Many lives might have been saved by such
intervention.

Cerebral Blood Flow

Retinal hemorrhages [19, 26] and (transient?) impairment of various intellectual

functions [53, 56] (psychomotor performance, verballeaming, short-term memory
etc.) are frequent findings in persons who have been at an altitude of more than
6000m for longer periods. The question "Do climbs to extreme altitude cause brain
damage?" [57] worries a lot of people. Since cerebral blood flow (CBF) reflects cere-
bral metabolism, CBF determination could be used as an objective parameter of
cerebral and cerebellar function.
CBF was measured by 133Xe inhalation tomography (Tomomatic 64) in eight
members of the expedition. Six were experienced high altitude climbers, and five
reached the summit of Mt. Everest. All stayed above 6500 m for a total of approxi-
mately 3 weeks. The reference group consisted of 25 healthy subjects from the hospi-
tal staff. Ten male diving instructors formed a second control group. They were
chosen because they, like the climbers, were familiar with the use of oxygen breath-
ing gear. Measurements were performed in Oslo before and immediately after the
completion of the expedition. Xenon radioactivity was recorded simultaneously in 3
transaxial tomographic slices of brain tissue, each 2 cm thick. Slice 1 passed through
cerebellum, slice 2 contained parts of the frontal, temporal, parietal and occipital
lobes and included also the basal ganglia and the thalamus, while slice 3 transsected
the upper part of the frontal and parietal lobes. The methodology has been described
in more detail elsewhere [39].
The individual flow values are shown in Fig. 3. After the expedition a small, but
not-significant reduction in CBF was observed. The most striking finding was, how-
ever, that already before the expedition took place, CBF was significantly lower than
in control subjects. Measured in per cent the largest flow difference between climbers
and the reference group was seen in mean hemispheric CBF (22% ), and the smallest
difference in cerebellar flow (16%). The average flow reduction was 19% compared
with the reference group and 16% compared with the second control group.
Reinvestigation of seven of the participants one year after the expedition and
four of them two years after showed the same "low" flow values (data not shown).
On both of these occasions their hematocrit values were normal.
High Altitude Physiology and Medicine 339

80 Reference group Expedition

members
A B

C, 70
e
...
0
52
R
c:
! 60 '"
J... e
E "
-e-
e
-"
~
.2
0
""

E
"
. Fig.3. Cerebral blood flow (CBF) measured by
means of Tomomatic 64. Reference group A com-
prised 25 healthy women and men at corresponding

..
0
:0 50
age, reference group B 10 male diving instructors.
e All values are corrected for differences in hemato-
.
..a
I!! crit. Horizontal lines represent the mean values
u

40
eX:
e e
for the respective groups and studies. CBF before
departure significantly lower than that of the refer-
ence groups (A: P<O.OOl; B: P<O.05). CBF after
~ Before After the expedition not significantly lower than flow val-
ues before departure. The flow values shown in this
ol figure are not PC02-corrected

Variations in hematocrit represent a problem in CBF measurements in general,

and in 133Xe CBF measurements in particular. Because of the uncertainties of CBF
corrections for hematocrit and oxygen delivery, capacity measurements performed
one year after the expedition may be the most reliable. At this time the hematocrit
of climbers and reference subjects were not significantly different, and no corrections
were necessary.
The reduced CBF in high altitude climbers should be interpreted with caution.
There are at least two explanations, either that hypoxia has led to long-lasting effects
on CNS neurons and a corresponding reduction in cerebral metabolism, or that cere-
bral metabolism is maintained at the normal level by increased oxygen and glucose
extraction from the blood. Repeated stays at high altitude seem to have no or only a
small additional effect.
The low initial CBF values may be explained by the fact that these measurements
in the climbers were not real pre-high altitude climbing measurements. Only one sub-
ject in the study group had never before been at high altitude and was the participant
with the highest hemispheric CBF, equal to the reference mean before the expedi-
tion. He also had the most marked fall in hemispheric CBF.
CBF was also studied after the i. v. injection of 1 g of acetazolamide (Diamox).
The relative increase in the climbers was higher than in the reference subjects, show-
ing that the functional capacity of the microvascular system of the brain was intact.
This hypernormal response supports the hypothesis of unchanged brain function and
metabolism. Further studies are needed to explore this interesting and intricate field.
Of particular interest might be quantitative EEG measurements which from a differ-
ent "angle" than does CBF, might elucidate whether the reduced CBF values ob-
served is due to reduced metabolism - or is compensated by increased efficiency in
oxygen extraction from the circulating brain blood. A detailed report on CBF in high
altitude climbers has been presented earlier [40].
340 K.T.Stokke

Concluding Remarks

Climbing at high or extreme altitude represents in every respect a challenge. The

body must gradually adapt to hypoxia and this may seem to require an unnecessary
long time. When acclimatization fails, acute mountain sickness is the punishment.
Being acclimatized, any stay at extreme altitude has to be as short as possible, as
bodily deterioration will inevitably take place. No climber should be allowed to stay
at an extreme altitude for more than a few days. He then should be forced down for
rest at "low" altitude « 5300m). The importance of sufficient fluid supply cannot be
overemphasized, so adequate fluid intake and regular rest periods at low altitude
may be the two most important rules for staying healthy at extreme altitude [32].
Climbing mountains higher than 8000 m without supplementary oxygen means an
even greater challenge, as thrilling as Russian roulette. There is no reason to believe
that this special kind of sport has any positive influence on health and well-being.
According to Houston [20] "mountaineers have much to teach those who are less
blessed". May we hope that climbing at extreme altitude without extra oxygen will
tell us more about the physiology and pathophysiology of hypoxia. May we also hope
that we shall not learn from this outre kind of sport that especially the brain is ex-
tremely sensitive to hypoxia.

References

1. Anonymous (1979) Proteinuria at high altitude. Br Med J 1: 508-509

2. Besa EC (1975) Physiological changes in blood volume. CRC Crit Rev Clin Lab Sci 6 : 67-79
3. Bradwell AR, Dykes PW, Coote JH, Forster PJE, Milles JJ, Chesner I, Richardson NV (1986)
Effect of acetazolamide on exercise performance and muscle mass at high altitude. Lancet 1:
1001-1005
4. Buick FJ, Gledhill N, Froese AB, Spriet L, Meyers EC (1980) Effect of induced erythrocytemia
on aerobic work capacity. J Appl Physiol: Respirat Environ Exercise Physiol 48 : 636-642
5. Cerretelli P (1976) Oxygen transport on Mount Everest: the effects of increased hematocrit on
maximal O 2 transport. Adv Exp Med BioI 75: 113-119
6. Clarke C (1985) High-altitude cerebral oedema. In: Rivolier J, Cerretelli P, Foray J, Segantini
P (eds) High altitude deterioration. Karger, Basel, pp 103-109
7. Duhm J (1982) Effects of high altitude (low arterial P0 2) and of displacements of the oxygen dis-
sociation curve of blood on peripheral O 2 extraction and P02 • In: Brendel W, Zink RA (eds)
High altitude physiology and medicine. Springer, Berlin Heidelberg New York, pp 54-65
8. Garruto RM, Dutt JS (1983) Lack of prominent compensatory polycythemia in traditional
native Andeans living at 4,200 meters. Am J Phys Anthropol 61: 355-366
9. Hackett PH (1980) Acute mountain sickness - the clinical approach. Adv CardioI27:6-1O
10. Hackett PH, Rennie D, Grover RF, Reeves IT (1981) Acute mountain sickness and the edemas
of high altitude: a common pathogenesis? Respir Physiol46: 383-390
11. Hackett PH, Rennie D, Hofmeister SE, Grover RF, Grover EB, Reeves IT (1982) Fluid reten-
tion and relative hypoventilation in acute mountain sicknes. Respiration 43: 321-329
12. Hackett PH, Reeves IT, Grover RF, Weil JV (1984) Ventilation in human populations native to
high altitude. In: West JB, Lahiri S (eds) High altitude and man. Williams and Wilkins, Balti-
more, pp 179-191
13. Hackett PH, Roach RC, Harrison GL, Schoene RB, Mills WJ Jr (1987) Respiratory stimulants
and sleep periodic breathing at high altitude. Almitrine versus acetazolamide. Am Rev Respir
Dis 135: 896-898
High Altitude Physiology and Medicine 341

14. Hamilton AJ, Cymmerman A, Black PMcL (1986) High altitude cerebral edema. Neurosurgery
19:841-849
15. Hartley LH, Alexander JK, Modelski M, Grover RF (1967) Subnormal cardiac output at rest
and during exercise in residents at 3,100m altitude. J Appl Physiol23: 839-848
16. Hauge A, Nicolaysen G, Thoresen M (1983) Acute effects of acetazolamide on cerebral blood
flow in man. Acta Physiol Scand 117: 233-239
17. Heath D, Williams DR (1977) Adaptation and acclimatization. In: Heath D, Williams DR (eds)
Man at high altitude. The pathophysiology of acclimatization and adaption. Churchill Living-
stone, Edinburgh, pp 225-235
18. Heath D, Williams DR (1977) Exposure to extreme altitudes. In: Heath D, Williams DR (eds)
Man at high altitude. The pathophysiology of acclimatization and adaption. Churchill Living-
stone, Edinburgh, pp 247-256
19. Houston CS (1983) Going higher: the story of man and altitude. Queen City, Burlington, Vermont
20. Houston CS (1985) Research in altitude illness. In: Rivolier J, Cerretelli P, Foray J, Segantini P
(eds) High altitude deterioration. Karger, Basel, pp 170-179
21. Johnson TS, Rock PB, Fulco CS, Trad LA, Spark RF, Maher JT (1984) Prevention of acute
mountain sickness by dexamethasone. N Engl J Med 310: 683-686
22. Kaung DT, Peterson RE (1962) "Relative polycythemia" or "pseudopolycythemia". Arch
Intern Med 110: 456-460
23. Larson EB, Roach RC, Schoene RB, Hornbein TF (1982) Acute mountain sickness and acet-
azolamide. Clinical efficacy and effect on ventilation. JAMA 248: 328-332
24. Lenfant C, Sullivan K (1971) Adaptation to high altitude. N Engl J Med 284: 1298-1309
25. Masuyama S, Kimura H, SUgita T, Kuriyama T, Tatsumi K, Kunitomo F, Okita S, Tojima H,
Yuguchi Y, Watanabe S, Honda Y (1986) Control of ventilation in extreme-altitude climbers. J
Appl Physiol 61: 500-506
26. McFadden DM, Houston CS, Sutton JR, Powles ACP, Gray GW, Roberts RS (1981) High-
altitude retinopathy. J AMA 245 : 581-586
27. McIntosh IB, Prescott RJ (1986) Acetazolamide in prevention of acute mountain sickness. J Int
Med Res 14:285-287
28. Meehan RT, Cymerman A, Rock P, Fulco CS, Hoffman J, Abernathy C, Needleman S, Maher
IT (1986) The effect of naproxen on acute mountain sickness and vascular responses to hypoxia.
Am J Med Sci 292: 15-20
29. Milledge JS (1983) Acute mountain sickness. Thorax 38:641-645
30. Moore LG, Harrison GL, McCullough RE, McCullough RG, Mieco AJ, Tucker A, Weil JV,
Reeves IT (1986) Low acute hypoxic ventilatory response and hypoxic depression in acute
altitude sickness. J Appl Physiol60: 1407-1412
31. Myhre LG, Brown DK, Hall FG, Dill DB (1968) The use of carbon monoxide and T-1824 for
determining blood volume. Clin Chern 14: 1197-1205
32. Oelz 0 (1982) How to stay healthy while climbing Mount Everest. In: Brendel W, Zink RA
(eds) High altitude physiology and medicine. Springer, Berlin Heidelberg New York, pp 298-300
33. Oelz 0 (1987) A case of high-altitude pulmonary edema treated with nifedipine. JAMA 257: 780
34. Oelz 0, Howald H, di Prampero PE, Hoppeler H, Claassen H, Jenni R, Biihlmann A, Ferretti
G, Bruckner J-C, Veicsteinas A, Gussoni M, Cerretelli P (1986) Physiological profile of world-
class high-altitude climbers. J Appl Physiol60: 1734-1742
35. Palareti G, Coccheri S, Poggi M, Tricarico MG, Magelli M, Cavazzuti F (1984) Changes in the
rheologic properties of blood after a high altitude expedition. Angiology (July) 451-458
36. Pugh LGCE (1964) Cardiac output in muscular exercise at 5,800m (19,000 ft). J Appl Physiol
19:441-447
37. Rennie D (1982) Water retention in the pathophysiology of acute mountain sickness. Bull Mem
Acad R Med Belg 137:627-640
38. Reynafarje B (1962) Myoglobin content and enzymatic activity of muscle and altitude adapta-
tion. J Appl PhysioI17:301-305
39. Rootwelt K, Dybevold S, Nyberg-Hansen R, Russell D (1986) Measurement of cerebral blood
flow with 133Xe inhalation and dynamic single photon emission computer tomography. Normal
values. Scand J Clin Lab Invest [Suppl] 46(184) :97-105
40. Rootwelt K, Stokke KT, Nyberg-Hansen R, Russell D, Dybevold S (1986) Reduced cerebral
blood flow in high altitude climbers. Scand J Clin Lab Invest [Suppl] 46(184): 107-112
342 K. T. Stokke: High Altitude Physiology and Medicine

41. Saltin B, Grover RF, Blomqvist CG, Hartley LH, Johnson RL Jr (1968) Maximal oxygen up-
take and cardiac output after 2 weeks at 4,300m. J Appl Physiol 25: 400-409
42. Sarnquist FR, Schoene RB, Hackett PH, Townes BD (1986) Hemodilution of polycythemic
mountaineers: effects on exercise and mental function. Aviat Space Environ Med 57:313-317
43. Schmid-Schonbein H (1982) Blood rheology in hemoconcentration. In: Brendel W, Zink RA
(eds) High altitude physiology and medicine. Springer, Berlin Heidelberg New York, pp 109-116
44. Schmid-SchOnbein H, Neumann FJ (1985) Pathophysiology of cutaneous frost injury: disturbed
microcirculation as a consequence of abnormal flow behaviour of the blood. Application of new
concepts of blood rheology. In: Rivolier J, Cerretelli P, Foray J, Segantini P (eds) High altitude
deterioration. Karger, Basel, pp 20-38
45. Schoene RB (1984) Hypoxic ventilatory response and exercise ventilation at sea level and high
altitude. In: West JB, Lahiri S (eds) High altitude and man. Williams and Wilkins, Baltimore,
pp 19-30
46. Schoene RB, Lahiri S, Hackett PH, Peters RM Jr, Milledge JS, Pizzo CJ, Sarnquist FR, Boyer
SJ, Graber DJ, Maret KH, West JB (1984) Relationship of hypoxic ventilatory response to exer-
cise performance on Mount Everest. J Appl Physiol: Respirat Environ Exercise Physiol 56:
1478-1483
47. Schoene RB, Hackett PH, Henderson WR, Sage EH, Chow M, Roach RC, Mills WJ Jr, Martin
TR (1986) High-altitude pulmonary edema. Characteristics of lung lavage fluid. JAMA 256: 63-69
48. Stokke KT, Rootwelt K, Wergeland R, Vale JR (1986) Changes in plasma and red cell volumes
during exposure to high altitude. Scand J Clin Lab Invest [Suppl] 46(184): 113-117
49. Sutton JR, Lassen N (1979) Pathophysiology of acute mountain sickness and high altitude pul-
monary oedema: a hypothesis. Bull Eur Physiopathol Respir 15: 1045-1052
50. Sutton JR, Houston CS, Mansell AL, McFadden MD, Hackett PH, Rigg JRA, Powles ACP
(1979) Effect of acetazolamide on hypoxemia during sleep at high altitude. N Engl J Med 301:
1329-1331
51. Sutton JR, Gray GW, McFadden MD, Houston CS, Powles ACP (1982) Sleep hypoxemia at
altitude. In: Brendel W, Zink RA (eds) High altitude physiology and medicine. Springer, Berlin
Heidelberg New York, pp 3-8
52. Sutton JR, Jones NL, Pugh LGCE (1983) Exercise at altitude. Ann Rev Physiol45 :427-437
53. Townes BD, Hornbein TF, Schoene RB, Sarnquist FR, Grant I (1984) Human cerebral function
at extreme altitude. In: West JB, Lahiri S (eds) High altitude and man. Williams and Wilkins,
Baltimore, pp 31-36
54. Ward A, Clissold SP (1987) Pentoxifylline. A review of its pharmacodynamic and pharmaco-
kinetic properties, and its therapeutic efficacy. Drugs 34: 50-97
55. West JB (1983) Climbing Mt.Everest without oxygen: an analysis of maximal exercise during
extreme hypoxia. Respir Physiol 52: 265-279
56. West JB (1984) Human physiology at extreme altitudes on Mount Everest. Science 223: 784-788
57. West JB (1986) Do climbs to extreme altitude cause brain damage? Lancet 2:387-388
58. West JB (1986) Highest inhabitants in the world. Nature 324: 517
59. West JB, Wagner PD (1980) Predicted gas exchange on the summit of Mt.Everest. Respir
Physiol 42: 1-16
60. West JB, Boyer SJ, Graber DJ, Hackett PH, Maret KH, Milledge JS, Peters RM Jr, Pizzo CJ,
Samaja M, Sarnquist FR, Schoene RB, Winslow RM (1983) Maximal exercise at extreme
altitudes on Mount Everest. J Appl Physiol: Respirat Environ Exercise Physiol 55: 688-698
61. West JB, Hackett PH, Maret KH, Milledge JS, Peters RM Jr, Pizzo CJ, Winslow RM (1983)
Pulmonary gas exchange on the summit of Mount Everest. J Appl Physiol: Respirat Environ
Exercise Physiol 55 : 678-687
62. West JB, Lahiri S, Maret KH, Peters RM Jr, Pizzo CJ (1983) Barometric pressures at extreme
altitudes on Mt. Everest: physiological significance. J Appl Physiol: Respirat Environ Exercise
Physiol 54: 1188-1194
63. Winslow RM (1984) High-altitude polycythemia. In: West JB, Lahiri S (eds) High altitude and
man. Williams and Wilkins, Baltimore, pp 163-172
64. Winslow RM (1984) Red cell function at extreme altitude. In: West JB, Lahiri S (eds) High
altitude and man. Williams and Wilkins, Baltimore, pp 59-72
65. Zink RA, Schaffert W, Messmer K, Brendel W (1982) Hemodilution: practical experiences in
high altitude expeditions. In: Brendel W, Zink RA (eds) High altitude physiology and medicine.
Springer, Berlin Heidelberg New York, pp 291-297
Chapter 29
Heart and Brain Under Hyperbaric Conditions in Man
A.O.BRUBAKK

Introduction

Man can work and function in many unfavorable environments. By using proper pro-
cedures and protective equipment he is able to protect himself from immediate harm
and injury. He cannot, however, completely eliminate all effects of that environ-
ment. These effects must be understood to exploit man's unique abilities and at the
same time allow him to escape harm.
The main characteristic of the hyperbaric environment is the increase in environ-
mental pressure above 100 kPa. Although man has dived successfully to an equiva-
lent depth of 680 msw, we still have an inadequate understanding of the mechanisms
that are responsible for the functional changes observed and the possible aftereffects
of this experience.
At present, routine diving is performed down to depths of approximately 200msw.
In the near future, dives down to 300-400 msw will have to be performed to develop
the subsea resources that have been found on the continental shelves.

Diving Methods

Surface-Based Diving

At depths down to approximately 50m, the diver can dive down from the surface.
This form of diving is the one used by scuba divers carrying their own air supply. For
serious work, however, it is severely limited, due to the short bottom times. Another
severe limit is the narcotic effect of nitrogen (see later), which represents a problem
at depths exceeding 30m. Thus, helium is usually substituted for nitrogen as the inert
gas in the breathing mixture for deeper dives. In the deepest range, hydrogen has
been considered due to its lower density.

Saturation Diving

This is the form of diving most used in commercial operations. The divers stay under
pressure for a considerable period of time (up to 3 weeks). At depths in excess of
50 m, helium is used instead of nitrogen as the inert gas, and oxygen partial pressure
344 A.O.Brubakk

is usually in the range or 40-50 kPa. An initial compression phase brings the diver to
the bottom. When the diver performs his work, the bottom phase is followed by a
saturation decompression phase to the surface. During the bottom stay the diver is
allowed to move freely inside a certain pressure range (excursions).

Effects of Increased Pressure

Brain and Nervous System

Down to a depth of approximately 180 msw, the diver can be compressed rapidly
without any apparent ill effects and with practically no functional impairment. If he
goes deeper, the so-called high-pressure nervous syndrome (HPNS) will occur to a
larger or smaller extent in all divers. This syndrome is characterized by dizziness,
nausea, vomiting, postural and intentional tremor, fatigue, and somnolence accom-
panied by poor sleep with nightmares, myoclonic jerking, stomach cramps, and re-
duction in intellectual and psychomotor performance. EEG shows a decrease in a-
activity, with an increase in slow waves [7]. This syndrome was first demonstrated in
1964-1965 during a series of oxygen helium dives down to 180-240m [5] and was
initially termed "helium tremors." However, in animal experiments, Brauer [10]
showed that the changes seen were most probably a direct effect of the increased
pressure and that the inert gases only played a modifying role. He also demonstrated
that the final stage was convulsions.
Both the absolute pressure and the rate of pressure change play a role in the
development of this syndrome. In a series of experiments, Brauer et al. [12, 13] dem-
onstrated that the total compression time, more than the rate of compression im-
mediately prior to reaching depth, is the important variable.
Usually the symptoms of HPNS are most severe immediately upon reaching the
bottom, and improvement will usually occur during stay at depth. However, both
EEG changes and tremor can be observed during the whole period at pressure and
even months after the dive [43, 51]. Figure 1 shows an example of EEG changes ob-
served during and after a dive to 500m using a helium/oxygen mixture.
Although the clinical picture of HPNS is quite well described, the underlying
mechanisms are still obscure. It has been known for quite some time that pressure is
able to counteract the effects of anesthesia [37]. Thus, the HPNS has been regarded
as a state of hyperactivity of the central nervous system. This hyperactivity probably
stems from subcortical structures, while the cortex only plays a modifying role [8,
14]. A reduction in noradrenaline levels [33] and an interaction with GABAergic
mechanisms [26] probably playa role in this.
Pressure has a direct effect upon the membranes of excitable tissue [54] and prob-
ably also on all membranes in the body [36, 41]. As is demonstrated in Fig. 2, an in-

Fig. 1. Distribution of EEG frequencies in diver 3 before, during, and after a dive on heliox to 500
msw. Note the decrease in center frequency at pressure and after the dive. (Data from Ellertsen et
al. [23])
Heart and Brain Under Hyperbaric Conditions in Man 345

LEFT SIDE RIGHT SIDE

5- 5.

>- >--
l-
....
III
...en
I-

Z Z
w w 0
0 0
.J .J
< <
a:: a::
l- I-
U U
W W
Q. Q.
Ul (I)

5. 10. 15. 20. 25. :30. 5. 10. 15. 20. 25.

1. 7.

>-
...
l-
(I)
..z
>-
I-
til

J
Z
w w
0 c
.J ...I
+ 70 minutes
< <
a:: a:: 198
l- I-
U U
W UJ
Q. Q.
IJ) Ul

0.
0- 5. HI. 15. 20. 25. :30. 5. 10. 15. 20. 25. :311l.

4. 4.

>- >-
....
l- I-
....
en til
z Z
w w
0 0
.J .J
+ 27 hours
< <
a:: a:: 500
l- I-
U U
w W
Q. Q.
Ul til

121.
121. 5. H'. 15. 2121. 25. :30. 5. 10. 15. 20. 25. 3121.

5. 5.

> >
.......
III
...III....
Z Z Postdive.
W W
0 0 0
-I -I
< <
....a: a:
....
U U
W w
n. n.
III III

5. 10. 15. 20. 25. :30. 5. 10. 15. 2111. 25. :3111.

FREClUENCY (HZ. ) FREClUENCY (HZ. )

346 A.O.Brubakk

-----!.----- 25 AIm.

50 AIm.
35
30
25
20
~ 1~
5
a 0 3OOAlm.
a.. --1~

~Jl~500Alm.
-~J
-10
-15
+ ~--- 600Alm
Fig. 2. Changes in resting potential (elec-
trical potential difference) in isolated frog
skin with increases in environmental pres-
sure. Cp, pressure application; D, pressure
I I I I release. (Pequeux and Gilles [41])
I
o 2 3 5 7 min.

Change in SEP at pressure

85.-----------------------,
D P3
80 ... P1
75
70
u 65
~ 60
c7c 55
-t 50 Fig. 3. Changes in peripheral nerve con-
-' 45 duction time (PI) and intracerebral con-
40 duction time (P3) during compression
35 and at pressure. (Data from Simpson et
30+-----,-------.----r---~ al. [46])
o 540 615 650 686
Depth, msw

crease in pressure will lead to an increased potential difference across the cell mem-
brane, probably caused by an increased permeability to sodium. This change is al-
ready apparent at pressures where diving has been performed.
While hyperexcitability can be recorded in the central nervous system and in
neurons, pressure seems to inhibit synaptic transmission [2]. This has also been ob-
served in humans. In a study performed during a dive to 686 msw on a helium/
nitrogen/oxygen mixture (Trimix), Simpson et al. [46] showed that the transmission
time from the popliteal nerve to the brain was delayed at pressure, while the intra-
cerebral conduction times were decreased. Figure 3 shows the maximum intracerebral
conduction time (difference between Pl and P3 wave) taken from this study. The in-
Heart and Brain Under Hyperbaric Conditions in Man 347

Compression strategies
0-300 msw
320
300
2BO
260
240
220
200
lBO
~ 160
E
140
120
100
BO
60
[J 19B3-proflle
40 + 19B4-proflle
20 <> 19B6-proflle

0
o 200 400 600
minutes

Fig.4. Three different decompression profiles used by a diving company during the time period
1983-1986

crease in conduction time at pressure can either be caused by pressure as such or be

an effect of nitrogen (see later).
In man, several different methods have been used to counteract the effects of
pressure. Due to the large difference in individual susceptibility, it is probably pos-
sible to select "resistant" individuals based on previous deep-diving experience [44].
The actual compression profile used will significantly influence the clinical symptoms
observed. Figure 4 shows three different profiles that have been used for commercial
diving operations using heliox. Although experience is limited, the 1986 profile will
lead to less HPNS symptoms than the 1983 profile, indicating that preventing symp-
toms to appear by allowing a period of adaptation in the 180-m zone offers some pro-
tection against HPNS at deeper depths.
Based on the observation that the narcotic effect of nitrogen could be reversed by
increasing the partial pressure of helium [52], it was suggested that adding a small
amount of nitrogen (5% -10%) to the breathing gas could control HPNS. Practical
testing of this concept both in the Deep Ex series of dives in Norway [52, 53], and in
the Atlantis dives in the United States [7], showed that the addition of nitrogen was
particularly effective in removing the tremor, while other symptoms still persisted.
Furthermore, as one of the divers in the Atlantis IV dive had psychotic symptoms
and two divers in the Deep Ex 81 dive had severe hallucinations upon reducing the
nitrogen content, some side effects of nitrogen narcosis cannot be discounted. It
must furthermore be borne in mind that the elimination of symptoms does not mean
a reversal of the pressure effects.
 348 A.O.Brubakk

Heart and Circulation

The increased hydrostatic pressure will lead to bradycardia due to a direct effect of
pressure upon the heart [35]. The effect is however small. Studies on rat atrial muscle
showed an increased cardiac contractility under pressure, with no change in fre-
quency [3]. These changes occurred at pressures as low as O.5mPa.
In the intact organism, the direct pressure effects will be masked by other effects
of the gas mixture used. As there is an increase in gas density with increasing pres-
sure, breathing resistance will increase, and thus CO 2 retention can occur. In most
diving operations an increased pressure of oxygen will be used. In a dive to 18.6ATA
(1.86mPa) using heliox, Smith et al. [47] showed a slight initial bradycardia and an
increase in resting cardiac output. During the dive, cardiac output tended to fall and
was significantly reduced after the dive. This can be caused both by the observed re-
duction in plasma volume and by hyperoxia. These changes are small and probably
have little practical importance. Figure 5 shows a recording obtained in a diver par-
ticipating in a 300-m dive; the flows were calculated from ultrasonic Doppler curves.
As can be seen, no significant changes occurred.

DIVER 4. DEEP-EX 80.

Day 1-111 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 1 10 1 11 11211311411511611 +1 1

Depth m. o~
100
200
300
% change 140
130
120
110
100
90
80
70 _ _ _ _e Heart rate
60 .._ .._ ...... Blood pressure
50 ._._.-.. Cardiac output
- Total peripheral resistance
FEMORAL ARTERY
% change 200

100

50
- Peripheral resistance

Fig.S. Changes in cardiovascular variables during a heliox dive to 300msw in one diver. Flow data
were calculated from Doppler ultrasonic measurements. Blood pressure was measured by the Riva
Rocci method. Peripheral resistance is the pressure/flow ratio. All data given as percentage of pre-
dive values (100%)
Heart and Brain Under Hyperbaric Conditions in Man 349

In some animal studies, local changes in blood flow have been shown as a result
of the increased pressure. In rats, Onarheim and Tyssebotn [39] showed an increase
in cerebral blood flow at 0.5 mPa.
Some data indicate that regulatory mechanisms are influenced by pressure. An
increase in the sensitivity of the respiratory system to CO2 had been observed after
long saturation dives [42], and an abolition ofthe bradycardia response to submersion
has also been described [16].

Effects of Increased Partial Pressure of Gas

Inert Gas

As was mentioned earlier, when diving deeper than approximately 30 m, air cannot
be used as a breathing gas. The reason for this is the narcotic effect of nitrogen at
high pressure, giving rise to euphoria, intoxication, and eventually narcosis. All inert
gases have some narcotic effect, with hydrogen having a narcotic effect of approxi-
mately 50% of nitrogen and helium one of 25% [6]. The most likely explanation for
the narcotic effect is a direct interaction between the lipids in the brain and the inert
gas as there is a close correlation between narcotic effect and lipid solubility.

Oxygen

During a normal diving operation, the diver will not be exposed to oxygen pressures
higher than 40-50 kPa. No central nervous effects have been described after such an
exposure for long periods of time. At partial pressures of 230-250 kPa, central ner-
vous symptoms have been seen, ranging from facial pallor, behavior changes, con-
striction of the visual field, and muscle twitching to convulsions [22]. It is important
to note that convulsions may occur suddenly without any other symptoms and that
there are large differences in susceptibility, both among individuals and in the same
individual at different times. In susceptible individuals, convulsions have been de-
scribed after breathing oxygen for 72 min at 189 kPa [18]. An increase in CO2 content
and exercise will reduce the threshold for convulsions. Data from animal experi-
ments indicate that there is a synergistic effect between oxygen and high pressure
[11], an effect that also lowers this threshold.

Effects of Other Substances

In the hyperbaric environment, the diver is exposed to a number of substances for a

prolonged period of time. The effect of this is largely unknown, but will in the future
have to be considered. A study on the uptake and elimination of toluene, an organic
solvent, in the rat showed that the uptake was significantly influenced by the hyper-
baric environment [38]; as is shown in Fig. 6.
350 A.O.Brubakk

30 100

80

20
10

40
10

20

0 0
11 21 31 41 11 21 31 41
ATA ATA
Blood Liver

Cone. (mg/kg) Cone. (mg/kg)

V
100

10

~
80

10 40

40

20
20

o~~--n~---21--~31--~~~ 0~~--1·1----21--~31--~~~

ATA ATA
Kidney Testis

80 Conc. (mg/kg) 2200 Conc. (mg/k )

10

40

20
V 2000

1800

1100

0 1400
n 21 31 41 11 21 31
ATA ATA
Brain Perirenal fat

Fig. 6. Changes in the concentration of toluene in different organs in the rat at increasing hydrostatic
pressures: 8-h exposure to 3.75 mg toluene!1. (Nilsen et a1. [38])

Effects of Increased Gas Uptake

During the diver's stay at pressure, there will be an increased uptake of gas which
must be eliminated upon returning to the surface. Ideally, this decompression should
be performed in a way that prevents gas from coming out of solution. Many different
Heart and Brain Under Hyperbaric Conditions in Man 351

Symptoms of neurological DeS

Fig. 7. Typical symptoms of serious (type II)

decompression sickness. (Data from Elliott
and Kindwall [24])

strategies have been developed for doing this. The first set of modern decompression
tables were developed by Boycott et al. [9] in 1908. They developed the concept that
bubble formation was determined by the degree of supersaturation and that the up-
take and elimination of inert gas were symmetrical. Based on theoretical and experi-
mental studies, they determined that a level of supersaturation existed under which
no bubbles would form. This was set to be equa1 to a 50% pressure reduction. Al-
though these procedures were very effective in reducing the incidence of decompres-
sion sickness, subsequent studies showed, in particular from deeper dives, that a con-
siderable amount of bubble formation would occur. Today it is assumed that all prac-
tical decompression schedules will lead to bubble formation [19].
It is generally agreed that decompression sickness is caused by gas bubbles. The
use of Doppler ultrasound has demonstrated that there can be a significant amount
of gas bubbles in the venous system without any symptoms of decompression sick-
ness [34]. These bubbles are termed "silent bubbles", and most investigators claim
that they have no pathophysiological significance as the lung is considered to be a
very effective filter [17]. If a large load of bubbles is presented to the lung or a rup-
ture of the alveoli occurs, gas may escape into the arterial system and lead to infarc-
tion of the central nervous system. Although many instances of serious decompres-
sion sickness follow this pattern, many discrepancies have been observed.
Figure 7 shows an overview of the symptoms seen in serious decompression sick-
ness. The majority of lesions found at autopsy are seen in the white matter of the spi-
nal cord, a site that does not conform with embolization from the heart. This has led
several investigators to conclude that the changes in the spinal cord can be caused by
gas bubble formation in the epidural vertebral venous system [29, 40]. At least in the
spinal cord, bubbles in the venous system can therefore have serious consequences.
Most of the decompression tables used today allow the diver to perform repeated
dives. The US Navy procedures [50] allow the diver to regard the second dive as a
part of the first one if the second is performed after less than 30 min surface time. Be-
tween 30min and 12h the second dive is regarded as a repetitive dive, requiring
slower decompressions. After 12h, enough excess gas is considered eliminated to
perform new dives without any restrictions. These procedures are based on models
 352 A.O.Brubakk
dB.H. 34 yrs
PEEP EX I
1. ASCENT
PULMONARY ARTERY FEMORAL VEIN
COMMON CAROTID ARTERY
llJ> IJ I~

r" II
~
~"

em /sec 50-
Vmax -..J k ~k~ U . 1f.. 1h I ~ I ~
0-

.+
50-
Vmean
0- ~ ~ At If.. I ~ l ~ · ! I

At 300m At 250m At 300m At 250m

1 1/2 Hours after 1 112 Hours after 1 1/2 Hours after

decompression decompressioo decompression

Fig. 8. Gas bubbles recorded from the femoral vein, pulmonary artery, and carotid artery in a diver
performing an ascending excursion from 300 to 250msw. The bubbles are recorded as high-intensity
events using ultrasonic Doppler equipment

of inert gas elimination which may change unpredictably when bubbles are present.
Large differences in gas elimination between different subjects have been demon-
strated: gas bubbles have been cited as the cause both for an increase [32] and a de-
crease [20] in elimination rate. Even in the same subject, large differences in the gas
elimination rate have been seen [21].
The time course of bubble occurrence in the central venous system shows large
individual differences (Brubakk, Matre, Getz, unpublished) and bubbles, once they
are formed, can be observed for long periods of time after decompression [19, 40]
(days to weeks). Furthermore, it has been demonstrated both in mice [27] and in
guinea pigs [25] that several repeated dives have a much higher mortality than a
single dive with a bottom time equal to the sum of the bottom times and the surface
interval. The mechanism for this is probably that the bubbles formed during the first
ascent are compressed and moved into areas where they can cause more damage
when they expand during the second decompression. There seems to be a relation-
ship between the weight of the animal and the critical time period between dives as
the interdive period giving the highest mortality is approximately four times higher
in the guinea pig than in the mouse. The weight ratio between a mouse and a guinea
pig is approximately 1: 10. In a man of about 75 kg, this will give a critical interdive
period of 20 h.
The data given above indicate that even following normal procedures, gas bubbles
in the vascular system of the nervous system may well be present. During a dive to
300 msw, gas bubbles were observed in the arterial system after excursions to 250
msW in all divers [15]. In Fig. 8 an example of bubbles recorded in this dive can be
seen. Similar results have been observed on excursions from 450 to 420msw in
another dive (Brubakk, unpublished). During decompression from a deep saturation
dive, bubbles in the carotid artery were observed in one diver [30]. These observa-
tions lead to the conclusion that the lung will not always be effective as a filter for
bubbles.
Heart and Brain Under Hyperbaric Conditions in Man 353

Effects of Gas Bubbles

As described above, the result of severe violation of decompression procedures will

be central nervous system damage, in particular of the white matter of the spinal
cord. Most modern decompression procedures, if adhered to, will prevent this. For
divers performing decompressions repeatedly, it is of importance to consider if small
amounts of bubbles can be harmful.
Following open heart surgery, Spencer et al. [48] concluded that small amounts
of gas bubbles did not give rise to any apparent neurological damage. Based upon
animal studies, Grulke and Hills [28] stated that bubbles with a diameter of less than
100!1 had no measurable effect.
It is, however, known that bubbles will serve as an interface for the aggregation
of thrombocytes [49]. Gas bubbles damage the endothelial layer of the arteries and
can lead to increased vessel permeability [34]. Bubbles between 10 and 20!1 diameter
will traverse the cerebral circulation and lead to marked increase in the permeability
of smaller veins even hours after the event [31]. Thus, possible decremental effects
of gas bubbles, in particular after repeated exposures, cannot be excluded.

Long-term Effects of Diving

It has until now been assumed that routine diving following accepted procedures will
not lead to any damage to the central nervours system. However, it has been de-
scribed that a considerable number of divers show changes in personality such as loss
of memory, aggressive behavior, excessive tiredness, difficulties in concentration,
and reduced tolerance for alcohol [4]. These "soft" clinical signs are difficult to inter-
pret, but cannot be disregarded as they describe changes in an otherwise young and
healthy population group.
In connection with the deep experimental dives performed in Norway, extensive
clinical and neurophysiological testing have been performed. Following dives to 350
msw, Aarli et al. [1] demonstrated that 4 of 23 divers showed clinical signs of focal
central nervous dysfunction. These signs were only present immediately after the
dive, and the possibility was discussed tat the dive could "unmask" previous minor
head traumas. These data are collected from three separate dives. In the first to 350
msw (n = 6), type I decompression sickness occurred in two divers during saturation
decompression to the surface. None of the divers in this dive showed any focal
changes. In another chamber dive to 350 msw (n = 6), no decompression sickness oc-
curred, but these divers performed ascending excursions of up to 10m. In this dive,
two divers showed focal changes. In the third, an open sea dive to 300msw (n = 11),
no decompression problems were described, but again lO-m excursions were per-
formed. After this dive, changes were seen in two divers. Gas bubble formation will
be more pronounced during excursions, indicating that bubbles may playa role in
the changes seen after the dive.
Investigations performed after a subsequent dive to greater depths showed a
similar pattern, with changes being more pronounced and longer lasting. In one
354 A.O.Brubakk

diver the possibility of permanent, focal changes cannot be excluded. In this dive, 30-
m excursions were performed with gas bubbles being observed in the carotid artery
in five of six divers (Brubakk, unpublished).
Until now, these changes have only been observed after deep experimental dives.
However, few divers have been subjected to such extensive investigations. Using
magnetic resonance imaging (MRI), focal changes were observed in the brain of 4 of
21 commercial divers [45]. These divers had extensive diving experience, but none
had experienced any episodes of decompression sickness.
All these data seem to indicate that the hyperbaric environment may lead to per-
manent changes in the central nervous system of divers who have never experienced
any accidents. The cause of these changes is not known, but the effect of gas bubble
formation seems to be a possibility.

Conclusions

The hyperbaric environment has significant effects upon the central nervous system,
both during and after the dives. The direct effect upon the heart and circulatory sys-
tem is less pronounced. We still have only limited understanding of the mechanisms
involved. The most important question that has to be answered is whether routine
diving can lead to permanent changes in the central nervous system.

References

1. Aarli JA, Vremes R, Brubakk AO, Nyland H, Skeidsvoll H, T9lnjum S (1985) Central nervous
dysfunction associated with deep-sea diving. Acta N eurol Scand 71 : 2-10
2. Ashford MU, Macdonald AG, Wann KT (1982) The effects of hydrostatic pressure on the
spontaneous release of transmitter at the frog neuromuscular junction. J Physiol (Lond) 333:
531-543
3. Ask JA, Tyssebotn I (1986) Positive inotropic effects of rat myocardium compressed to 5, 10,
and 30 Bar. In: Schrier LM (ed) Diving and hyperbaric medicine. Proceedings XII Annual
Meeting EUBS, Rotterdam, pp 65-70
4. Baddely A (1983) Discussion. In: Shields TG, Minsaas B, Elliott DH, McCallum RI (eds) Long
term neurological consequences of deep diving. EUBS, Stavanger, pp 164-167
5. Bennett PB (1965) Psychometric impairment in men breathing oxygen-helium at increased pres-
sure. Underwater Physiology Subcommittee Report no 251. MRC RNPRC, London
6. Bennett PB (1982) Inert gas narcosis. In: Bennett PB, Elliott DH (eds) The physiology and
medicine of diving. Bailliere Tindall, London, pp 239-261
7. Bennett PB, McLeod M (1984) Probing the limits of human deep diving. Philos Trans R Soc
Lond [Bioi] 304: 105-117
8. Bowser-Riley F (1984) Mechanistic studies of the high pressure neurological syndrome. Philos
Trans R Soc Lond [Bioi] 304:31-41
9. Boycott AE, Damant GCC, Haldane JS (1908) The prevention of compressed air illness. J Hyg
8:342-443
10. Brauer RW (1968) Seeking man's depth level. Ocean Ind 3: 28-33
11. Brauer RW, Beaver RW (1982) Synergism ofhyperoxia and high helium pressures in the causa-
tion of convulsions. J Appl Physiol 53: 192-202
12. Brauer RW, Beaver RW, Lahser S, Mansfield WM, O'Connor F, White L (1975) Rate factors
in the development of the high pressure nervous syndrome. J Appl Physiol 38: 220-227
Heart and Brain Under Hyperbaric Conditions in Man 355

13. Brauer RW, Beaver RW, Lahser S, Mansfield WM, Sheehan ME (1977) Time, rate and temper-
ature factors in the onset of high pressure convulsions. J Appl Physiol43: 173-182
14. Brauer RW, Mansfield WM, Beaver RW, Gillen HW (1979) Stages of development of high
pressure neurological syndrome in the mouse. J Appl Physiol46: 756-765
15. Brubakk AO, Peterson R, Grip A, Holand B, Onarheim J, Segadal K, Kunkle TD, T~njum S
(1986) Gas bubbles in the circulation of divers after ascending excursions from 300 to 250msw.
J Appl Physiol60:45-51
16. Brubakk AO, Pasche A (1986) Changes in the diving response after saturation dives. In:
Brubakk AO, Kanwisher JW, Sundnes G (eds) Diving in animals and man. Tapir, Trondheim,
pp 283-299
17. Butler BD, Hills BA (1979) The lung as a filter for microbubbles. J Appl PhysioI47:537-543
18. Butler FK, Thalmann ED (1984) CNS oxygen toxicity in closed-circuit scuba divers. In: Bacharach
AJ, Matzen MM (eds) Underwater physiology VIII. Undersea Medical Society, Bethesda, pp
15-30
19. Daniels S (1984) Ultrasonic monitoring of decompression. Philos Trans R Soc Lond [Bioi] 304 :
153-175
20. D'Aoust BG, Smith KH, Swanson HT (1976) Decompression-induced decrease in nitrogen
elimination in awake dogs. J Appl Physiol 41 : 348-355
21. Dick APK, Vann RD, Mebane GY, Feezor MD (1984) Decompression induced nitrogen elimi-
nation. Undersea Biomed Res 11: 369-380
22. Donald KW (1947) Oxygen poisoning in man. I & II. Br Med J 1: 667-672, 712-717
23. Ellertsen B, Hammerborg D, Lindrup A, Roby J, Vrernes R (1982) Central nervous reactions
during Deep Ex 81. NUTEC Report No 7-82. NUTEC, Bergen
24. Elliott DH, Kindwall EP (1982) Manifestations of the decompression disorders. In: Elliott DH,
Bennett PB (eds) The physiology and medicine of diving. Bailliere Tindall, London, pp 461-472
25. Gait DJ, Miller KW, Paton WDM, Smith EB, Welch B (1975) The redistribution of vascular
bubbles in multiple dives. Undersea Biomed Res 2: 42-49
26. Gran L, Coggin R, Bennett PB (1980) Diazepam under hyperbaric conditions in rat. Acta
Anaesthesiol Scand 24: 407 -411
27. Griffiths HB, Miller KW, Paton WDM, Smith EB (1971) On the role of separated gas in decom-
pression procedures. Proc R Soc Lond [Bioi] 178 : 389-406
28. Grulke DC, Hills BA (1978) Experimental cerebral air embolism and their resolution. In: Schil-
ling CW, Beckett MW (eds) Underwater physiology VI. FASEB, Bethesda, pp 587-594
29. Hallenbeck JM, Bove AA, Elliott DH (1975) Mechanisms underlying spinal cord damage in de-
compression sickness. Neurology 25: 308-316
30. Hjelle JO (1987) Arterial bubbles during decompression in a deep saturation dive. Undersea
Biomed Res [Suppl]14 : 6
31. James PB, Hills BA (1987) Micro-embolism, multiple sclerosis and the perivenous syndrome.
Lancet (in press)
32. Kindwall EP, Bax A, Lightfoot EN, Lanphier EH, Seireg A (1975) Nitrogen elimination in man
during decompression. Undersea Biomed Res 2:285-297
33. Koblin DD, Little HJ, Green AR, Daniels S, Smith EB, Paton WDM (1980) Brain monoamines
and the high pressure neurological syndrome. Neuropharmacology 19: 1031-1038
34. Levin LL, Stewart GJ, Lynch PR, Bove AA (1981) Blood and blood vessel wall changes induced
by decompression sickness in dogs. J Appl Physiol 50: 944-949
35. Lundgren CEG, 0rnhagen H (1976) Heart rate and respiratory frequency in hydrostatically
compressed, liquid breathing mice. Undersea Biomed Res 3: 303-320
36. Macdonald AG (1984) The effects of pressure on the molecular structure and physiological func-
tions of cell membranes. Philos Trans R Soc Lond [Bioi] 304:47-68
37. Miller KW, Paton WDM, Smith RA, Smith EB (1973) The pressure reversal of anaesthesia and
the critical volume hypothesis. Mol Pharmacol 9: 131-143
38. Nilsen A, Zahlsen K, Nilsen OG (1987) Uptake, distribution, accumulation and elimination of
toluene in the rat. Report no STF23 A87013. SINTEF, Trondheim
39. Onarheim J, Tyssebotn I (1980) Effect of high ambient pressure and oxygen tension on organ
blood flow in the anesthetized rat. Undersea Biomed Res 7: 47-60
40. Palmer AC, Blackmore WF, Payne JE, Sillence A (1978) Decompression sickness in the goat:
nature of brain and spinal cord lesions after 48 hours. Undersea Biomed Res 5: 275-286
356 A.O.Brubakk: Heart and Brain Under Hyperbaric Conditions in Man

41. Pequeux A, Gilles R (1986) Effect of hydrostatic pressure on ionic and osmotic regulation. In:
Brubakk AO, Kanwisher JW, Sundnes G (eds) Diving in animals and man. Tapir, Trondheim,
pp 161-188
42. Pasche A, Paciorek J, Hauge A, Segadal K (1982) Ventilatory CO2 response following deep
saturation dives. In: Elliott D (ed) Diving and hyperbaric medicine. Proceedings VIII Annual
Meeting EUBS, Lubeck, pp 281-309
43. Rostain JC, Lemaire C, Gardette-Chauffour MC, Doucet J, Naquet R (1983) Estimation of
human susceptibility to the high-pressure nervous syndrome. J Appl Physiol 54: 1063-1070
44. Rostain JC, Naquet R (1978) Human neuropsychologic data obtained from two simulated dives
to a depth of 610 m. In: Shilling Beckett MW (eds) Underwater physiology VI. FASEB, Bethesda,
pp 52-60
45. Shields TG, Smith FW (1987) MRI observations in a random group of divers. In: Medical re-
search council workshop. Diagnostic techniques in diving neurology. MRC, London (in press)
46. Simpson DM, Harris DJ, Bennett PB (1983) Latency changes in the human somatosensory
evoked potential at extreme depths. Undersea Biomed Res 10: 107-114
47. Smith RM, Hong SK, Dressendorfer RH, Dwyer HJ, Hayashi E, Yelverton C (1977) Hana Kai
II: a 17-day dry saturation dive to 18.6 ATA. IV. Cardiopulmonary functions. Undersea Biomed
Res 4:267-281
48. Spencer MP, Lawrence GA, Thomas GI, Sanvage LR (1969) The use of ultrasonics in the deter-
mination of arterial air embolism during open heart surgery. Ann Thorac Surg 8: 489
49. Thorsen T, Brubak A, 0vstedal T, Farstad M, Holmsen H (1986) A method for production of
N2 microbubbles in platelet-rich plasma in an aggregometer-like apparatus, and the effect on the
platelet density in vitro. Undersea Biomed Res 13: 271-288
50. US Navy diving manual (1980) NAVSEA 0994-LP-OOl-9020, Washington DC
51. Vrernes RJ (1983) Reversible and possible irreversible CNS changes of deep diving. A discussion
of some empirical studies. In: Shields TG, Minsaas B, Elliott DH, McCallum RI (eds) Long
term neurological consequences of deep diving. EUBS, Stavanger, pp 31-47
52. Vrernes R, Bennett PB, Hammerborg D, Ellertsen B, Peterson RE (1982) Central nervous sys-
tem reactions during heliox and trimix dives to 31 ATA. Undersea Biomed Res 9: 1-14
53. Vrernes R, Hammerborg D, Ellertsen B, Peterson R, T!IInjum S (1983) Central nervous reac-
tions during heliox and trimix dives to 51 ATA, DEEP EX81. Undersea Biomed Res 10: 169-192
54. Wann KT, Macdonald AG (1980) The effect of pressure on excitable cells. Comp Biochem
Physiol 66A: 1-12
Chapter 30

Effects of Positive End-Expiratory Pressure Ventilation

on Intracranial Pressure and Cerebral Blood Flow
O.HEVR0Y, N.-E.KL0W, 0. NYGAARD, and J.H.TRUMPY

Use of Positive End-Expiratory Pressure Ventilation

The interplay between physiological dynamics relating to cardiac, pulmonary, and

central nervous system function is of great concern during intensive care of the
neurosurgical patient. Patients suffering from multiple trauma including intracranial
injuries frequently develop acute respiratory failure requiring mechanical ventilation
[12]. In addition, mechanical ventilation is often needed in comatose patients follow-
ing head trauma without other injuries. Pulmonary complications with increased
venous admixture are also reported in these patients [6].
Application of positive end-expiratory pressure (PEEP) during mechanical venti-
lation in patients with acute respiratory failure may be useful. PEEP improves re-
gional lung ventilation and reduces the ventilation-perfusion imbalance in diseased
lungs and may relieve life-threatening arterial hypoxia without use of toxic oxygen
concentrations [3].

Side EtTects of Positive End-Expiratory Pressure

During PEEP ventilation, cardiac output (CO) is reduced due to decreased venous
return and increased pulmonary vascular resistance lowering the left ventricular pre-
load [19]. The cardiovascular effects of PEEP depend on transmission of the ele-
vated airway pressure to the pleura and to the heart, and this is affected by lung com-
pliance [19]. These hemodynamic side effects have to be considered carefully when
PEEP ventilation is used in critical care.

EtTects of Positive End-Expiratory Pressure on Intracranial Pressure

Increased intracranial pressure (ICP) during PEEP ventilation has been reported
both in clinical and experimental studies [1, 2, 7, 10, 12, 14, 15,20]. PEEP increases
the pleural and thereby the central venous pressure (CVP) [3]. Increased impedance
to cerebral venous outflow raises intracranial venous pressure and blood volume,
and also ICP [17]. This is a major concern when PEEP is to be used in patients fol-
lowing head trauma [1, 2]. It has been suggested that the increased ICP caused by
358 O. Hevrllly et al.

PEEP may seriously reduce the cerebral perfusion pressure (CPP) and cerebral
blood flow (CBF) in patients with intracranial lesions [1, 16]. However, conflicting
results have been presented, and the magnitude of the increase in ICP due to PEEP
ventilation varies considerably between different studies [1, 5, 12,20].

Previous Studies

Aidinis et al. [1] studied the effects of PEEP in cats with artificially created intracra-
nial masses. Noncardiac pulmonary edema was induced in some of the animals. Cats
with normal lungs had significant decreases in CPP due to increased ICP, resulting
in electroencephalographic abnormalities in 77% when end-expiratory pressure of
15 cmH20 was applied. Cats with decreased lung compliance due to pulmonary
edema had significantly less reduction in CPP, and none of these animals showed
EEG changes during PEEP ventilation.
Also Huseby et al. [15] studied the effects of PEEP in dogs with pulmonary
edema induced by oleic acid. They found that ICP increased when end-expiratory
pressure was raised from 0 to 20 cmH20 although pulmonary compliance was low. In
their study this resulted in seriously reduced CPP.
Shapiro and Marschall [20] studied the effects of PEEP in 12 patients with head
injuries. ICP was monitored using a subarachnoid bolt or an intraventricular catheter.
End-expiratory pressure ranged between 4 and 8cmH20. ICP increased by 10mmHg
or more in six patients, and CPP decreased significantly. They concluded that in pa-
tients with decreased intracranial compliance, PEEP should be used carefully and al-
ways with monitoring of ICP.
Frost [12], however, studying the effects of PEEP on ICP in seven comatose and
brain-injured patients, found no increase in ICP when PEEP was applied even to
levels of 40cmH20, and Cooper and Boswell [5] recently reported a slight, though
clinically insignificant increase in ICP when PEEP was applied in patients following
head trauma, concluding that end-expiratory pressure of 10 cmH20 was safe to use
in such patients.

Effects of Positive End.Expiratory Pressure on Cerebral Blood Flow

We have studied the effects of PEEP on ICP and CBF in dogs with normal and arti-
ficially elevated ICP.
Seven dogs were anesthetized with sodium pentobarbital and placed in the right
supine position. The head was kept at the heart level. The dogs were ventilated
through a cuffed endotracheal tube by a volume-controlled ventilator, and ventila-
tion was adjusted to keep PaC02 within the normal range. Mean aortic blood pres-
sure (MAP) and central venous pressure (CVP) were measured through fluid-filled
catheters. ICP was measured through a flat fluid-filled (0.1 ml) balloon placed sub-
durally over the right hemisphere through a burr hole. All pressures were measured
by Statham 23 dB pressure transducers.
Effects of Positive End-Expiratory Pressure Ventilation 359

CBF was measured by the hydrogen (Hz) de saturation technique [4, 18] using a
platinum electrode placed in the saggital sinus through a midline burr hole. CO was
measured by thermodilution technique. Burr holes were closed by wax before the
start of recordings. Measurements were made at 0, 10, and 20 cmHzO end-expiratory
pressure with normal ICP. In five of the dogs ICP was raised by inflation of a balloon
placed subdurally over the right hemisphere, and measurements were repeated at
the same PEEP levels. CPP was calculated as MAP minus ICP.
Data are presented as means ± SD. Differences were evaluated by analysis of
variance for repeated measurements and/or Student's t test. P < 0.05 was considered
significant. Linear regression analysis was used to evaluate correlation between two
variables.

Results

By inflating the balloon over the right hemisphere, ICP increased from 11 ± 4 mmHg
to 25 ± 3mmHg (Fig. 1). This was accompanied by a slight decrease in heart rate,
while MAP and CO remained unchanged (Table 1). CPP decreased from 124 ± 10
mmHg to 110 ± lOmmHg (Table 1). CBF decreased slightly but not significantly.
When PEEP was applied, CO decreased significantly in both groups at both
PEEP levels (Table 1). ICP increased significantly when PEEP was applied both in
dogs with normal and artificially elevated ICP (Fig. 1). The increase was not signifi-
cantly different between dogs with normal ICP compared with dogs with elevated
ICP. The increases in ICP correlated well with the increase in CVP induced by PEEP
in both groups (Fig. 2). MAP was not significantly reduced by PEEP, and the slight
increase in ICP caused by PEEP did not result in any significant reduction in CPP in
either group (Table 1).
CBF decreased when PEEP was applied both in dogs with normal and dogs with
elevated ICP compared with zero end-expiratory pressure (Fig. 3). The decreases

40
DZEEP
III PEEP10 *
~PEEP20

Fig.t. Increases in intracranial pressure (Iep) induced

by ventilation with positive end-expiratory pressure
(PEEP) in dogs with normal and dogs with elevated
ICP. ZEEP, zero end-expiratory pressure. Values are
o means ± SD. *P<O.05
Normal Elevated
ICP ICP
Table 1. Hemodynamic effects of PEEP during normal and elevated ICP VJ
~
NormalICP Elevated ICP
ZEEP PEEP lO PEEP20 ZEEP PEEP lO PEEP20
Heart rate 160 ±25 159 ±25 155 ±20 139 ±31 150 ±25 159 ±24
MAP (mmHg) 135 ± 9 138 ± 7 124 ±13 137 ±1l 137 ±1l 130 ±26
CPP (mmHg) 124 ±1O 125 ± 7 113 ± 16 110 ± lOb 109 ±11 101 ±24
CVP (mmHg) 1 ± 2 5 ± 2' 8 ± l' 2 ± 1 5 ± 1a 9 ± 2a
CO (I . min-I) 4.6 ± 1.0 3.8± 0.7' 2.3 ± 0.6' 4.5 ± 1.3 3.5 ± 0.8 a 2.3 ± 0.8 a

ap<0.05 compared with ZEEP; bp<0.05 compared with baseline level at normal ICP.
ICP, intracranial pressure; ZEEP, zero end-expiratory pressure; PEEP lO • 20 , 10 and 20cmH20 end-expiratory pressure; MAP, mean aortic pressure ; CPP,
cerebral perfusion pressure; CVP, central venous pressure ; CO , cardiac output. Values are means ± SD

<: ... 61CP 61CP

0
8~ I!l~ ~
;(bg~ I\) Co) .j>. I\) Co) .j>.
01 01
Cerebral Blood Flow 5.~~~
(percent change)
g. q-o ~
I I aa~~[ ~ m
-: ~ ~
(JI PI) ~ 5. I!l o· ,0 p 00
0,"
o (JI o ~
I\)
ION
! I\) g~
~. ~ ~ ::r
::r (b ~
o 0.. -..~ '
• • ;;
<bc-C"'ic-
<:'< • .. ~• •
r:::::lO ::;l(b .j>. [> .j>.
~ ~ [> • •
S~~~ () •
.
i i! [~:::::::;' ~ 0..tT1p>(b ()
< _tT1::s(b t> t> t>
~

it ~ [
::::::::: 5 () '" 0.. : . ~(l) t> ~ (l)
• t> t> t>
~ ~ S· 5' ~
t> 9
.o
~
~~~
Q Vl t> •
t> ::c:
" "II rJl ~ ~ (Xl (Xl t>
m
~. t>
.. ..
m
m
m
m
N'TI ...."TJ
~ ::r- n
S· S· o 0
~
'<
~ [ *" *" I I
t>
:::::::::::::ym:mm:jU ~f!d ::s -0 g ..... ..... N N ~
o o m m
m m
~ ~, §" - . . ~
Effects of Positive End-Expiratory Pressure Ventilation 361

were significant only at the highest PEEP level used. CBF decreased slightly more in
response to PEEP in dogs with normal ICP, but no significant differences between
groups were observed.

Discussion

It is well established that ICP increases in response to raising end-expiratory pressure

through its mechanical effect on cerebral venous drainage [17], although the impor-
tance of this effect has been disputed [1, 2, 5,12,20]. This mechanism is also respon-
sible for increases in ICP observed during abdominal compression and Valsalva
maneuver, and during direct venous compression in the neck (Queckenstedt's test).
When intracranial compliance is reduced, as in patients with intracranial masses or
cerebral edema, a greater increase in ICP in response to increased airway pressure
has been suggested [1, 20].
In the present study ICP increased similarly when PEEP was applied both in dogs
with normal and dogs with artificially elevated ICP. Therefore, we believe that the
increase in intracranial venous blood volume caused by the elevated airway pressure
was less pronounced when intracranial compliance was reduced by balloon inflation.
Although ICP increased significantly when PEEP was applied both in dogs with
normal and dogs with elevated ICP, the increases were small, and no significant re-
ductions in CPP due to PEEP could be shown. Still, CBF was reduced by 16% and
31% at 10 and 20cmHzO of PEEP, respectively, in dogs with normal ICP and by
10% and 29% in dogs with raised ICP. This is in harmony with the observations by
Doblar et al. [10] who found that CBF, measured by electromagnetic flow probe, de-
creased by 32% when 15 cmHzO of PEEP was applied in goats. Gioia et al. [13] re-
cently reported an 18% reduction in CBF, measured by the radiolabeled micro-
sphere technique, when 13cmH20 of end-expiratory pressure was used in dogs.
The mechanism by which CBF is reduced during PEEP ventilation has not been
established. By autoregulation, CBF is normally kept relatively constant over a wide
range of perfusion pressures and cardiac outputs [9]. Following head trauma the
autoregulation may be distorted, CBF becoming directly dependent on CPP [20]. In
the present study, CPP was not significantly influenced by PEEP in either group. CO
decreased significantly when PEEP was applied, but no significant correlation was
observed between reductions in CO and CBF. Reductions in CBF during mechanical
ventilation with high levels of PEEP are therefore probably a direct cerebrovascular
response to the elevated intrathoracic pressure.
Interpretation of these findings depends on the accuracy of CBF measurements.
It is possible that the saggital sinus drains blood from extracranial structures as well
as the brain, and the flow measured from the saggital sinus may not be the true CBF.
In dogs, obstruction to venous outflow via the saggital sinus has been found to divert
blood through alternative cerebral venous drainage pathways [8]. This may be true

..
Fig.3. Reductions in cerebral blood flow caused by PEEP ventilation in dogs with normal and dogs
with elevated intracranial pressure (Iep). Values are means ± SD. *P< 0.05
362 O.Hevrflly et al.

also when venous outflow pressure is increased by PEEP and when saggital sinus
integrity is disturbed by catheter insertion. However, in spite of this, the H2 washout
in the saggital sinus probably reflects the changes in CBF in the individual dog [4].
In general, CBF may be reduced as a result of reduced CPP or increased cerebro-
vascular resistance [9]. As no significant reduction in CPP was observed, PEEP ven-
tilation probably increased the vascular resistance in the brain due to increased CVP.
The exact mechanism by which vascular resistance is increased when PEEP is
applied has not been settled. The following theories may be suggested:
1. The existence of venoarterial reflexes, i.e., arteriolar vasoconstriction in re-
sponse to venodilation, has been observed in several vascular beds [10]. How-
ever, flow regulation in the brain by this mechanism has not been shown.
2. It has also been suggested that increased CVP increases interstitial fluid pressure
and causes swelling of the brain, reducing the cerebral microvascular blood flow
[8,10,21]. The importance ofthis response in the range of venous pressures seen
during PEEP ventilation has not been established [11].
3. Another possible explanation may be cerebral vasoconstriction elicited by lung
distention mediated by nervous reflex mechanisms.
Based on the present study and data earlier presented, it is not possible to con-
clude how PEEP ventilation reduces CBF in dogs. In patients with head injuries with
or without elevated ICP it remains to be shown whether CBF is reduced in response
to PEEP ventilation or not.
Although several studies have been conducted, the clinical importance of the side
effects of PEEP ventilation in patients with intracranial injuries is still controversial
[5]. A slight increase in ICP is probably harmless in most patients. However, a major
reduction in CBF in patients with severe intracranial lesion or expansive masses may
be disadvantageous [20].

Summary and Conclusions

Mechanical ventilation with positive end-expiratory pressure (PEEP) has been

shown to increase intracranial pressure and thereby possibly reduce cerebral perfu-
sion pressure. We have studied these effects of PEEP ventilation in dogs both during
normal and artificially elevated intracranial pressure. PEEP increased intracranial
pressure in both groups, but increases were small, and no decrease in cerebral perfu-
sion pressure was found. However, PEEP reduced cerebral blood flow, probably
due to increased cerebral vascular resistance.
These possible serious side effects of PEEP ventilation have to be considered
carefully during mechanical ventilation of patients with head injury or following intra-
cranial surgery. Cautious use of high levels of PEEP, i.e., above lOcmH20, is recom-
mended.
Effects of Positive End-Expiratory Pressure Ventilation 363

References

1. Aidinis SJ, Lafferty J, Shapiro HM (1976) Intracranial responses to PEEP. Anesthesiology 45:
275-286
2. Apuzzo MU, Weiss MH, Petersons V, Small B, Kurze T, Heiden JS (1977) Effect of positive
end expiratory pressure ventilation on intracranial pressure in man. J Neurosurg 46: 227-232
3. Asbaugh DG, Petty TL (1972) Positive end-expiratory pressure. Physiology, indications and
contraindications. J Thorac Cardiovasc Surg 65: 165-170
4. Aukland K, Bower BF, Berliner RW (1964) Measurements of local blood flow with hydrogen
gas. Circ Res 14: 164-187
5. Cooper KR, Boswell PA, Choi SC (1985) Safe use of PEEP in patients with severe head injury.
J Neurosurg 63: 552-555
6. Cooper KR, Boswell PA (1983) Reduced functional residual capacity and oxygenation in pa-
tients with severe head injury. Chest 84: 29-35
7. Cotev S, Paul WL, Ruiz BC, Kuck EJ, Modell JH (1981) Positive end-expiratory pressure
(PEEP) and cerebrospinal fluid pressure during normal and elevated intracranial pressure in
dogs. Intensive Care Med 7: 187-191
8. Cuypers J, Matakas F, Potolicchio SJ (1976) Effects of central venous pressure on brain tissue
pressure and brain volume. J Neurosurg 45: 89-94
9. D'Alcey L (1972) The cerebral circulation. In: Ruch TC, Patton HD (eds) Physiology and bio-
physics. Saunders, Philadelphia, pp 262-276
10. Doblar DD, Santiago TV, Kahn AU, Edelman NH (1981) The effects of positive end-expiratory
pressure ventilation (PEEP) on cerebral blood flow and cerebrospinal fluid pressure in goats.
Anesthesiology 55: 244-250
11. EkstrlZlm-Jodal B (1970) Effect of increased venous pressure on cerebral blood flow in dogs.
Acta Physiol Scand [Suppl] 350 : 51-61
12. Frost EAM (1977) Effects of positive end-expiratory pressure on intracranial pressure and com-
pliance in brain-injured patients. J Neurosurg 47: 195-200
13. Gioia FR, Harris AP, Traystman RJ, Rogers MC (1986) Organ blood flow during high-frequency
ventilation at low and high airway pressure in dogs. Anesthesiology 65: 50-55
14. Hoffmann P, Schockenhoff B, Wauquier A (1986) Intrakranieller Druck und Kreislaufpara-
meter bei erhohtem endexpiratorischen Druck (PEEP). Anaesthesist 35: 721-733
15. Huseby JS, Pavlin EG, Butler J (1978) Effects of positive end-expiratory pressure on intra-
cranial pressure in dogs. J Appl Physiol 44: 25-27
16. Johnston IH, Rowan JO, Harper AM, Jennett WB (1973) Raised intracranial pressure and
cerebral blood flow. J Neurol Neurosurg Psychiatry 36: 161-170
17. Luce JM, Huseby JS, Kirk W, Butler J (1982) Mechanism by which positive end-expiratory pres-
sure increases cerebrospinal fluid pressure in dogs. J Appl PhysioI52:231-235
18. Michenfelder JD, Messick JM, Theye RA (1968) Simultaneous cerebral blood flow measured by
direct and indirect methods. J Surg Res 8: 475-481
19. Quist J, Pontoppidan H, Wilson RS (1975) Hemodynamic response to mechanical ventilation
with PEP. Anesthesiology 42: 45-55
20. Shapiro HM, Marshall LF (1978) Intracranial pressure responses to PEEP in head-injuries pa-
tients. J Trauma 18: 254-256
21. Wagner EM, Traystman RJ (1983) Cerebral venous outflow and arterial microsphere flow with
elevated venous pressure. Am J PhysioI244:H505-512
Chapter 31
Positive End-Expiratory Pressure and Cardiac Function:
The Role of Extraventricular Constraint
O.A.SMISETH, I. KINGMA, N.W. SCOTI-DOUGLAS, E.R.SMITH, and J.Y.TYBERG

The clinical usefulness of positive end-expiratory pressure ventilation (PEEP) is

limited by reductions in both blood pressure and cardiac output. PEEP leads to a de-
crease in stroke volume at an increased left ventricular (LV) filling pressure [2].
Thus, by conventional ventricular function analysis, LV contractility apparently de-
creases. The precise mechanism(s) responsible for this decreased cardiac perfor-
mance have not been resolved although several hypotheses have been advanced.
These include:
1. Increased right ventricular afterload secondary to increased airway pressure and
lung volume [10];
2. decreased contractile state of the myocardium possibly due to release of a cir-
culating agent [6] or due to stimulation of reflexes that inhibit the vasomotor
center [2];
3. restriction to venous return due to increased extraventricular constraint [9] or de-
creased myocardial compliance [3].
The disagreement concerning the relative importance of each of these proposed
mechanisms is related to the problem of precisely measuring extraventricular con-
straint. In the present paper we will present data which demonstrate that a flat,
liquid-containing pericardial balloon accurately measures extraventricular con-
straint. However, an open-ended fluid-filled pericardial catheter seriously underesti-
mates extraventricular constraint unless the pericardium contains more than the
physiological amount of fluid. Pleural surface pressure and esophageal pressure are
also inadequate measures of extraventricular constraint. By accurate measurement
of pericardial surface pressure it can be demonstrated that the downward and right-
ward shift of the LV stroke work - LV diastolic pressure (LVEDP) relation with
PEEP is only a reflection of increased extraventricular constraint and does not re-
flect depression of LV contractility. Finally, we will present data demonstrating that
right atrial pressure is an accurate reflection of pericardial surface pressure, thereby
providing a means to calculate LV transmural pressure or "true" LV preload.

Measurement of Extraventricular Constraint

The ideal measure of extraventricular constraint is intrapericardial pressure. How-

ever, it is important to appreciate the distinction between "liquid pressure" and "sur-
face pressure" [1]. Liquid pressure can be easily measured in any fluid by an open-
Positive End-Expiratory Pressure and Cardiac Function 365

pericardium present

calculated
pericardial pressure

~pericardium removed

Diameter of the left ventricle

Fig.l. Definition of calculated pericardial pressure. Schematically, the upper curve represents LV
diastolic pressure vs. diameter with the pericardium closed, and the lower curve that after the
pericardium had been removed. The lower curve therefore represents the directly measured LV
transmural pressure, and the upper curve represents the sum of LV transmural pressure and a pres-
sure caused by the presence of the pericardium. The decrease in LV pressure (at a given diameter)
after removal of the pericardium is defined as the calculated pericardial pressure. P LV = LV pressure;
[from 13]

ended catheter. The concept of surface pressure is more difficult; it is the force per
unit surface area exerted by, for example, the LV surface on the overlying parietal
pericardial membrane. We wanted to determine whether pericardial surface pres-
sure could be accurately measured. The experiments were done in open-chest anes-
thetized dogs [13].
To define a standard of reference against which the measured values of pericar-
dial pressure could be compared, we assumed that at the endocardial surface at end-
diastole, LV intracavitary pressure (PLV) was balanced by the sum of LV transmural
pressure (PLVTm ) and pericardial surface pressure (PP) (i.e., PLV = PLVTm + PP).
Therefore PP = PLV - PLVTm at any given LV diameter. In other words, we postu-
lated that the correct pericardial surface pressure is the pressure which must be
added to LV transmural pressure to equal LV intracavitary pressure at a common LV
volume. In these experiments we could measure LV transmural pressure directly. At
the end of the experiment, after removal of the pericardium and with the lungs re-
tracted, pericardial pressure equalled zero so that intracavitary LV pressure was, by
definition, equal to LV transmural pressure. Thus, we defined this calculated
pericardial pressure as the difference between LV end-diastolic pressure measured
with the pericardium closed and the LV end-diastolic pressure measured at the same
diameter with the pericardium open.
Figure 1 shows schematically how this calculated value of pericardial pressure
was obtained. LV anteroposterior and septal to free-wall diameters were recorded by
sonomicrometry and LV pressure by a micromanometer-tipped catheter. Pericardial
liquid pressure was also recorded, using a liquid-filled open-ended catheter. The
pericardium was sealed, and pressures were recorded with 0-50ml of saline in the
pericardium. Intravenous saline infusion was used to expand the vascular volume
and produce significant pericardial constraint.
 ,
366 O.A.Smiseth et al.

l ~ §
~~ 1
15
~ T
t T
c,
J:
E
.5
? 1

f
w
a: 10
::J
en

t
en
w
a:
0.
...J
4(
C o Calculated value
a: 5
4(
U o Balloon
~ e Catheter
W

f
0.
mean ±SE
n=6

0
0 10 20 30 40 50
PERICARDIAL FLUID (ml)
Fig.2. Measured and calculated pericardial pressures at various pericardial fluid volumes. The flat
balloon measures pressures that approximate the calculated values. The open-ended catheter, how-
ever, measures pressures substantially below the calculated value (P<0.05) unless the pericardium
contains at least 30 ml of fluid. LVEDP was kept at 20 mm Hg by adjusting an intravenous saline infu-
sion. (0), calculated value; (0) balloon; (e) catheter; number = 6; mean ± SEM; [from 13]

Figure 2 compares pericardial pressure recorded by each of these techniques with

the calculated value. Note that the flat balloon recorded a pressure which was similar
to the calculated value regardless of the amount of fluid in the pericardium. The
open catheter, however, recorded a pressure which was significantly less than the
calculated value unless the pericardium contained at least 30 ml of fluid. These data
indicate that such a pericardial balloon may be used to measure extraventricular con-
straint. However, an open-ended catheter seriously underestimates extraventricular
constraint unless the pericardium contains a substantial fluid volume. Pericardial
surface pressure equals liquid pressure plus the "contact stress" [11]. In our study
[13], as fluid was infused into the pericardium, liquid pressure increased rapidly and,
with 30-50ml in the pericardium, it approached the surface pressure (Fig. 2). When
sufficient fluid is infused and the heart floats freely, there is no contact stress and
liquid pressure equals surface pressure.
In previous studies of the effects of PEEP on LV function different measures of
extraventricular constraint have been used. From the data presented above it should
be clear that extracardiac pressure recorded by an open-ended catheter seriously
underestimates external constraint. The same objection applies to the use of a micro-
manometer inside an open-ended catheter; although free of transmission artifacts it
still records liquid pressure [9]. Since previous studies have often used pleural sur-
face pressure (PPL) or esophageal pressure (PES) to assess extraventricular con-
Positive End-Expiratory Pressure and Cardiac Function 367

PRESSURE
(mm Hg)
20

10

o
(mm Hg)
20

10

o
LOW VOLUM E HIGH VOLUME

Fig.3. Comparison of end-diastolic pressures during baseline conditions (upper left), following vol-
ume loading with i.v. saline (upper right), with 20mmHg PEEP without volume loading (lower left) ,
and with 20mmHg PEEP after volume loading (lower right) . P LV = LV pressure; PPE = pericardial
surface pressure; PPL = pleural juxtacardiac surface pressure; PES = esophageal pressure; mean ±
SD; number = 9

straint, we undertook an experimental study in which we compared these measures

to pericardial surface pressure, a better measure of extraventricular constraint [8].
The experiments were done in acutely instrumented closed-chest dogs. We used bal-
loons identical to those described above to record PP (over the anterolateral LV sur-
face) and PPL (lateral to the pericardium overlying the pericardial balloon), and an
air-containing cylindrical balloon catheter to record PES (the balloon was placed in
the esophagus just cephalad to the site where maximum cardiac artifact was re-
corded) . The pressures were compared during control conditions and following
blood volume expansion with intravenous saline with and without PEEP. Figure 3
shows the effects of 0 and 20 mm Hg PEEP and volume loading. PPL was always less
than PP indicating that PPL does not represent an accurate assessment of extraven-
tricular constraint. The transmural pericardial pressure (PP-PPL) increased with vol-
ume loading due to pericardial distension secondary to cardiac dilation. However,
with increasing level of PEEP transpericardial pressure decreased as PPL increased
despite the fact that PP also increased. This implies that with 20mmHg PEEP extra-
ventricular constraint is accounted for predominantly by mechanical compression by
the expanding lungs. Transmural LVEDP, reflecting LV preload, decreased with
PEEP.
We also found a significant increase in PPL secondary to volume loading
(P < 0.05). This indicates increased contact stress in the heart-lung interface,
perhaps related to increased pressures in the pulmonary vasculature following vol-
ume loading. The increased pulmonary vascular turgor might then lead to an in-
crease in the "deformation pressure" which is necessary to make the lungs fit the
shape of the heart . This phenomenon would be expected to be augmented by the ex-
368 O.A.Smiseth et al.

pansion of the left ventricle following volume loading. More studies need to be
undertaken to test this hypothesis.
Esophageal pressure was consistently less than PPL and is therefore an unreliable
estimate of the mechanical compression exerted on the heart by the lungs. It is im-
portant to be aware that in the present study PPL was measured only in the lung-
heart interface. PPL between the lung and chest wall has been shown to be lower
than PPL between the pericardium and the lungs [5].

Positive End-Expiratory Pressure and LV Contractility

The data presented above imply that the practice of using PPL or PES to assess
extraventricular constraint leads to incorrect estimates of transmural LVEDP. This,
in turn, may lead to false conclusions concerning the effect of PEEP and other inter-
ventions on LV diastolic compliance and on LV function curves. Figure 4 shows
stroke work plotted against both intracavitary LVEDP and transmural LVEDP

20.00
. 20.000
..
15.000

.
x

..
19.
•c
15.000

.• •
, c x

.. ..
X
10,000 C'J. 10.000 c
c
0 00 0 00
~ x
~ 5.000 C
5.000 c
E lIc
O+---~~__r - - - - - - - - -
~ C II
E 0
E
N E
E -5.000'+--,---.---r--.--,---, N -5,000 I I
E 0 5 10 15 20 25 30 E 52 53 54 55 56 57
LV Pressure (mmHg) E LV Diameter (em)
-"
-"
~ (;

~ 2~OOO ..
~

..
20.000
2 oX
iii 2
15.000- iii 15.000
.~'i. x
10.000- 10.000 • c
.... c
o 00
5.000 x ..
c 5.000 c
c x
O+-~~~------------
~ 0

-5.000+---1------r--.----. -5000
-10 0 10 20 30 • 2300 2400 2 00 2600 2700 2800
LV Transmural Pressure (mmHg) LV Area (mm2)
Fig. 4. Effect of PEEP on LV function in a representative experiment. A range of filling pressures
was obtained by volume loading. PEEP (20mmHg) shifted the stroke work-intracavitary LVEDP
relation downwards (upper left). However, the stroke work-transmural LVEDP relation (lower left)
was not affected by PEEP, indicating that there was no decrease in LV myocardial contractility with
PEEP. Similarly, the relations of stroke work against two measures of end-diastolic volume (upper
and lower right) were also unaffected by PEEP. (&), PEEP = OmmHg; (x), PEEP = SmmHg;
(0), PEEP = lOmmHg; (.), PEEP = lSmmHg; (0), PEEP = 20mmHg
Positive End-Expiratory Pressure and Cardiac Function 369

(LVEDP-PP). As expected, PEEP shifted the stroke work - intracavitary LVEDP

relation downward. However, there was no downward-shift of the stroke work -
transmural LVEDP relation. Therefore, contractility did not decrease. The present
data indicate that PEEP decreases LV performance according to the Frank-Starling
mechanism: increased extraventricular constraint raises left and right ventricular
intracavitary pressures impeding filling, and decreases transmural pressure (i.e., de-
creases preload); thus decreasing stroke volume. Another factor which seems to con-
tribute to the decrease in LV preload with PEEP is increased impedance to blood
flow through the lungs [9]. LV filling may also be impaired by diastolic leftward dis-
placement of the ventricular septum secondary to the increased right ventricular
afterload [7].

Pericardial Surface Pressure Assessed by Right Atrial Pressure

Since PEEP changes the LV diastolic pressure-volume relationship LVEDP or pul-

monary-capillary wedge pressure cannot be used to interpret changes in LV perfor-
mance according to the Frank-Starling mechanism. As shown in Fig. 4 such a prac-
tice may be misleading. We have shown in the normal and in the failing dog heart
that right atrial pressure is an excellent estimate of pericardial surface pressure [12,
14]. This is shown in Fig. 5. More recently we have confirmed these data in an intra-

50

40
-=:t:E
/

/
/
,
.
! /

.
/ /
w 30 / /
a:: /
/
/
/

::J /
en / /
en ;' .;' /
w /, /
a:: 20 ;' /. -/
A.

.,..
;,. ;' /
-J
<t /

. -.. ..
.
-/
a:: /.. /
I- /. /

<f 10
;'. /.
;, . . e . . ,.

.
I- / /
:t:
/:~: ~/
S!
a:: 0 /,..",-- ~'/
./
;'

/
/

-10 .
-10 o 10 20 30 40 50
PERICARDIAL PRESSURE (mmHg)
Fig. 5. Right atrial vs. pericardial mean diastolic pressure in dogs (number = 8). Recordings before
as well as after induction of acute ischemic LV failure are included. This suggests that right atrial
pressure may be used to estimate pericardial surface pressure. 95% confidence limits are indicated.
r=O.96; y=O.95x + 2.51; P<O.OO1 [from 12]
370 O.A.Smiseth et al.

30 30

"
20 20
t,"c 'Cl
'Cl :J:
E
:J:
E " 0
S
S 10 " D D 0

~
10
~
" 0
I

I
D
u

0,/8"
0
~
0 b oJ!)
"
0

-10 -10
70 75 80 85 90 15 100 70 75 80 85 to 95 100

AREAlved (%) AREA Ived (%)

Fig. 6. A representative dog experiment showing that myocardial diastolic properties are unchanged
by PEEP. Although the relation between pulmonary capillary-wedge pressure (ppcw) and end-
diastolic area is shifted progressively upward by PEEP, suggesting that diastolic compliance is de-
creased (left panel), the relation between estimated transmural pressure [Ppcw - PRA (right atrial
pressure)] and area is unchanged (right panel). Thus it should be possible to estimate true preload
(i.e., transmural pressure) clinically using a triple lumen catheter. (0), PEEP = OmmHg; (0) PEEP
= 10mmHg; (.6.), PEEP = 20mmHg

operative study in patients [15]. Figure 6 shows pulmonary capillary-wedge pressure

and pulmonary capillary-wedge pressure minus right atrial pressure vs. LV end-
diastolic area in a dog study in which we applied different levels of PEEP [4].
Whereas diastolic compliance appears to be changed by PEEP (left panel), the rela-
tion between estimated transmural pressure and area shows that myocardial diastolic
properties are unchanged.
On the basis of these observations we propose that LV transmural pressure (LV
preload) can be reliably estimated during PEEP in man by subtracting right atrial
pressure from LV intracavitary or pulmonary capillary-wedge pressures. The re-
lationship observed between right atrial and pericardial pressure implies that right
atrial transmural pressure is low. Of course, this might be different in patients with
right-sided heart disease (e.g., secondary to chronic lung disease) in whom there
might be significant right-sided hypertrophy. This will require further study. How-
ever, it may still be possible to estimate the change in PP from changes in right atrial
pressure even though absolute PP might be significantly underestimated by measure-
ment of right atrial pressure. The latter concept could be tested clinically by deter-
mining whether absolute diastolic pressure or estimated transmural-diastolic pres-
sure better predicts LV diastolic dimensions and systolic performance according to
the Frank-Starling relationship.

Conclusions

Depression of the LV function curve with positive end-expiratory pressure does not
indicate decreased contractility, but merely reflects increased extraventricular con-
Positive End-Expiratory Pressure and Cardiac Function 371

straint. Thus, decreased stroke work with positive end-expiratory pressure can be ac-
counted for by decreased LV preload.
Correct assessment of extraventricular constraint requires measurement of peri-
cardical surface pressure. An open-ended pericardial catheter measures liquid pres-
sure and seriously underestimates extraventricular constraint. Both pleural surface
pressure and esophageal pressure are inadequate measures of extraventricular con-
straint.
Finally, our data show that right atrial pressure is a good estimate of pericardial
surface pressure, an observation which may provide a means of calculating trans-
mural LV end-diastolic pressure or LV preload during positive end-expiratory pres-
sure.

References

1. Agostoni EE (1972) Mechanics of the pleural space. Physiol Rev 52: 57-121
2. Cassidy SS, Robertson CH Jr, Pierce AK, Johnson RL Jr (1978) Cardiovascular effects of posi-
tive end-expiratory pressure in dogs. J Appl Physiol44: 743-750
3. Ditchey RV, Costello D, Shabetai R (1983) Effects of airway pressure and lung volume on left
ventricular transmural pressure-volume relationships in humans. Am Heart J 106: 46-51
4. Douglas N, Kingma I, Smiseth OA, Smith ER, Tyberg JV (1984) Assessment of left ventricular
preload during PEEP from pulmonary capillary wedge pressure and right atrial pressure in dogs.
J Clin Invest Med [Suppl 3]7: 41
5. Fewell JE, Abendschein DR, Carlsson CJ, Rapaport E, Murray J (1980) Mechanism of decreased
right and left ventricular end-diastolic volumes during continuous positive pressure ventilation
in dogs. Circ Res 47: 467 -472
6. Grindlinger GA, Utsunomiya T, Vegas A, Levine LL, Shepro D, Hechtman HB (1982) Prosta-
glandin mediation of unstable hemodynamics during lung perfusion. Surgery 92: 52-60
7. Jardin F, Farcot J-C, Boisante L, Curien N, Mangairaz A, Boundarias J-P (1981) Influence of
positive end-expiratory pressure on left ventricular performance. N Engl J Med 304: 387-392
8. Kingma I, Smiseth OA, Frais MA, Smith ER, Tyberg JV (1987) Left ventricular external con-
straint: relationship between pericardial, pleural and esophageal pressures during positive end-
expiratory pressure and volume loading in dogs. Ann Biomed Eng 15: 331-346
9. Rankin JS, Olson CO, Arentzen CE, Tyson GS, Maier G, Smith PK, Hammon JW, Davis JW,
McHale PA, Anderson RW, Sabiston DC (1982) The effect of airway pressure on cardiac func-
tion in intact dogs and man. Circulation 66: 108-120
10. Scharf SM, Brown R (1982) Influence of the right ventricle on canine left ventricular function
with PEEP. J Appl Physiol 52: 254-259
11. Sears FW, Zemansky NW, Young HD (1980) University physics, reading. Addison-Wesley,
Reading
12. Smiseth OA, Refsum H, Tyberg JV (1984) Pericardial pressure assessed by right atrial pressure:
a basis for calculation of left ventricular transmural pressure. Am Heart J 108: 603-605
13. Smiseth OA, Frais MA, Kingma I, Smith ER, Tyberg JV (1985) Assessment of pericardial con-
straint in dogs. Circulation 71 : 158-164
14. Smiseth OA, Frais M, Kingma I, White A, Knudtson M, Cohen J, Manyari D, Smith ER,
Tyberg JV (1986) Correlation between pericardial and right atrial pressure during removal of
pericardial effusion. J Am Coli CardioI7:307-314
15. Tyberg JV, Taichman GC, Smith ER, Douglas NWS, Smiseth OA, Keon WJ (1986) The rela-
tion between pericardial pressure and right atrial pressure. An intraoperative study. Circulation
73:428-432
Part VIII
Cerebral Function and Cardiac Surgery
Chapter 32
Neuromonitoring in High Risk Surgery:
Physiological Tolerance Limits for Central Nervous System
LA.SULG

The term neuromonitoring refers to the semicontinuous recording, analysis, and dis-
play of some vital nervous system functions in high risk procedures in order to detect
spontaneous or induced changes in a patient's condition when special circumstances
(such as anesthesia, neuromuscular blocking, artificial ventilation, induced drug
coma, or primary coma of various origin) make neurological observation difficult.
Neuromonitoring delivers information about the integrity or possible disturbances of
certain nervous system functions [10, 76, 102]. Some alterations of monitored vari-
ables during surgical or other critical procedures from their preoperative characteris-
tics can be interpreted as a warning, indicating that there may be a deficiency in tis-
sue oxygenation [11, 14,30, 152]. These neurofunctional warning signs usually pre-
cede irreversible changes as a result of definite structural damage. When these signs
(such as EEG slowing or decrement in evoked response amplitudes) appear, there
may still be time to prevent permanent damage. The following functions of the cen-
tral nervous system can be monitored in quantified terms: EEG, evoked cerebral re-
sponses (EvCR), cerebral blood flow (CBF), cerebral oxygen consumption (CMR02),
and intracranial pressure (ICP).
This review surveys monitoring of electrical brain activity. In addition, simulta-
neous monitoring of other vital functions is also considered, especially if multivari-
able monitoring would offer a better understanding of the findings and trends in
monitoring procedure.

Neuromonitoring in Anesthesia

There is an imperative need to assess the depth of general anesthesia in order to

avoid pain and awareness of the surgical procedure as well as to minimize the risk as-
sociated with greater than necessary doses of anesthetic drugs. Direct quantitative
measurements of anesthetic agents during anesthesia have not proven helpful for
several reasons. Rapid determination of anesthetic levels is currently feasible only
for inhaled drugs. The equipment is expensive and the anesthetic concentration in
the system varies considerably between different anesthetic levels in the patient,
especially during induction and unexpected emergencies. Furthermore, a number of
factors other than drug concentration determine the intensity of anesthetic effects in
different patients (e.g., age, premedication) as well as in the same patient under dif-
ferent conditions (e.g., changes in body temperature, imbalance of electrolyte levels
in blood and tissues). The solution to this problem would be for the anesthesiologist
376 I. A.Suig

to develop clinical skills in assessing anesthetic depth from sign that are characteristic
to each class of anesthetic drugs. This task, however, is hard to achieve because it re-
quires personal experience from several years of anesthesiological practice. Objec-
tive measures of anesthetic depth based on computerized monitoring of some quan-
titative neurovariables appear to offer another, probably shorter, and perhaps also
more reliable way to solve this task. A device which would monitor a meaningful
data-reduced and compressed display of brain activity parallel to cardiovascular vari-
ables, blood gases, and electrolytes could be an optimal solution for reliable and con-
tinuous orientation in high risk surgery.

EEG

Since EEG was developed to a valuable clinical tool for studies on brain activity, evi-
dence has been accumulating about a close, but complex coupling between electrical
activity of the brain, CBF, and oxygen-glucose consumption of brain tissue. This evi-
dence was first based on the fact that general impairment of brain function (e.g., in
hypoxia, hypoglycemia, dysmetabolic coma) usually is accompanied by progressive
slowing of EEG. It can thus be hypothesized that the various wave patterns in an
EEG contain some information about brain metabolism. From here the question
arises: is the EEG pattern specifically connected to the individual, or does the de-
pendence stay at the general physiological level, where the plain neuronal metabolic
process designs the wave pattern and frequency content? The answer to this question
may be deduced from some essential EEG characteristics. At a certain functional
level EEG exhibits a stable, individually characteristic wave and frequency pattern,
also involving the well-known alpha rhythm. Before the human brain matures to its
adult form, EEGs pass through many different stages until the mature brain provides
an individual pattern in it, which we then retain for most of our life. This strict indi-
viduality has been convincingly illustrated in twin studies [129]. The two individuals
in a monozygotic twin pair are genetically identical. Consequently, their EEG pat-
terns are also so alike as to be nearly indistinguishable. EEG could therefore be con-
sidered almost as our inner physiognomy, or the "finger prints" of the brain. Its
appearance changes somewhat, as does our face when we mature and grow older,
but it always has our individual characteristics, as does our physiognomy. An indi-
vidually persistent pattern of EEG implies an intimate functional coupling to
physiological functions at different levels, although here the electrical wave pattern
is only a paraphenomenon of these higher functions, carried by energy-consuming
neuronal metabolism. But still, despite the substantial evidence about close relation-
ships between EEG, cerebral metabolism, and blood flow, there are some doubts re-
garding the nature of this dependence. Indeed, there are conditions which seem to
contradict the statement concerning a close relationship between EEG, cerebral
metabolism, and CBF: For example, CBF does not increase by eye opening as EEG
frequency does; the slow wave stage of sleep is not accompanied by a proportionally
reduced CMR02 ; in the immature brain EEG shows various slow activity patterns
despite high cerebral metabolism and CBF. These contradictions are, however,
mostly illusional. The brain is an extremely complex organ in its integrated func-
Neuromonitoring in Surgery 377

tions. The pattern of cognition, reasoning, thinking, and acting can pass rapid
changes without any significant shifts in brain metabolism, although there are consid-
erable changes in the EEG wave pattern. However, if the eyes-open EEG is moni-
tored through a progressing hypoxia, there is certain to be the same slowing as in the
eyes-closed EEG. Similarly, the slower immature EEG will become still slower when
metabolic processes are depressed, e.g., in encephalitis or in postictal state.
With the advent of quantitative techniques to measure blood circulation and
oxygen uptake [51, 52, 67, 74, 99], it became possible to correlate EEG and EvCRs
to these parameters quantitatively as well as regionally [97, 118, 146]. The most reli-
able method to measure regional CBF (rCBF) is by means of intracarotid injection
of isotopes [51, 74]; this has been supplemented by noninvasive techniques such as
inhalation and intravenous injection of isotopes [72, 93]. These alternative methods
can be chosen at the cost ofresolution and reliability. This deficit, however, is com-
pensated by the fact that CBF measurements, if necessary, can be repeated without
any harm to the patient.

Relationships Between EEG, Cerebral Oxygen Consumption and Blood Flow

For clinical purposes EEG is usually interpreted in descriptive terms. This is suffi-
cient for conventional purposes, but is less suitable for statistical evaluation. By
studying the quantitative relationships between EEG, CBF, and CMR02 , we can
learn how to utilize EEG more rationally in order to extract as much relevant infor-
mation as possible from these interrelationships. Could EEG then - analyzed and
interpreted by a computer - be a key to the code of cerebral metabolism? Having
again and again seen, for example in progressive ischemia or hypoxia, successive
EEG slowing and thereafter fading away - over burst-suppression pattern - into
electrocerebral silence, one cannot avoid being convinced that EEG really has some-
thing essential to tell us about the host organ [156]. The healthy brain in anesthesia
is, however, much more complex to understand than the dying one and therefore
there are still many problems to solve.
Obrist, the first to use statistical analyses of the relationships between EEG,
CBF, and CMR02 , showed that in natural aging in healthy individuals there is no
major change in EEG. In an age-matched group of psychiatric patients, however,
there was significant slowing of EEG, and decrease of CBF and CMR02 [92]. The
statistical significance was higher for the EEG/CMR02 relationship than for that of
EEG/CBF.
Ingvar et al. contributed with new experimental and clinical results on EEG and
rCBF, measured by means of isotope clearance analysis of intracarotid injected iso-
tope [51, 52]. One part of these studies, an elaborate statistical evaluation of clinical
material (144 neurological and psychiatrical patients) and 6 healthy controls, was
performed for an academic thesis [130, 131]. The statistical evaluation incorporated
regression and correlation methods and discriminative and profile clustering. De-
spite the heterogeneity of the material, statistically significant correlations between
EEG and CBF were found. In Fig. 1 the spectral distribution of quantifitated EEG
activity in five levels of mean rCBF in grey matter is shown. The groups differ by
378 LA. Sulg

EEG
PERCENT ACTIVITY TIME
in frequency classes 1-20 c/sec

40 cortical rCBF
in ml/100g/min

>100
35
99- 80
79-60
30 59-40
39 - 20

25 Shaded { 82.1!5.8
area (healthy controls)

20

15

10

2 4 6 8 10 12 14 16 18 20

EEG c/sec
Fig. I. Graphic display of quantitated EEG activity at different CBF levels (flow in cortical and sub-
cortical grey matter). Different EEG polygons represent group mean values of EEG activity dis-
tribution along the frequency scale (X axis) in five groups of neurological patients. Grouping is
based on CBF. Group mean values differ from each other by about 20 ml/l00g brain weight/min.
Shaded area, Distribution of EEG activity in a group of healthy controls with mean cortical CBF of
82.1 ml (S.D. = 5.8)/100 g brain weight/min. From Sulg [130]

20 ml flow per 100 g grey matter per min. For each group a mean quantitative EEG
(QEEG) profile was calculated and compared to the frequency distribution in a
group of healthy controls. It can be seen that for the subnormal flow classes there
was an increasing shift of EEG activity to slower frequences. However, in the class
with the highest flow a markedly abnormal pattern of QEEG was seen, which in some
respects was even more abnormal than in the group with the lowest CBF. This dis-
crepancy can be explained as a result of passive hyperperfusion in severely damaged
brain tissue with maximally dilated, nonreactive vessels: a so-called luxury perfusion,
an abnormal hyperperfusion, where the physiological coupling (the autoregulation)
between cerebral metabolism and CBF has been lost [46, 75]. Although CMR0 2 was
not measured in that material, it was apparent that the severity of tissue damage was
more adequately reflected in EEG than in CBF measurements [130].
Neuromonitoring in Surgery 379

Correlation Between Single EEG Variables and CBF

Since the EEG is the only brain activity variable which can be easily recorded and
quantified continuously it is well suited for brain monitoring in high risk conditions
or in situations where continuous documentation of brain activity is needed for some
other reason. Bickford et al. [9] initiated monitoring of compressed spectral power
profile arrays. When, however, the emphasis lies on an easily interpretable display,
comprehensible at a glance, then a semicontinuous graph of some few variables has
to be chosen. Therefore, there is an imperative need for meaningful data reduction.
To this end a multivariate stepwise regression analysis of several QEEG variables in
relation to CBF was performed [130, 135]. From single frequency classes the lO-
II Hz and 1-3 Hz bands achieved significant correlations, but the mean frequency
(mF) rose to a still higher significance. This close correlation between mF and CBF
was confirmed also in a later study on brain infarction [145]. Here the computerized
power density analysis of EEG and an intravenous isotope method for CBF mea-
surement were used [72]. In another study on 32 neuropsychiatric patients this EEG
parameter was compared with CMR02 , also here the correlation coefficient achieved
high significance [52].

EEG in Dysmetabolic Encephalopathies

Quantitative evaluation of EEG changes in metabolic brain disorders is needed

mainly for diagnostic purpose and for monitoring in coma. In liver cirrhosis a statisti-
cally significant positive correlation between galactose elimination capacity and mF
of EEG has been shown [63]. Successive trends in mF values usually move parallel
to clinical improvement/impairment or may precede it. Quite similar results were
obtained in patients with uremia. There was a shift towards lower frequencies in
EEG and a significant correlation between the degree of azotemia and mF [40].
These findings emphasize the fact that the degree of metabolic disturbance in brain
functions is reflected in EEG, but that EEG slowing is an unspecific sign common
both to hypoxia and various dysmetabolic states of the brain [130].

QEEG as a Measure of Cerebral Dysfunction Before and After Open Heart Surgery

The risk of central nervous system (CNS) disorders arising during open heart surgery
has been established in a number of clinical, EEG and CBF studies [4, 11,56, 98].
Despite advances in surgical techniques and extracorporeal circulation devices, CNS
disorders are still reported to occur quite frequently [1, 58, 107, 108, 123, 125, 149,
151]. Because of these risks attempts have been made to find some prognostic pre-
operative measures. A project was designed by the author to study effects of open
heart surgery on the brain [133, 134]. The purpose was (a) to find some prognostic
variables from QEEG preoperatively, (b) to establish better understanding of re-
lationship between conventional and QEEG interpretation, and (c) to compare the
clinical usefulness of various quantitative EEG analysis methods. Two preoperative
380 I.A.Sulg

and five postoperative EEG examinations were performed in 65 cardiac valve re-
placement patients in a 1-year follow-up study [119, 120]. Patients were also exam-
ined neurologically and neuropsychologically. EEG quantification was carried out
using three different analysis methods: (1) combined period and amplitude analysis
giving the following numeral outputs for frequencies 0.5-25.5 Hz: mean frequency,
mean voltage and mean energy; (2) normalized slope descriptor (NSD) time domain
analysis giving following parameters: activity (mean amplitude), mobility (mean fre-
quency) and complexity (measure of deviation from the sine-wave concept); and (3)
fast Fourier transform (FFf) analysis, giving power density spectra, mean fre-
quency, its standard deviation and skewness, kurtosis, 25 percentile value of quan-
titative polygon, as well as alpha mean frequency and activity areas for alpha, beta,
theta and delta in percentages of total activity [79, 122, 133, 137].
When the preoperative FFf mean frequency was low « 7 Hz) the incidence of
postoperative clinical complications was a little more than double that in cases with
higher frequency values. The presence of slow wave activity in the first postoperative
EEG, recorded 10 days after surgery, correlated clearly with the clinical findings:
CNS disorders were observed in 78% of cases with either episodic or continuous
delta activity. Of the QEEG changes the following proved to have the most signifi-
cant correspondence to a poor clinical outcome: (1) slowing of FFf mean frequency
with 2Hz or more, (2) slowing of alpha mean frequency with 1 Hz or more, and (3)
fall in the percentage alpha activity with 30% or more. The most essential conclusion
of this study was the fact that preoperative EEG seems to carry prognostic informa-
tion in cardiac surgery. Our observations indicate that EEG and especially QEEG can
provide warning of impaired tolerance to exceptional strains during extracorporeal
circulation and that these measures should not be overlooked in preoperative patient
evaluation [119, 120]. A long-term (5 years) follow-up study of that patient material
has been performed and is presented elsewhere in this volume (Chap. 36).

Pharmacoelectroencephalography

A prerequisite for a proper interpretation of monitored EEG is knowledge of the ef-

fect of anesthetics and other neurotropic drugs on brain activity [12, 13, 24, 33, 47,
54, 121]. Advances in neuro- and psychopharmacology have made available a large
number of novel CNS-active agents and have made possible their wide-ranging appli-
cations in the treatment of mental disorders as well as in anesthesia. Because of the
importance of recognizing drug effects, it is necessary to review some known effects
onEEG.

Psychotropic Agents [12, 24, 33, 54, 55]

Phenothiazines. All phenothiazines may alter the dominant frequencies of EEG. The
alpha rhythm may be slowed by 1 Hz or more, and there may be an increase in slow
activity. Some of these drugs, particularly chlorpromazine, may induce intermittent
high voltage, slow activity, and even seizures in some susceptible individuals [33,
54,55].
Neuromonitoring in Surgery 381

Meprobamate taken orally, when it affects EEG at all, produces a change somewhat
similar to that following small doses of barbiturate, namely the apparance of some
fast activity in the 20-30 Hz range [33, 54, 55].

Tri- and Tetracyclic Antidepressants. All these antidepressants can induce an in-
crease in slow EEG activity with episodic accentuation, especially after about 1
week's administration [33, 55, 106].

Imipramine elicits theta activity in patients with depression, whereas it may act as an
activator in patients with epilepsy [33, 54].

Lithium can induce diffuse, continuous slowing (usually frontal dominant) and even
focal slow wave abnormalities. These drug-induced EEG changes may be very pro-
nounced and may be mistaken for abnormalities due to an organic brain disorder [33,
54,57].

Anticonvulsants

Most anticonvulsants, although potent in reducing epileptogenic discharges, have

vague effects on background activity unless given in toxic doses. At toxic levels all
anticonvulsants cause diffuse EEG slowing, most pronounced by fenytoin intoxica-
tion [12, 33, 54, 110].

Narcotics

Morphine and all opium derivatives can induce slowing of EEG, usually first indi-
cated by a decrease in alpha frequency [12, 24, 33]. With chronic morphine adminis-
tration, alpha rhythm first slows, then returns to about predrug frequency [12, 24, 33].

Normeperidine, a meperidine metabolite, is a cerebral irritant. It is more likely to

cause seizures when given parenterally because meperidine is metabolized faster
than it is absorbed, and the ratio of normeperidine to its parent compound increases
[17].

Inhalational Anesthetics [17, 50, 85, 114, 127]

Nitrous Oxide. Up to a concentration of 30% N20 in oxygen, EEG changes are

minimal. As loss of consciousness develops, alpha rhythm progressively disappears
and short episodes of fast activity enter EEG. In curarized patients on 80% nitrous
oxide, the fast rhythm disappears and only 4-8 Hz activity remains. N20 associated
convulsions have been reported [22, 31, 114, 154].

Volatile anesthetics cause generally distinctive trends in the EEG pattern. The EEG
frequency falls and the amplitude increases as the concentration of anesthetic in the
382 I. A.Sulg

brain increases. For mean frequency and mean amplitude there is thus the same
reciprocal pattern as for cerebral hypoxia. Any accentuation of this pattern without
increased dosage of the drug indicates hypoxia.

Diethyl Ether. The first effect is fast 20-30Hz activity. Thereafter some 6-8Hz
waves appear in the background. With increasing alveolar concentration a low am-
plitude sinusoidal 15-30Hz activity usually appears [22, 29, 31]. Anesthesia with
ether should not be carried to the level of burst-suppression pattern (BSP) [114, 124].

Halothane (Fluothane) affects EEG in serial steps, passing through a low-voltage,

fast-frequency stage before the patient loses contact with his environment. At sub-
anesthetic levels, halothane produces fast 12-18Hz activity until consciousness is
lost. As anesthesia deepens, the frequency of halothane-induced rhythm decreases
as a near-linear function of anesthetic concentration. As anesthesia is further deep-
ened, 0.5 Hz activity becomes dominant until BSP appears, and by a minimum alve-
olar concentration (MAC) of about 3.5 the EEG has usually become isoelectric.
There is a continuous CBF and CMR02 decrement if anesthesia is further deepened,
and generalized seizure activity may appear suddenly, especially when also N20 is
added. Hypocapnia considerably accentuates EEG slowing. At the same time CBF
decreases progressively until systemic hypotension reduces perfusion pressure below
the autoregulatory threshold. This decrease in CBF continues beyond the point of
EEG isoelectricity. At an extreme high dose brain energy metabolism may become
irreversibly disturbed with ensuing lactic acidosis [22, 32, 35, 85, 114, 128, 147, 148].

Isoflurane is considered the inhalation anesthetic "of choice" in modern neuroanes-

thesia. Sub anesthetic concentrations produce diffuse 15-20Hz activity. With in-
creasing concentration a characteristic pattern develops with 4-8 Hz waves of high
amplitude dominating. As anesthesia deepens these waves further slow and BSP ap-
pears around MAC 1.5 followed by electrocerebral silence around MAC 2.0-2.5
[6, 50, 64, 114, 126].

Enflurane up to MAC 1 has EEG effects similar to those described for isoflurane,
but around MAC 1.5 BSP suddenly may appear with high voltage spikes within
bursts of EEG activity. In deep anesthesia EEG may show pronounced epileptiform
bursts separated by suppression sequences of 5-15 s. The EEG changes brought
about by hypercapnia differ from those typically seen with other agents: at a steady
anesthetic depth, hypercapnia usually produces the appearance on EEG of deepen-
ing anesthesia: the suppression periods lengthen and the bursts shorten. During en-
flurane anesthesia, however, hypercapnia may have the reverse effect: bursts
lengthen and suppressions shorten [6,17,60,77,90,126,128,147].

Methoxyflurane in low concentrations produces a regular 15-25 Hz beta-activity of

moderately large amplitude. With the onset of anesthesia there is some slowing, but
beta-activity still predominates. Small amplitude 1-4Hz activity develops later, but
high frequencies continue. BSP occurs only when this anesthetic is used together
with another agent such as nitrous oxide [6,17,150].
Neuromonitoring in Surgery 383

Intravenous Agents

Barbiturates. A slow injection of a short-acting barbiturate, e.g., thiopentone, pro-

duces a 15-30 Hz beta-rhythm. Next, 5-12 Hz waves superimpose on the fast activ-
ity, often in spindle-shaped bursts (barbiturate spindles). When these appear, con-
sciousness is usually lost. Large irregular I-3Hz waves then develop and indicate a
depth of anesthesia at which a skin incision will be tolerated. If the drug is given
rapidly, the earlier stages may be by-passed, and I-3Hz waves will dominate the
EEG. By increasing the dosage, BSP begins to break into the record, signaling a
stage at which minor surgery can be performed. For major surgery the duration of
suppression periods ought to be 3-4 s. In barbiturate anesthesia or coma the prog-
nosis with BSP or electrocerebral silence is not as poor as without these drugs, prob-
ably due to the brain-protective effects provided by barbiturates [12, 13,25,33,36,
48,54,66,84,111].

Benzodiazepines are particularly potent in provoking continuous and generalized

high frequency activity in EEG. Such beta-activity may persist as long as 2 weeks fol-
lowing a large dose [24, 25, 91, 96].

Ketamine. After an intermediate anesthetic dose of 2 mg/kg i. v. the EEG pattern

shows a mixture of 5-6 Hz and 30-40 Hz frequencies. Rapid i. v. administration of
20-30mg/kg usually results in generalized seizure activity. There can be an increased
incidence of convulsions in epileptics when ketamine has been given in normal
anesthetic dosage [7, 112, 114, 141, 144].

Fentanyl produces a slow frequency, high amplitude EEG when used in high dosage;
the reversal of these EEG changes by naloxone has been described by Kubicki [70,
114].

Butyrophenones. Of clinical importance as an adjuvant intravenous anesthetic is

droperidol (dehydrobenzperidol). A combination of a narcotic and a butyrophenone
is called neuroleptanalgesia, a state of a trance-like demeanor in which immobility
and analgesia exist. When high doses of these two drugs have been employed for
major surgical procedures, a prolonged postoperative CNS depression may result
and will be reflected also in extreme EEG slowing [88, 114].

Propanidid is an ultra-short-acting hypnotic that produces amnesia and unconscious-

ness as fast as thiopentone does. Propanidid has no analgesic effect. Although the
site of action of propanidid in the brain is unknown, EEG changes are similar to
those found with the barbiturates [114].

Steroid Anesthetics

Althesin, Minaxolone, Alphaxolone, Alphadolone. The EEG changes and the effects
on CMR02 are similar to those from rapidly injected barbiturates: oxygen consump-
tion and intracranial pressure are significantly decreased, even by induction dosages
384 I.A.Sulg

(50-100 jlg/kg). At the onset of unconsciousness, high voltage 1-3 Hz waves pre-
dominate in the EEG. Higher doses lead to BSP [18,19, 100, 114].

Other Anesthetics

Etomidate, a carboxylated imidazole, decreases CMR02 and intracranial pressure.

EEG changes are similar to those of thiopental. While myoclonies are sometimes
seen, they are not associated with epileptogenic EEG signs. Combinations of benzo-
diazepines or narcotics with etomidate suppress motor phenomena but prolong the
anesthesia and potentiate the EEG signs of depression [20, 53, 71].

Vasoactive Agents

Norepinephrine, epinephrine, histamine, isoproterenol and acetylcholine do not exert

any specific effect on EEG itself. These agents, however, can directly alter cerebro-
vascular resistance or may indirectly change the CBF by shifting the brain metabolic
rate. Many of these agents are monoamines which possess putative roles as neuro-
transmitters and in this way may alter brain metabolism. Changes in CMR0 2 and/or
in CBF will then be reflected in EEG, especially the deteriorating trends [114].

Nitroprusside is a potent vascular muscle relaxant and can therefore increase total
CBF, provided the arterial blood pressure is not significantly reduced [143]. Intra-
cranial pressure (ICP) increases in this situation [81]. Slowing of EEG and deteriora-
tion of neurologic status, caused by elevation in ICP, have occurred in association
with nitroprusside administration. Similar increases in CBF and ICP have been ob-
served in patients of this category with administration of hydralazine [94].

Muscle Relaxants

Muscle relaxants have no known direct effects upon cerebral activity or upon brain
vasculature. However, their indirect effects can be significant in neurosurgical pa-
tients. Succinylcholine, for example, may elicit an increase of ICP during anesthesia
[41]. Since succinylcholine does not cross the blood-brain barrier, factors such as CO 2
retention may be the indirect cause of increased cerebral blood volume [86, 87,142].

The Story of Brain Monitoring

Hans Berger (1873-1941) was the first to record EEG systematically in man; he
examined changes associated with chloroform anesthesia [8]. Since then the effects
of anesthetic agents on the electrical activity of nervous tissue have been extensively
studied. Critical reviews of the role of EEG in various monitoring situations, how-
ever, suggest that conventional EEG is a too expensive tool in terms of manpower
Neuromonitoring in Surgery 385

because it ties up highly trained personnel for long periods. Further, even the trained
eye has considerable limitations in grasping information from an ever-changing com-
plex wave pattern.

Computerized EEG Monitoring

By applying the fast Fourier transform algorith (FFf) on powerful computers it be-
came possible to perform digital analysis of multichannel EEG on line [21, 69, 79,
109, 133, 157]. This technique results in power density spectral profiles, where the
mixed frequencies from the original "analog" EEG are "sorted" into an orderly se-
quence from low to higher frequencies, and where the more abundant activity in cer-
tain frequency bands appears as peaks, readily seen visually. This display technique
has been further developed by plotting successive power profiles serially into com-
pressed spectral array (CSA) so that changes and trends can be seen and followed
over longer periods [9,16,89,113, 155].
This technique provides a quantitative "summary" of EEG activity. A major limi-
tation is that it is relatively insensitive to transient events. Spikes are not resolved un-
less they occur often and repetitively, as for example during a seizure. Also burst
activity-suppression pattern is suboptimally covered by CSA.
Stockard and coworkers used CSA to monitor EEG in 25 consecutive patients
undergoing cardiopulmonary bypass [123]. A close relationship was observed among
the degree of hypotension and neurologic outcome. Irreversible CSA changes (peak
power frequency < 3Hz or amplitude < 10 pV) occurred in three patients; two died,
and the other was left with permanent neurological impairment. Reversible CSA
changes occurred in another three patients, all of whom experienced transient post-
operative neurologic deficits. In each of these six cases, EEG abnormality appeared
coincident with hypotensive episodes. The severity of the episode correlated both
with degree of EEG abnormality and severity of neurological complication.
Quantitative EEG profile arrays can also be derived from combined period and
amplitude analysis, where every single wave is measured, using zero-line-crossing
techniques [15, 21,122,133]. This method has been used extensively and successfully
by Simons and Pronk in open heart and carotid surgery for more than 20 years [117].
In this system brain activity is analyzed in 15-s epochs; values from both hemispheres
together with blood pressure and temperature are displayed on a video screen and
can be printed as a hard copy (Fig. 2). Thus, not only the trends in the condition of
the patient may be followed closely, but the system also generates warning signals if
a critical condition is reached. Danger to the brain is thus detected and signalled in
time, so that measures can be taken to counteract complications [117].

Monitoring Single EEG Variables

In contrast to the quite sophisticated and expensive devices described above, there
has been a search for techniques for meaningful data reduction in long-term EEG
monitoring. Obrist and Sulg were the first (1970) to program a computer for
monitoring mean EEG frequency and integrated amplitude by means of period anal-
386 l.A.Sulg

EEC/AHAESTHESIE OOK~TATIE
ST·AHTOHIUS ZIEKEHHUIS
KLIHISCHE HElIROF'YSIOLOGIE
I'IEDISCH FVSISCH IHSTITUUT-THO
UTRECHT
PROCRAtt-VERSICIH: HS09t2

tIAI1E :
BIRTH [VVt1l'lDD):
SEX nOVF): 1'1
STAZU CODE : C2182tSl1
EEC CODE : 1-19848
DATE [VVI9lDD) : 8t8tUS
PRDI UIUIAN "' HRL.DOL
SURGERV RE-OP . AD-i<LEP
OPERATING ROOM: 01<9
SURCEON DR
AHAESTH. DR

._-----
HEUROPHVS . DR
__
...TECHNICIAN
8:24 START PROCRAI'I

8 : 29 H20 0104
8 28 VALIUM 28
3 : 29 F'ENTAHVL 10
8 : 29 HALOTHANE 0e;
8:55 F"ENTAHVL OFF
8 : 55 INFUSION
8 : 57 FENTANVL 28
921 HALOTHANE OFF
10 : 32 PERF'US I ON ON
10 35 COOLING
11 : '5 REWARMING
11 : 15 VALIUM 10
11 17 N20 OFF"
11 : 3? I-lARNING LOW ASVM EEG
11 : 39 N20 ON
11 : 42 PERF'US I ON OFF"
11'44 PERF'US 1 ON ON
11: 45 ARTEFACT OF"F
11' 45 PERF'US I ON OFF'
11 : 4? WARNING HIGH ASVM EEC
11: 48 END IJARNING HIGH ASVI'I EEG
11 : 53 IJARHIHG LOW ASvtt ££G
12: 0 ATT KEY LOU ASVI'I EEC
12:31 I-lARN I HG LOW ASVI'I ££G

1242 WARNING HIGH RSVM EEC

12 47 A~ARM HICH ASVM EEC
1247 ATT KEY ~~ ASVI'I EEC
12 : 47 END ALARM HICH ASYM EEG
lZ'49 WARNING HIGH ASVM EEC
1~ : 51 ALARM HIGH ASYM EEG
1252 ATT KEY HICH ASVI'! EEG
12: 54 END A~ARM HIGH ASVM EEG
12:59 WARNING HIGH ASVM EEG
1~59 WARNING ~OW ASVI'I EEG
13 . 1 ENO WARNING HIGH ASVM EEO
13: 10 VALIUM 10
13.11 END WARNING LOIJ ASVI'I EEC
13: 16 WARNING LOW ASVM EEC
t3 . 21 ATT KEY ~OW ASYM EEG
13.27 WARNING SLOW EEC
13:27 END WARNING SLOU EEC
13:28 STOP PROGRAM
CONCLUSIE:
ER HEBBEN ZICH ENICE CEREBRALE
PROBLEMEN VOQRCEDAAN, IJAARVAN DE
UITSLAC NOG NIET Z£KER IS · ICU-
CONTROLE IS WEHS£LVK . ELECTROOEH
ZO MOCELYK NIET VERIJYOEREN

Fig. 2. Intraoperative monitoring of quantitative EEG profiles from the two hemispheres. Print-out
every 15 s. The channel between hemisphere profiles shows mean of rectified EEG amplitudes. In
the next channels blood pressures, temperatures (surface and core), and administration of anes-
thetics are monitored. From about 10: 30 the effect of cooling and rewarming on EEG is seen; there
is slowing and amplitude decrement when cooling. About the same pattern is seen in hypoxia and
ischemia. From Simons and Pronk [117]
Neuromonitoring in Surgery 387

M 0 NIT 0 R I N G of Q E E G
from PARIETO-OCCIPITAL REGION in PROGRESSIVE HYPOXIA

MEAN
AMPLIT.
MEAN 11
FREQ. , 'I 4 5 6 7 A 9 H\ 11 *
- s-
10 20 30 40 50 60 70 80 90 100 110 V
-22.034 9.501 • v
36.964 9.970 •
-'..1.374
32.369
9.015 I.
9.92b. 11 v mean frequency : * :JI ....
32.">39
35.745 B.6b4.
9.0;09 III
~
X ...... ,* ::I
"". *.., :r
.. -
38.985 9.291. III
ct _~v._Y~'___V:
III

32.057 10.12;3.
mean amplitude .... PI
-.J~'-+--'9u'•..........,
'HI 2......... .... ' I-'
28.706 B• 856. ct V,"" ~' PI
30.525 8.979. ....
<
V
ct
-.
• .. c ...
, ....

;
27.899 9.557.
~.£2.2.6-+....:9z..u'7",'.7.£2...j..L'
33.211 9.569. III v........ '!
_38.467 9.41). •v "
35.939 8.889. 11
v'" v
~).842 8.1.21. III
:;*
38.44+1
410167
39.030
39.680
9.292.
II. 5'lo.
80341.
6.715.....
III
Q
ct
v,
_ _ _-+v!!..-_ _ _ _ _ _ _ _ _ ~_---
V
V,
\
45.3,2 8.062. 0 ',v *,

1
42.683 8.525. v> ,'*
49.222 7.999. ::I ;V ,*
45.504 70399. v........ ('
50.a81 7.363. ct v, ".
55.863 6.569. .... v *' Q
55.946 7.12). !:I V "', ... " III
;;J , "
57.21.6 6.315. III _ _ _ _ _ _ _--lVL~...,~l~~~--------- ::I

1
64.146 0.073. ,. V
70.028 '1.7H. ct .::::,.,*~.______ ct
IN\.298 b.08b. y

5Z.37fJ 6.520. 111 V::: ;.

62.54b 5.581. UI *;....... V
ct *... €v

1
63.128 S.654.
54.867 6.458. v:?" '*
_..§l-!..O~_~_I-"b'-!.• .:!.4~88_1_!_. "y.~
65.130 5.750. *~» it
~~.9~7~'1~~7~O~7~1~.L-----------------~v'~~---J-4-~-------------

)
59.755 b.524. v~'"
71.880
58.542
5.1.19.
6.073. ,'V,.-
• ... ---.V

--2L~~3~~5~,~9'~:81~5.L-------------------~.~-----------------
58.391 b.035. v* inhalation
__~l~~_~l~~i..n~'A~~~~~--------------------~~~v~~-
75.064
72.019
5.708.
5.056.
,* of 100 "
.'
--,V oxygen

-.
bO.669 b.]9",.

~
~v::-
53.967 70120. .v~ --*-
ItO.60b 9.267. V"

Fig.3. Monitoring EEG mean frequency and mean amplitude in experimental, voluntary hypoxia.
A healthy person inhales 6% oxygen. Mean frequency (*) and mean amplitude (V) are plotted every
10 s against corresponding scales at the top of the figure. Very soon these variables show reciprocal
trends: a decrement in frequency and an increment in amplitude. Reaction time was also measured
during this experiment. When the proband did not react adequately any more, 100% oxygen was
administered; within about 20 s EEG parameters returned to prehypoxic levels. This kind of recip-
rocal change in frequency and amplitude is a typical sign of slight, reversible brain hypoxia. From
Sulg and Ingvar [132]
388 1. A. Sulg

ysis technique. The value and reliability of this technique were demonstrated in ex-
perimental hypoxia (Fig. 3) [132].
Harmel et al. developed a simultaneous QEEG and QEMG monitoring tech-
nique. From two scalp electrodes three separate variables are derived: from EEG,
the average frequency and amplitude; from the frontal muscle, the integrated EMG
amplitude. On the basis of about 1000 surgical procedures they stated that frontal
EMG is a reliable guide to muscular relaxation in anesthesia and that this combina-
tion with EEG serves as a valid indicator in assessing the anesthetic state [43, 68].
Still more compressed data reduction was achieved by the cerebral function monitor
(CFM). This simple and portable device also derives its output from a single pair of
electrodes. Only frequencies of 2-15 Hz contribute to a single voltage trace. As am-
plitudes fluctuate and because of very slow paper speed (6cm/h) the trace appears
as a thick oscillating band on the chart [82, 101]. This system has been useful, espe-
cially in intensive care. The lack of higher frequency resolution, however, was a
severe handicap for CFM. This apparatus has later been developed further to a more
advanced form, which measures spectral power density in all conventional frequency
bands and displays the results in detailed, but perhaps too complex patterns [82,
103]. This new design, the cerebral function analysis monitor (CFAM), is a very
valuable tool in the hands of an experienced neurophysiologist [83, 148].

Monitoring EEG Mean Frequency and Amplitude

During Extracorporeal Circulation

The first part of this study was performed at the University Hospital of Oulu, Finland
[134, 136]. A Hewlett-Packard computer was programmed to perform quantitative
analysis (zero-crossing technique) in consecutive epochs from both hemispheres, and
to display and print out EEG variables (mean frequency, mean amplitude) every lOs
together with the measured perfusion pressure in systemic circulation. The results
were presented in numerical values as well as in semicontinuous plots along appro-
priate scales. Very soon some characteristic mean frequency (mF) and mean ampli-
tude (rnA) patterns ensued from hypotensive/hypoxic sequences of different sever-
ity. In slight and moderate hypoxia, often seen initially in extracorporeal circulation,
there was usually a reciprocal change: mF decreased and rnA increased. This recip-
rocal pattern (Fig. 4) is a sign of benign hypoxia without any risk for the patient.
When, however, hypoxia was more pronounced or appeared suddenly, rnA declined
together with mF toward the ominous isoelectricity. This EEG change is a warning
about the more dangerous stage of hypoxia. Here immediate measures are needed,
such as elevation of perfusion pressure. If this is not done in time, brain activity will
be suppressed and the deleterious events following this stage can no longer be moni-
tored by EEG.
The next example (Fig. 5) illustrates a catastrophe caused by human error: air
was accidently pumped into the left ventricle. EEG was the first physiological vari-
able which signaled danger: the amplitude dropped suddenly. Some 60-70 s later the
perfusion pressure turned to a steep paradoxical rise, a sign of severe, acute cerebral
ischemia. The surgical team reacted immediately, identified the cause and succeeded
 Neuromonitoring in Surgery 389

MEAN AMPLITUDE and MEAN FREQUENCY

of' the E E G
PERFUSION PRESSURE j mmH9
Hz
o
-----.---------------
50 100
I '~
EPOCH 15 . 1 1 3 . 0
'0 5 10
·a------.-------. ----------------
50

P?

P
70 . 9
71.9
70 . 2
.9 . 9
10"
No
:U
13. 1
13. 1
12 . 6 13 . 0
P 71 . I ~ 5 I I .~ 12 . I
,p 69 . 0 H 13 . 2 1 2 . 2
15 . 3 1 2 . 4

-p
P £7 . 2 57
r H .7 58 15 . 7 I I . 7
59 13 . 8 I I 7
P'I' ~U ,; 0 13 . 3 12 . 9
l 70 . 9 ,; I
6,
15 . " 13. t
17 . 4 11 . 8
g.~ 63 16 . 9 12 . S
,.........'1'7, .• 64 II.~ 12 . J
11 . 9

r<
.f' 71 . 2 65 I~ . J
",P 69 . 5 GG 23 .2 t o .9
G7 t 7 .4 t 2.0
;g . ~ ji8 t b 7 12 .0
f ~I .5 69 16 . ' I I .9
P G2 . I 70 14 . 0 t I .J
p - -P ;U
31 . 2
71
72
?J
12 . 3 I I
I ~ 1 10
13.. I I
.0
.•
.3
29 . 3 7~ 1 3 . I II .9
15 . 2 75 17 . 5 9.
n 5.

....
7.I H .2
9.1 (7 52 . 4 4
12 . 2 78 56 . 'l 3 .•
7 .7 79 ) 3.4 4 .4
10 ~O 26 . 0 4 .8

..
14 . 1 81 20 . 0 5 .2
14 82 1 2.6 • 5
15 . 0 83 14 . 5 1 5 . I
H 84 17 . I 4 . .;
1S . 9 65 II JI .. 5 5 5
24 ~ H 4 1 5 .8
30 . 1 87 9 .0 Ii . '}
30 3
J~
, ~8
89
10 . 9
9 .9
5 .7
~ . J
4 1 'i 311 I ~ . 3 4.I
42 . 5 91 9 . ~ 7.J
4 a. 4 92 8 .4 ~ ,9
41 . 1 93 l ij . J 5 .,
39 . 0 H 9.I
I
}- ;) , 0
7.G

>
/p 38 . 0 95
33 . 3 % 1 8I . J7
H .I H 10 2 ~ -~
30 . 01 ';8 9 .9 o. 1
CRITICAL VALUE!__ _ 7.7
27 . 9 ,9 9 . "l I 8 .
P 27 . 3 100 8 .0 I S .

Fig. 4. Computer output. Every lOs the following values were printed: mean perfusion pressure (P)
along pressure scale and corresponding numerical value ; critical level and area of P (i.e., cumulative
area beneath the 40mmHg level); consecutive number of analysis epoch; mean EEG amplitude (A)
and frequency (M) in numerical values and along corresponding scale. A typical EEG response to
slight hypoxia (due to decreased perfusion pressure) is seen

in eliminating most of the air from the heart and aorta. After some minutes EEG
amplitude slowly began to rise but stayed below the pre accident level for the rest of
the anesthesia. Postoperatively this resulted in coma with intermittent convulsions.
Fortunately these symptoms of severely disturbed brain function decreased succes-
sively, and 3 weeks postoperatively the patient was discharged without major neuro-
logic deficits. This was the only serious complication in these 20 cases of extracor-
poreal circulation in whom EEG was monitored intraoperatively. This case demon-
strated impressively that the brain signals danger without delay.

Anesthesia and Brain Monitor

On the basis of the above-mentioned studies on the relationship between EEG and
CBF/CMR0 2 a novel computerized device was developed - the Anesthesia and
390 I.A.Sulg

PERFUSION E E G E E G
PRESSURE mean meaD
in syst.em. AMPLIT. FREQ.
circulat.ion p.V Hz

t.ime
1 minute

air
emboli?

Fig. 5. In principle the same display as in Fig. 4. Here a marked response in EEG and perfusion pres-
sure to a massive air embolus is seen

Brain Monitor (ABM) - in cooperation with the departments of anesthesiology at

the University Hospitals of Oulu and Helsinki [134, 136]. This easily transportable
equipment records and calculates at regular lO-s intervals the mF and rnA from EEG
and the integrated voltage from spontaneous muscle activity, usually from frontal
muscles. Because carbon dioxide has a paramount role in the regulation of CBF,
oxyhemoglobin dissociation, hydrogen ion activity, and cardiac output, monitoring
of end-tidal CO2 were included in the ABM system (Fig. 6). It is well known that
alveolar pC0 2 , reflected in end-tidal CO 2 , follows pC0 2 changes in blood closely as
long as there is no pulmonary pathology interference [62, 65]. The CO 2 relation with
CBF is linear; if arterial pC0 2 is, e.g., doubled, then the CBF is, too. Hyperventila-
tion is a usual routine in neurosurgery and intensive care, but a reduction of 25%-
30% in CBF may become critical in patients with cerebral arteriosclerosis [139,140].
Neuromonitoring in Surgery 391

TIME

®
©

@
/~ /~
\ ,

Fig. 6. Display of anesthesia and brain monitor (ABM): A, Spontaneous EMG; B, EEG mean fre-
quency; C, EEG mean amplitude; D, neuromuscular responses; E, common display; F, alarm
regarding A; G, alarm regarding B; H, alarm regarding C; I, spontaneous EMG, latest value ; J,
EEG mean frequency, latest value; K, EEG mean amplitude, latest value; L, first muscle response;
M, TOF ratio (see Fig. 10); N, CO 2 alarm levels; 0, latest value; Q, Capnogram; X, overload message

In addition, effect of neuromuscular relaxants can also be monitored in terms of

muscle responses to serial automatically applied stimuli to a peripheral nerve [2, 27] .
Later on another version of this device was developed (ABM II), mainly for carotid
surgery and postoperative recovery as well as for intensive care [65, 136]. This ver-
sion is equipped with two channels for EEG so that both hemispheres can be moni-
tored. In order to be able to detect possible BSP, the analogue EEG can also be
visualized alternatively to the trend display. The EMG and CO 2 monitoring facilities
are the same as in the first version, but in addition two pressures can be monitored:
one on a higher level (e.g., arterial pressure) and the other for a low level, such as
venous blood pressure or intracranial pressure (Figs. 7 and 8). ABM I and II are al-
ready in use in several university hospitals [138].
Kay [65] evaluated both EEG and EMG monitoring functions of ABM and made
the following comments: During surgical anesthesia, increased intensity of narcosis
is indicated by a fall in mF and rnA that is easily measurable on ABM. In general,
EEG is useful in indicating an unwanted intensity of narcosis. Hypoxia has the same
effect as narcotics and ABM allows warning limits to be set, with an alarm to indicate
presence of hypoxia or unduly intense narcosis (Fig. 6). The frontalis EMG indicates
impending awareness during surgery on a paralysed patient. This muscle is quite re-
sistant to curarization and produces marked responses if the patient is aroused by
surgical stimulation under inadequate anesthesia, even in the presence of consider-
392 LA. Sulg

'""~~"'l:ivsltolic

Diastolic
Heart rate

Non-invasive BP1 graph

Systolic
mm •
US
'sioc?,I"'T Systolic
o:::5 Dias tolic
I1RIOO Heart rate
30 BPt
C02='t·B
Fig. 7. Display of ABM II for
Diastolic Heart rate blood pressure and intracranial
pressure recording. This moni-
Invasive BP2 graph
tor is compatible with most
pressure monitoring equipment
used in surgery and intensive
care. Measurements from these
lic
monitors can be transmitted to
ABM and displayed parallel
""1r U l.as,oolic to EEG and other monitored
variables
Diastolic

able neuromuscular block (Figs. 9 and 10). Use of intravenous anesthesia is regarded
more practicable with this monitor; ABM can ensure unconsciousness throughout
surgery, yet allow infusion to be discontinued before the end of surgery for early
recovery. Imminent recovery during surgery is readily apparent, and a bolus of
anesthetic will prevent awareness. Unwanted, too early awareness is the major prob-
lem in total intravenous anesthesia. Monitoring of neuromuscular blocking is in-
creasingly important in clinical practice, especially since the introduction of some
new relaxants (atracurium and vecuronium) recovery may be so rapid as to allow un-
expected and unwanted patient movement during surgery [2, 27, 65]. Rating and
Kuypers [104] monitored more than 200 anesthesias with ABM. They found that the
value of this monitor can be best demonstrated under conditions of near-accidents
and accidents during anesthesia. Examples are : accidental overdose of anesthetics
with CNS depression (low EEG frequency, high amplitudes, BSP) [104].
Edmond [26] evaluated ABM in 34 surgical patients and states that the ratio of
the lowest rnA in anesthesia to the lowest preoperative rnA (EEG amplitude ratio or
EAR) proved to be a valuable indicator of apparent adequacy of cerebral perfusion/
oxygenation. Despite a wide variety in patient ages (0.5-76 years), surgical proce-
dures, preanesthetic medications, and inhaled anesthetic agents, this ratio never fell
to less than 0.6 for> 1 min during anesthesia maintenance in 32 uneventful cases.
Correction of hypotension in one of the two remaining cases promptly elevated EAR
that had been continuously depressed for 6 min. In the other case, ABM indicated an
 Neuromonitoring in Surgery 393

DATEX Dual Channel Brain Act i vity Monitor

% C02 ( KP) __ ___ ICP(mmHg). ___ Ampl ___ EEG ___ Frq EMG BPA(mmHg )
o 5 10 0 20 50 25 uV 0 Hz 10 0 1 uV 10 30 100
Cl ock ------+-------+ +----+-----+ +--------------+ +--+--- -- + +------
.~ '\
1~ 1
e.
l
"

!....
;-
.... 1'10 ,"'- .~:
-;...
~ .

~-'-

lt
..:
55 - -~
.~

:£
~
-~ "
~
,if .!4~

(I, 0()
,~
1- t.
~~

56 ~
f l ~.

-~, t
?:
1-'.....
~~
r
;
J
j
1:00
~

.,,;-
I t
-. I .:'-"

J-:i
'

" ,;',.

"
57 - "
,{
~)
-;:
,<.
""
58 " ~ A
J
~
2,131'1 •
;.<C.
--;i
!
i };
:t
59 -
~ :-
- - ',
J rv_ 'r

,~ :t
;~ : (H) )~ ;;~
' -I~
.~
-,~: ;.! -
-';'"
,~;~~ .
.~.~ .
~; ~
~~
-~::"::>:
......:-;:.- /-

.'
:-::'
..", .
'.
,
._;-". ,:~
"
:)- -:.~

Fig, 8. Monitoring severe posttraumatic coma, Arterial blood pressure , heart rate, frontal muscle
EMG , EEG frequency and amplitude from right and left hemisphere , and intracranial pressure were
monitored, At about 00:30 high ICP was followed by a hypoxic sequence in EEG , At event mark 56
the patient was hyperventilated for 10 min
394 I. A.Sulg

C02 ____ •. HMT _._ • .•••.___ Aoopl.EEB_Frq/Hr EIIII ____ _____ _ BPC_g) _____ _
o+ - - - -, + - - -.10+ 0 ~ 100 100 0 20 0 ~ 100 30 100 200
+---+----+-.. +----_._ +---- -+ +-- ---+-----+ + •._- - .. ------- - .

3
4

Fig. 9. ABM II print-out of arterial blood pressure and heart rate (first channel) , from the right spon-
taneous EMG activity in frontal muscle (2nd channel), EEG (3rd channel), NMT (4th channel) and
capnogram (left channel) during surgery. It can be seen that the anesthesia is not deep enough to
eliminate the patient's reactions to painful stimuli in EMG and blood pressure , yet this reaction can-
not be seen in the NMT trace. Thus, although the neuromuscular block in the limbs is sufficient, the
frontal muscle still reacts to painful stimuli
Neuromonitoring in Surgery 395

MEASUREMENT OF DRUG EFFECT (DOSE· RESPONSE TO METHOHEXITONE)

X C02__________ NHT____________ ~pl __ ££O __ Fr~q/Hz £HO ___________ _

--
~------;-----!~ ~-----~~----!~~ Z----!~Z ~-----~Z 2----~----!~
Methohexitone 110 mg
00 N2070% in02
2 Alcuronlum 20 mg
...,.", """""'f'"U' ,..
.3
.It
~" mOlnltorlnR

001 12 r · Methohexitone SS mg. Alcuronium 5 mg

00136

(11100

It-
01112

. Atropine I "'I. mg .
Immediftl. on~t or !!pOnt.neou~ ventilation. rise In scaln EMO.
'rlIIIa;..- - - - l1li.-- - - - . - - __
(11136
CL - _ ...........

Fig. 10. ABM display. The effect of methohexitone on EEG and the alcuronium effect on NMT are
demonstrated. Note the reciprocal response of EEG parameters on every single dose of the anes-
thetic. From Kay [65]

extended period of unusually low EAR « 0.3) in the presence of moderate hypoten-
sion. This EEG pattern signified marked depression of cerebral activity, which was
manifested by a prolonged recovery from anesthesia. Thus, simply by comparing
mean EEG amplitude in conscious and anesthetized states, ABM can reliably assess
the adequacy of cerebral perfusion during anesthesia with common agents [26].
396 I. A.Sulg

Our experience with both ABM models in Norway indicates that the following
modifications would give the ABM system still better monitoring properties: (1) The
resolution of EEG scales has to be increased; they are now too compressed and
therefore the sensivity and resolution for both frequency and amplitude are too low.
(2) The time axis of the video display and print-out has to be more flexible; it is now
too rigid and works only with lO-s analysis epochs. With this time resolution the BSP
can hardly be visualized in a trend display or documented in print-out. Therefore a
5-s analysis epoch would be a useful alternative. When BSP is not the task for
monitoring, analysis epochs of 20s would usually be more appropriate.
In the further development of ABM system one also has to consider the increas-
ing need for monitoring evoked cerebral responses (EvCR), especially in abnormal
states or in drug-induced coma where the EEG has become silent [5, 102, 103].
EvCR are also indicated in situations where specific information is needed about in-
tactness and conductivity in somatosensory baths of spinal and cerebral structures [3,
38]. Here the SER (somatosensory evoked cerebral response), VER (visual evoked
response), and BAER (auditive evoked brainstem response) can offer valuable in-
formation to the surgeon. Several reports have been published about successful ap-
plications of EvCR technique in neuromonitoring [3, 38, 39, 57, 59].

Physiological Tolerance Limits for the Brain in Hypoxia and Other Critical Conditions

There are many factors, physiological as well as pathological, which may affect the
electrical activity of the brain. Low serum glucose concentration and hypoxia have
nearly analog effect. However, an abundance of these two nutrients has quite differ-
ent effects. Hyperoxemia may induce cortically generated convulsions, whereas
hyperglycemia (> 600mg/dl) may cause hyperosmolar coma with slowing of EEG
due to brain dehydration. Carbon dioxide, the most essential regulator of CBF, af-
fects the EEG directly as well as indirectly by changing CBF. Mild hypercapnia acti-
vates EEG through stimulation of the ascending reticular activating system; pro-
nounced hypercapnia induces narcosis and reduces EEG frequency. Hyponatremia
(Na < 120mEq/l) results in progressive EEG slowing. Hypernatremia (Na > 150
mEq/l) causes hyperosmolar coma with slow EEG.

Perfusion Pressure

The systemic arterial blood pressure and the intracranial pressure present, in their
combined effect, another tolerance limit for the brain [37, 41, 42, 45, 61, 141, 153,
158]. In normal man, CBF is autoregulated at 45-50 mllmin/100 g (brain-tissue
weight) within mean arterial pressures between 60 and 130mmHg [75]. Below this
range CBF becomes increasingly dependent on the cerebral perfusion pressure (CPP
= the difference between mean arterial and mean intracranial pressure) and on the
resistance of intervening blood vessels. As CPP falls a patchy pattern of ischemia
develops in the brain owing to regionally varying requirements for oxygen and glu-
cose [23]. The EEG begins to slow down (in man) when the regional CBF drops from
the normal (about 50 ml) to 20 mllmin/100 g. The central autoregulation of the CBF,
Neuromonitoring in Surgery 397

however, already becomes ineffective at mean systemic arterial pressures below 50

mmHg, as usually signalized by the slowing EEG [37, 45].
As long as the CPP is adequate, moderate fluctutations in intracranial pressure
(ICP) are not reflected in EEG. Falling systemic arterial pressure or increasing ICP
can lead to equalization of both, and thus to a global cerebral ischemia and to flatten-
ing of EEG. But an exceedingly high ICP can also leave the CBF and EEG unaf-
fected as long as CPP is maintained. It is thus important to note that CBF may be
adequate with a relatively small difference between systemic arterial pressure and
ICP, although the former may be greatly elevated [34, 73, 80].

Neuronal Death

The cerebral metabolic consumption rate for oxygen is 3.2-3.8 mllmin/lOO g, and for
glucose 60 mg/min/lOO g. Each molecule of glucose gives 38 molecules of ATP, which
is needed for neurotransmitter synthesis, transport mechanisms, and maintenance of
the membrane pumps [74, 115]. Selective vulnerability to hypoxia exists, not only be-
tween different cellular elements in the brain, neurons being most sensitive and
microglia least sensitive, but also between different neurons, with susceptibility de-
clining in following order: neocortex, basal ganglia, hippocampus, and cerebellum
[37]. Progressive ischemia is associated with increasing glycolysis and rise in cellular
lactate production, while high-energy phosphate levels are at first unaltered [78, 95].
Electrical activity progressively fails: EEG and somewhat later evoked cerebral re-

Table 1. Neuromonitoring and tolerance thresholds for electrical activity of the brain

Perfusion Oxygenation Brain electrical activity

CBF CPP Pa02 PV02 EEG EvCR
(mlll00 g!min) (mmHg) (mmHg) (mmHg) (HzandJlV) (JIV and ms)

Reference 50 70-80? 85-95 50-60 NormalEEG NormaISER,

values VER,BAER

Pathological 25-20 60-50 <40 28-25 Reciprocal pattern Amplitude

conditions in frequency! reduction
amplitude SER
<25 Declining VER
amplitude BAER
20-15 40-25 30-25 19-17 Burst-suppression Increasing
pattern latencies
Loss of auto-
regulation Electrical silence Distortion
at scalp level
<15 <25 <25 <17 Absent at
scalp level
Increased passive volume conduc-
tance in brain tissue

Irreversible brain tissue damage

398 I.A.Sulg

sponses are abolished at a flow of 1O-20mllmin/100g. When CBF falls to about

20 mllmin/100 g, intracellular migration of water from extracellular fluid leads to
swelling of astrocytes (cytotoxic edema) initially in grey matter. This may reduce re-
gional CBF by raising local tissue pressure or by distorting blood vessels, thus in-
creasing resistance to flow. When CBF falls to about 10 ml/min/lOO g there is sudden
membrane depolarization of the neurons with rapid efflux of potassum into the extra-
cellular fluid [5]. At even lower flows (6-9 ml/min/100 g) calcium influx begins [45].
When the abrupt failure of ATP production is induced by intracellular acidosis, the
result will be Na +IH+ exchange. If membrane failure allows calcium influx, this acti-
vates a self-propagating energy-wasting cycle, causing an accelerated fall in ATP
concentrations. Influx of calcium into mitochondria leads to failure of mitochondrial
function, the point at which neuronal death is probably inevitable (Table 1) [45, 116].

Conclusion

The purpose of neuromonitoring is to prevent invalidating and fatal complications

during surgery and anesthesia. This task is achieved by allowing the prognostically
most important organ, the brain, a possibility to signal alarm about immediate risks
to itself. Simultaneous monitoring of other vital variables offers additional informa-
tion needed for proper interpretation [57]. The need for neuromonitoring in high-
risk surgery and anesthesia is well expressed in an editorial note to a recent article in
the Journal of Neurology [39]:
Much of the neurosurgeon's work is done "blind" with regard to the production of a neurological
deficit even when no complications have occurred during the operation. It is as the patient awakens
in the recovery room that the surgeon must look for signs of cord or hemisphere damage, facial
nerve function, etc. It seems likely that in a few years time such uncertainty will be looked on as
primitive curiosity. With the recent advances in neurophysiological studies and evoked potential
analysis in particular, it will surely become possible to monitor every aspect of a patient's neuro-
logical function as the operation continues.

References

1. Aberg T, Ronquist G, Tyden H, Brunnkvist S, Hultman J, Bergstrom K, Lilja A (1984) Adverse

effects on the brain in cardiac operations as assessed by biochemical, psychometric, and radio-
logic methods. J Thorac Cardiovasc Surg 87: 99-105
2. Ali HH, Savarese 11 (1976) Monitoring of neuromuscular function. Anesthesiology 45: 216-249
3. Allen A, Starr A, Nudleman K (1981) Assessment of sensory function in the operating room
utilizing cerebral evoked potentials: a study of 56 surgically anesthetized patients. Clin Neuro-
surg 28: 457-481
4. Arfel G, Weiss J, DuBouchet N (1961) EEG findings during open-heart surgery with extra-
corporeal circulation. In: Gastaut H, Meyer JS (eds) Cerebral anoxia and the electroence-
phalogram. Thomas, Springfield, pp 231-249
5. Astrup J, Symon L, Branston NM, Lassen NA (1977) Cortical evoked potential and extra-
cellular K+ and H+ at critical levels of brain ischaemia. Stroke 8: 51-57
6. Bart AJ, Homi J, Linde HW (1971) Changes in the power spectra of electroencephalograms
during anesthesia with fluroxene, methoxyflurane and ethrane. Anesth Analg 50:53-63
Neuromonitoring in Surgery 399

7. Bennett DR, Madsen JA, Jordan WS, Wiser WC (1973) Ketamine anesthesia in brain damaged
epileptics: electroencephalographic and clinical observations. Neurology 23: 449-460
8. Berger H (1929) Uber das Elektroenzephalogramm des Menschen. Arch Psychiatr Nervenkr
87:527-570
9. Bickford RG, Brimm J, Berger L, Aung M (1973) Application of compressed spectral array in
clinical EEG. In: Kellaway, Petersen I (eds) Automation of clinical electroencephalography.
Raven, New York, pp 55-66
10. Binnie CD (1983) Telemetric EEG monitoring in epilepsy. In: Pedley TA, Meldrum BS (eds)
Recent advances in epilepsy, 1. Churchill Livingstone, Edinburgh, pp 155-178
11. Branthwaite MA (1975) Prevention of neurological damage during open-heart surgery. Thorax
30:258-261
12. Brazier MAB (1964) The effect of drugs on the EEG in man. Clin Pharmacol Ther 5: 102-116
13. Brazier MAB (1972) The neurological background for anesthesia. Thomas, Springfield
14. Brechner VL (1964) Current status of electroencephalography in the practice of clinical anes-
thesia. Clin Anesth 2: 87
15. Burch NR, Nettleton WJ, Sweeney J, Edwards RJ (1964) Period analysis of the electroen-
cephalogram on a general purpose digital computer. Ann NY Acad Sci 115: 827 -843
16. Chiappa KH, Burke SR, Young RR (1979) Results of electroencephalographic monitoring dur-
ing 367 carotid endarterectomies. Stroke 10: 381-388
17. Clark DL, Hosick EC, Rosner BS (1971) Neurophysiological effects of different anesthetics in
unconscious man. J Appl Physiol 31: 884-891
18. Clarke RSJ, Dundee JW, Carson IW (1972) Some aspects of the clinical pharmacology of
Althesin. Postgrad Med J [Suppl 2] 48: 62-65
19. Clarke RSJ (1974) The eugenols. In: Dundee JW, Wyant GM (eds) Intravenous anesthesia.
Churchill, Livingstone, Edinburgh, pp 162-192
20. Cold GE, Eskesen V, Eriksen H, Blatt Lyon B (1986) Changes in CMRO z, EEG and concen-
tration of etomidate in serum and brain tissue during craniotomy with continuous etomidate
supplemented with N20 and fentanyl. Acta Anaesthesiol Scand 30: 159-163
21. Cooper R, Osselton JW, Shaw JC (1980) EEG technology, 3rd edn. Butterworths, London
22. Courtin RF, Bickford RG, Faulconer A Jr (1950) The classification and significance of electro-
encephalographic patterns produced by nitrous-oxide and ether anesthesia during surgical
operations. Proc Staff Meet Mayo Clin 25: 197-208
23. Dearden NM (1985) Ischemic brain. Lancet II: 255-260
24. Dundee JW, Haslett WHK, Keilty SR (1970) Studies of drugs given before anesthesia: diazepam-
containing mixtures. Br J Anesth 42: 143-150
25. Dundee JW, Wyant GM (1974) Intravenous anesthesia. Churchill Livingstone, Edinburgh, pp
249-273
26. Edmond HL Jr, Yoon YK, Sjogren SJ, Maguire HT, McGraw CP (1983) In: Prakash 0, Mey
SH, Patterson RW (eds) Computing in anesthesia and intensive care. Nijhoff, The Hague, pp
279-292
27. Edmond HL Jr, Paloheimo M, Wanquier A (1988) Computerized EMG monitoring in anes-
thesia and intensive care. Malherbe, Schoutlaan, p 116
28. Edwards G, Morton HJV, Pask EA, Wylie WD (1956) Deaths associated with anesthesia.
A report on 1000 cases. Anaesthesia 11 : 194-220
29. Eger EI, Saidman LJ, Brandstater B (1965) Minimum alveolar anesthetic concentration: a stan-
dard of anesthetic potency. Anesthesiology 26: 756-763
30. Evans JM, Fraser A, Wise CC, Davies WL (1983) In: Prakash 0, Mey SH, Patterson RW (eds)
Computing in anesthesia and intensive care. Nijhoff, The Hague, pp 279-292
31. Faulconer A Jr (1952) Correlation of concentration of ether in arterial blood with electroen-
cephalographic patterns during ether-oxygen and nitrous oxide-oxygen of human surgical pa-
tients. Anesthesiology 13: 361-369
32. Findeiss JC, Kien JA, Huse KOW, Linde HW (1969) Power spectral density of the electroen-
cephalogram during halothane and cycloprane anesthesia in man. Anesth Analg 48: 1018-1023
33. Fink M (1964) A selected bibliography of electroencephalography in human psychopharmacol-
ogy 1951-1962. Electroenceph Clin Neurophysiol Suppl23
34. Forster FM, Nims LF (1947) EEG effects of acute increase of intracranial pressure. Arch
Neurol Psychiat 47: 449-453
400 I.A.Sulg

35. Galla SJ, Olmedo AK, Ketchmer HE (1962) Correlation of EEG patterns with arterial concen-
trations and clinical signs during halothane anesthesia. Anesthesiology 23: 147
36. Gisvold SE, Safar P, Hendrick HHL, Rao G, Moossy J, Alexander H (1984) Thiopental treat-
ment after global brain ischemia in pigtailed monkeys. Anesthesiology 60: 88-96
37. Graham DJ (1985) The pathology of brain ischemia and possibilities for therapeutic interven-
tion. Br J Anaesth 57:3-17
38. Grundy BL (1983) Intraoperative monitoring of sensory evoked potentials. Anesthesiology 58:
72-87
39. Hacke W (1985) Neuromonitoring. J Neuro1232: 125-133
40. Hagstam KE (1971) EEG frequency content related to chemical blood parameters in chronic
uremia. Scand J Urol Nephrol [SuppI7] (thesis)
41. Halldin M, Wahlin A (1959) Effects of succinylcholine on the intraspinal pressure. Acta Anaes-
thesiol Scand 3 : 155-161
42. Halley M, Reetsma K, Creech 0 (1958) Cerebral blood flow, metabolism and brain volume in
extracorporeal circulation. J Thorac Cardiovasc Surg 36: 506-518
43. Harmel MR, Klein FF, Davis DA (1978) The EEMG - a practical index of cortical activity and
muscular relaxation. Acta Anaesth Scand [Suppl] 70:97-102
44. Harris RJ, Symon L, Brantston NM, Mayhen M (1981) J Cereb Blood Flow Metabol1: 203-209
45. Hass HK (1981) Beyond cerebral blood flow, metabolism and ischemic thresholds. In: Meyer
JS, Lechner H, Reinch M, Ott ED, Arbinar A (eds) Proceedings of X Salzburg Conference of
Cerebral Vascular Disease, vol 3. Excerpta Medica, Amsterdam, pp 20-21
46. Henriksen L (1986) Brain luxury perfusion during cardiopulmonary bypass in humans. A study
of the cerebral blood flow response to changes on CO2 , O 2 , and blood pressure. J Cereb Blood
Flow MetaboI6:366-378
47. Herrmann WM (1981) Some examples for the possibilities and limitations ofpharmacoeiectro-
encephalography as a method in clinical pharmacology. Methods Find Exp Clin Pharmacol
[Suppl 1] 3: 55-76
48. Hicks RG, Kerr DR, Horton DA (1986) Thiopentone cerebral protection under EEG control
during carotid endarterectomy. Anaesth Intensive Care 14:22-28
49. Hjorth B (1970) EEG analysis based on time domain properties. Electroenceph Clin Neuro-
physiol 29: 306-310
50. Hosick EC, Clark DL, Adam N, Rosner BS (1971) Neurophysiological effects of different
anesthetics in conscious man. J Appl Physiol 31 : 892-898
51. Ingvar DH, Lassen NA (1965) Methods for cerebral blood flow measurements in man. Br J
Anaesth 37: 216-224
52. Ingvar DH, SjOlund B, Ardo A (1976) Correlation between dominant EEG-frequency, cere-
bral oxygen uptake and blood flow. Electroenceph Clin Neurophysiol41: 268-276
53. Ingram GS, Payne JP, Perry IR (1976) Electroencephalographic patterns during anesthetic in-
duction with etomidate. Br J Clin Pharmacol 3: 356-357
54. ltil1M, Guven F, Cora R, Hsu W, Polvan N, Ucok A, Sanseigne A, Ulett GA (1971) Quanti-
tative pharmacoeiectroencephalography using frequency analyzer and digital computer meth-
ods in early drug evaluations. In: Smith WL (ed) Drugs, development, and brain functions.
Thomas, Springfield, pp 145-166
55. ltil1M, Polvan N, Hsu W (1972) Clinical and EEG effects of GB-94, a "tetracyclic" antidepres-
sant (EEG model in discovery of a new psychotropic drug). Curr Ther Res 14: 395-413
56. Javid H, Tufo HM, Najafi H, Dye WS, Hunter JA, Julian OC (1969) Neurological abnormalities
following open-heart surgery. J Thorac Cardiovasc Surg 58: 502-509
57. John ER, Alter I, Ransehoff J (1982) Evaluation of coma patients with the brain state analyzer.
In: Grossmann RG, Greildenberg PL (eds) Head injury: basic and clinical aspects. Raven,
New York, pp 259
58. Johansson L, Lundberg S, SOderlund S (1967) Renal complications following heart surgery with
extracorporeal circulation. Scand J Cardiovasc Surg 1: 52-56
59. Jones SJ, Edgar MA, Ransford AD, Thomas NP (1983) A system for the electrophysiological
monitoring of the spinal cord during operations for scoliosis. J Bone Jt Surg 65B: 134-139
60. Julien RM, Kavan EM (1972) Electrographic studies of a new volatile anesthetic agent; Enflurane
(Ethrane). J Pharmacol Exp Ther 183: 393-403
61. Juneja I, Flynn RE, Berger RI (1972) The arterial, venous pressures and the electroencephalo-
gram during open heart surgery. Acta Neurol Scand 48: 163-168
Neuromonitoring in Surgery 401

62. Kalenda Z (1978) Capnography during anesthesia and intensive care. Acta Anaesthesiol Belg
29:201-228
63. Kardel T, Stigsby B (1975) Period-amplitude analysis of the electroencephalogram correlated
with liver function in patients with cirrhosis of the liver. Electroenceph Clin Neurophysiol38:
605-609
64. Kavan EM, Juliene RM (1972) Central nervous system effects of isoflurane (Forane). Can
Anesth Soc J 21 : 393-403
65. Kay B (1984) The anesthesia and brain monitor (ABM) concept and performance. Acta Anaes-
thesiol Belg [Suppl] 35
66. Keirsey DK, Bickford RG, Faulconer A Jr (1951) Electroencephalographic patterns produced
by thiopental sodium during surgical operations: description and classification. Br J Anesth 23 :
141-152
67. Kety SS, Schmidt CF (1948) Nitrous oxide method for quantitative determination of cerebral
blood flow in man. J Clin Invest 27: 476-483
68. Klein FF (1976) A wave form analyzer applied to the human EEG. IEEE Trans Biomed Eng
BME 23: 246-252
69. Kritikou PE, Branthwaite MA (1977) Significance of changes in cerebral electrical activity at
onset of cardiopulmonary bypass. Thorax 32: 534-538
70. Kubicki ST (1976) Fenthatienyl: EEG studies presented at VI World Congress of Anesthesia,
Mexico
71. Kubler J, Doenicke A, Laub M (1977) The EEG after etomidate. In: Doeniccke A (ed) Etomi-
date. Springer, Berlin Heidelberg New York , pp 31-48
72. Kuikka J, Ahonen A, Kouvula A, Kallanranta T, Laitinen J (1977) An intravenous isotope
method for measuring regional cerebral blood flow and volume. Phys Med Bioi 22 : 958-970
73. Langfitt TW, Weinstein JD, Kassell NF (1965) Cerebral blood flow with intracranial hyper-
tension. Neurology 15:761-773
74. Lassen NA (1959) Cerebral blood flow and oxygen consumption in man. Physiol Rev 39: 183-
238
75. Lassen NA (1966) The luxury-perfusion syndrome and its possible relation to acute metabolic
acidosis localized within the brain. Lancet II: 1113-1115
76. Levy WJ (1984) Intraoperative EEG patterns: implications for EEG monitoring. Anesthesiol-
ogy 60: 430-434
77. Levy WJ (1986) Power spectrum correlates of changes in consciousness during anesthetic induc-
tion with enflurane. Anesthesiology 64 : 688-693
78. Ljunggren B, Norberg K, Siesjo BK (1974) Influence oftissue acidosis upon restitution of brain
energy metabolism following total ischemia. Brain Res 77: 173-174
79. Lopes de Silva FH (1987) Computerized EEG analysis: a tutorial overview. In: Halliday AM,
Butler SR, Paul R (eds) A textbook of clinical neurophysiology. John Wiley & Sons, Chichester,
pp 61-102
80. Lundberg N (1969) Continuous recording and control of ventricular fluid pressure in neuro-
surgical practice. Acta Psychiat Neurol Scand [Suppl] 149: 1-193 (thesis)
81. Marsh ML, Shapiro HM, Smith RW (1979) Changes in neurologic status and intracranial pres-
sure associated with sodium nitroprusside administration. Anesthesiology 51: 336-338
82. Maynard DE (1979) Development of the CFM: the cerebral function analysing monitor (CFAM).
Ann Anesth Franc 20: 253-255
83. Maynard DE, Jenkinson JL (1984) The cerebral function analysing monitor. Initial clinical ex-
perience, application and further development. Anaesthesia 39: 678-690
84. Miller JD (1979) Barbiturates and raised intracranial pressure. Ann Neurol 6: 189-193
85. Miller RD (ed) (1986) Anesthesia. Churchill Livingstone, Edinburgh
86. Mori K (1973) Excitation and depression of C.N.S. electrical activities induced by general
anesthetics. In: Fujita M, Iwatsuki K, Miyazaki M (eds) Anesthesiology. International Con-
gress Series, no 292, Symp 1. Elsevier, New York
87. Mori K, Iwabuchi K, Fujita M (1973) The effects of depolarizing muscle relaxants on the elec-
troencephalogram and the circulation during halothane anesthesia in man. Br J Anesth 45 : 604
88. Morrison JD (1974) Neurolept techniques. In: Dundee JW, Wyant GM (eds) Intravenous anes-
thesia. Churchill Livingstone, Edinburgh, pp 207-218
89. Myers RR, Stockard JJ, Saidman U (1977) Monitoring of cerebral perfusion during anesthesia
by time-compressed Fourier analysis of the EEG. Stroke 8: 331-337
402 I.A.Sulg

90. Neigh JL, Garman JK, Harp JR (1971) The electroencephalographic patterns during anesthesia
with Ethrane. Anesthesiology 35: 482-487
91. Ngai AH, Tseng DTC, Wang SC (1966) Effect of diazepam and other central nervous system
depressants on spinal reflexes in cats: a study of site of action. J Pharmacol Exp Ther 153: 344
92. Obrist WD, SokoloffL, Lassen NA, Lane MH, ButlerRN, Feinberg I (1963) Relation ofEEG
to cerebral blood flow and metabolism in old age. Electroenceph Clin Neurophysiol15: 610-619
93. Obrist WD, Thompson HK, Wang HS, Wilkinson WE (1975) Regional cerebral blood flow
estimated by 133Xenon inhalation. Stroke 6: 245-256
94. Overgaard J, Skinh9lj E (1975) The effects of hydralazine upon intracranial pressure and cere-
bral blood flow. In: Harper M, Jennett WB, Miller JD (eds) Blood flow and metabolism in the
brain. Churchill Livingstone, Edinburgh, pp 16-17
95. Overgaard J, Mosdal C, Tweed WA (1981) Cerebral circulation after head injury. Part 3. Does
regional cerebral blood flow determine recovery of brain function after blunt head injury?
J Neurosurg 55: 63-74
96. Parsonage MJ, Norris JW (1967) Use of diazepam in treatment of severe convulsive status epi-
lepticus. Brit Med J 3: 85-88
97. Paulson OB, Sharbrough W (1974) Physiologic and pathophysiologic relationships between the
EEG and the regional cerebral blood flow. Acta Neurol Scand 50: 194-220
98. Pedley TA, Emerson RG (1985) Neurological complications of cardiac surgery. In: Mathews
WB, Glaser HG (eds) Recent advances in clinical neurology. Churchill Livingstone, Edin-
burgh, pp 159-178
99. Phelps ME, Hoffmann EJ, Huang SC (1978) A new computerized tomographic imaging system
for positronemitting radiopharmaceuticals. J Nucl Med 19: 635-647
100. Pickerodt VWA, McDowall DG, Coroneos NJ (1972) Effect of Althesin on cerebral perfusion,
cerebral metabolism and intracranial pressure in the anesthetized baboon. Br J Anesth 44: 751
101. Prior PF (1979) Monitoring cerebral function: long term recording of cerebral activity. Elsevier,
Amsterdam
102. Prior PF (1981) Electroencephalography in cerebral monitoring: coma, cerebral ischemia and
epilepsy. In: Statberg E, Young RR (eds) Butterworths international medical reviews, neurol-
ogy, clinical neurophysiology. Butterworths, London, pp 347-383
103. Prior PF, Maynard DE (1986) Monitoring cerebral function. Long-term monitoring of EEG
and evoked potentials. Elsevier, Amsterdam
104. Rating W, Kuypers R (1984) Frontal EMG and brain activity during and after anesthesia. Acta
Anaesthesiol Belg [Suppl] 35
105. Rossmann KA, Sate KF (1970) Recovery of neuronal function after prolonged cerebral isch-
emia. Science 168: 375
106. Saletu B (1982) Pharmaco-EEG profiles of typical and atypical antidepressants. Adv Biochem
Psychopharmacol32: 257-268
107. Salerno TA, Lince DP, White DN, Lynn RB, Charrette EJD (1978) Monitoring of electro-
encephalogram during open-heart surgery. A prospective analysis of 118 cases. J Thorac
Cardiovasc Surg 76: 97-100
108. Sawyer KC Jr, Sawyer RB, Robb WC (1963) Postoperative renal failure. Am J Surg 106: 668-
672
109. Schils GF, Sasse FJ, Rideout VC (1987) Automatic control of anesthesia using two feedback
variables. Ann Biomed Eng 15: 19-34
110. Schmidt D (1982) The influence of antiepileptic drugs on the electroencephalogram: a review
of controlled clinical studies. Electroenceph Clin Neurophysiol [Suppl] 36: 453-466
111. Schwartz J, Feldstein S, Fink M, Shapiro DM, Itil T (1971) Evidence for a characteristic EEG
frequency response to thiopental. Electroencephal Clin Neurophysiol3l: 149-153
112. Schwartz MS, Virden S, Scott DF (1974) Effects of ketamine on the electroencephalogram.
Anesthesia 29: 135
113. Schwilden H, Stoeckel H (1980) Investigations on several EEG-parameters as indicators of the
state of anaesthesia; the median - a quantitative measure of the depth of anaesthesia. Anaesth
Intensivther Notfallmed 15: 279-286 (in German)
114. Shapiro HM (1986) Anesthesia effects upon cerebral blood flow, cerebral metabolism, and the
electroencephalogram. In: Miller R (ed) Anesthesia, vol II. Churchill Livingstone, Edinburgh,
pp 795-824
Neuromonitoring in Surgery 403

115. Siesjo BK (1978) Brain energy metabolism. Wiley, Chichester

116. Siesjo BK (1981) Cell damage in the brain: a speculative synthesis. J Cereb Blood Flow Metabol
1:155-185
117. Simons AJR, Pronk RAF (1983) Automatic EEG monitoring during anesthesia. In: Prakash
0, Mey SH, Patterson RW (eds) Computing in anesthesia and intensive care. Nijhof, The
Hague, pp 227-257
118. Smith AL, Wollmann H (1972) Cerebral blood flow and metabolism: Effects of anesthetic
drugs and techniques. Anesthesiology 36: 378
119. Sotaniemi KA, Sulg lA, Hokkanen TE (1980) Quantitative EEG as a measure of cerebral dys-
function before and after open-heart surgery. Electroenceph Clin NeurophysioI50:81-95
120. Sotaniemi KA (1980) Cerebral disorders in open-heart surgery patients. A neurological, elec-
troencephalographical and neuropsychological follow-up study. Thesis. Acta Universit Oulu
(Finland) series D, no 56
121. Steen P, Michenfelder JD (1979) Neurotoxicity of anesthetics. Anesthesiology 50:437-453
122. Stigsby B, Obrist WD, Sulg IA (1973) Automatic data acquisition and period-amplitude analy-
sis of the EEG. Comput Programs Biomed 3 : 93-104
123. Stockard JJ, Bickford RB, Schauble JF (1973) Pressure-dependent cerebral ischemia during
cardiopulmonary bypass. Neurology 23: 521-529
124. Stockard JJ, Bickford RG, Smith TN, Saidman U, France CJ (1973) Structure-activity relation-
ships of ether anesthetics: an electroencephalographic study of diethyl ether, fluoxene, iso-
flurane and enflurane. Electroenceph Clin Neurophysiol 34: 713
125. Stockard J, Calanchini P, Pickford R, Billinger T (1974) Electroencephalographic seizures dur-
ing and after cardiopulmonary bypass. J Neurol Neurosurg Psychiat 37: 181-190
126. Stockard JJ, Calverly R, Myers RR, Smith TN (1975) Electrographic seizures induced by
fluroxene, diethyl ether and enflurane in man and cats. Electroenceph Clin Neurophysiol
38:97
127. Stoeckel H, Schwilden H (1987) Comparative pharmacodynamics of halogenated anesthetics:
quantitative EEG analysis to objectify central nervous effects. In: Peter K, Brown BR, Martin
E, Norlander 0 (eds) Inhalation anesthetics, new aspects. Springer, Berlin Heidelberg New
York, pp 26-32
128. Stullken EH, Milde JH, Michenfelder JD (1977) The nonlinear response of cerebral metabo-
lism to low concentrations of halothane, enflurane, isoflurane and thiopental. Anesthesiology
46:28-34
129. Sulg lA, Dencker SJ (1968) Electroencephalographic findings in MZ twin pairs, discordant for
closed head injury. Acta Genet Med Gemellol 17: 389-401
130. Sulg IA (1969) The quantitated EEG as a measure of brain dysfunction. Thesis, University of
Lund (Sweden)
131. Sulg lA, Lindgren G, Stigsby B (1969) Statistical evaluation of EEG data derived by period-
and amplitude analysis. A computer study. Comput Progr Biomed 1:20-31
132. Sulg lA, Ingvar DH (1970) Data reduction in EEG monitoring. Proc Annu Meet Aerospace,
Med Assoc, StLouis, pp 27-32
133. Sulg lA, Hokkanen TE, von Post B, Reunanen M (1977) Computerized quantitative multi-
parameter analysis of the EEG. In: Sicuteri F (ed) Headache: new vistas. Biomedical, Florence,
pp 213-224
134. Sulg lA, Hokkanen TE, Hollmen AI (1977) Simultaneous monitoring of perfusion pressure and
quantitative EEG in anesthesia and in other conditions with risk for hypoxia. In: den Hartog-
Jager WA, Bruyn GW, Heystee APJ (eds) 11th World Congress of Neurology. Excerpta
Medica, Amsterdam, p 152
135. Sulg lA, Sotaniemi KA, Tolonen U, Hokkanen E (1981) Dependence between cerebral metab-
olism and blood flow as reflected in the quantitative EEG. In: Mendlewicz J, van Praag HM
(eds) Adv Bioi Psychiat, vol 6. Karger, Basel, pp 102-108
136. Sulg lA, Hollmen AI, Tammisto T, Hokkanen ET (1982) Quantitative neuro-monitoring in
intensive care and in anesthesia. In: Advances in neurotraumatology. Internat Congress Series,
no 612. Excerpta Medica, Amsterdam, pp 235-237
137. Sulg IA (1984) Quantitative EEG as a measure of brain dysfunction. In: Pfurtscheller G, Jonk-
man EJ, Lobes da Silva FH (eds) Brain ischemia: quantitative EEG and imaging techniques.
Progr Brain Res 62: 65-84
404 I. A. Sulg: Neuromonitoring in Surgery

138. Sulg lA, Stenseth R, Hotvedt R (1987) Experiences with the Anesthesia & Brain Monitor
(ABM) in peroperative and in intensive care monitoring. Abstracts from heart & brain seminar.
University of TromS!ll, pp 78-79
139. Symon L (1970) Flow thresholds in brain ischemia and the effect of drugs. Br J Anaesth 57 :34
140. Symon L (1970) Regional cerebrovascular responses to acute ischemia in normocapnia and
hypercapnia. An experimental study in baboons. J Neurol Neurosurg Psychiat 33: 756-762
141. Takeshita H, Okuda V, Sari A (1972) The effects of ketamine on cerebral circulation and
metabolism in man. Anesthesiology 36: 69-75
142. Tarkkanen L, Laitinen L, Johansson G (1974) Effects of d-tubocurarine on intracranial pres-
sure and thalamic electrical impedance. Anesthesiology 40:247-251
143. Thomas WF, Cole PV, Etherington NJ, Prior PF, Stefansson SB (1985) Electrical activity of
the cerebral cortex during induced hypotension in man: a comparison of sodium nitroprusside
and trimetaphan. Br J Anaesth 57: 134-141
144. Thompson GE (1972) Ketamine-induced convulsions. Anesthesiology 37: 662-663
145. Tolonen U, Sulg IA (1981) Comparison of quantitative EEG parameters from four different
analysis techniques in evaluation of relationships between EEG and CBF in brain infarction.
J Electroenceph Clin Neurophysiol51: 177-185
146. Trojaborg W, Boysen G (1973) Relation between EEG, regional cerebral blood flow and inter-
nal carotis artery pressure during carotid endarterectomy. Electroenceph Clin Neurophysiol
34:61-69
147. Ty Smith N, Rampil IJ, Sasse FJ, HoffBH, Flemmings DC (1979) EEG during rapidly chang-
ing halothane or enflurane. Anesthesiology 51 : 35-54
148. Wark KJ, Sebel PS, Verghese C, Maynard DE, Evans SJ (1986) The effect of halothane on
cerebral electrical activity. An assessment using the cerebral function analysing monitoring
(CFAM). Anaesthesia 41 : 390-394
149. Witoszka M, Tamura H, Indeglia R, Hopkins RW, Simeone FA (1973) Electroencephalo-
graphic changes and cerebral complications in open heart surgery. J Thorac Cardiovasc Surg
66:855-864
150. Wolfson B, Siker ES, Ciccarelli HE, Gray GH Jr, Jones L (1967) The electroencephalogram as
a monitor of arterial blood levels of methoxyflurane. Anesthesiology 28: 1003-1009
151. Wollmann H, Stephen GW, Clement AJ, Danielson GK (1966) Cerebral blood flow in man
during extracorporeal circulation. J Thorac Cardiovasc Surg 52: 558-564
152. Wollmann H, Alexander SC, Cohen PJ, Smith TC, Chase PE, Vander Molen RA (1965) Cere-
bral circulation during general anesthesia and hyperventilation in man. Thiopental induction to
nitrous oxide and d-tubocurarine. Anesthesiology 26: 329
153. Wollmann H, Smith TC, Stephen GW, Colton ET, Gleaton HE, Alexander SC (1968) Effects
of extremes of respiratory and metabolic alkalosis on cerebral blood flow in man. J Appl
Physiol 24: 60
154. Yamamura T, Fukuda M, Takeya H, Goto Y, Furukawa K (1981) Fast oscillatory EEG activity
induced by analgesic concentration of nitrous oxide in man. Anesth Analg 60: 283-288
155. Young WL, Ornstein E (1985) Compressed spectral array during cardiac arrest and ressuscita-
tion. Anesthesiology 62: 535-538
156. Zaret BS (1985) Prognostic and neurophysiological implications of concurrent burst suppres-
sion and alpha pattern in the EEG of post-anoxic coma. Electroenceph Clin Neurophysiol 61 :
199-209
157. Zetterberg LH (1969) Estimation of parameters for a linear difference equation with applica-
tion to EEG analysis. Math Biosci 5: 227-275
158. Zwetnow NN (1968) Cerebral blood flow and autoregulation of blood pressure and intracranial
pressure variations. Scand J Clin Lab Invest [Suppl]102
Chapter 33

Monitoring Brain Function During Cardiovascular Surgery:

Hypoperfusion vs Microembolism as the Major Cause
of Neurological Damage During Cardiopulmonary Bypass
E.R.JoHN, L.S.PRICHEP, R.J.CHABOT, and W.O.1SOM

Introduction

As a result of improvements in the filters, oxygenators, and tubing materials used in

cardiopulmonary bypass (CPB) apparatus, the incidence of coma or paralysis due to
intraoperative brain damage has decreased but still remains significant. Most institu-
tions report an incidence of about 2% for these gross neurological complications [6,
12, 23, 24, 29, 49]. Using neuropsychological and psychometric tests to evaluate
more subtle functions, comparison of postoperative with preoperative performance
reveals impairment in a much higher proportion of patients. A number of studies
have concurred that significant behavioral deficits are demonstrable in about 3S%
of patients in the 1st week after CPB and persist in 7% -14 % for as long as 6 months
[1,22,41,42].
Numerous papers have established the fact that changes in the frequency spec-
trum of the electroencephalogram (EEG) reflect regional cerebral blood flow
(rCBF) [16,20,28,39,44-48, SO]. Some quantitative studies have suggested that the
frequency spectrum may display only minor changes until rCBF falls to a "critical
flow" level about 20% of normal (about 1S-20ml/100 g/min), below which marked
EEG slowing occurs [2, 7, 39, 43, 46, 48, SO]. Study of the effect of decreased mean
arterial pressure on the EEG suggests that in most patients the critical flow cannot
be maintained when mean arterial pressure (MAP) falls below SOmmHg [21].
There is widespread agreement that the EEG reflects the cerebral ischemia and
hypoxia which occur in many patients undergoing CPB [2-5,7,9,13-15,17,20,21,
26-28,31-33, 3S, 36, 44-47]. Persistent EEG slowing is followed by postoperative
neurological deficits, while such deficits are seldom noted if the EEG changes are
reversed by modifying the surgical procedure or if no intraoperative EEG changes
are observed [2, S, 8, 10, 2S, 30, 37-39, 41-43, 46].
An interesting recent paper compared 133Xe measures of rCBF and performance
on behavioral tests before and after CPB [40]. Both rCBF and behavioral perfor-
mance displayed moderate to severe postoperative deficits in 60% of the patients
tested, which persisted in 30% 6 months after surgery. Extremely high and consistent
correlations were noted between decreased rCBF and behavioral impairment.
There appears to be widespread consensus that the primary cause of intraopera-
tive brain damage during CPB is microembolism due to micro bubbles or showers of
debris released into the circulation by surgical manipulation of vessels containing
atherosclerotic plaques. We believe that another explanation may exist for the dif-
fuse EEG slowing and diminished rCBF which are correlated with postoperative
deficits.
406 E.R.John et al.

It seems reasonable to expect a large proportion of patients undergoing heart sur-

gery with CPB to have compromised cerebral vasculature. Not only might cerebral
atherosclerosis be more pronounced in some regions than others, but local fluctua-
tions of cerebral hemodynamics and autoregulation during the operation may vary
the relationship between blood flow in any particular brain region and the total cere-
bral blood flow or the perfusion pressure. A high proportion of the numerous obser-
vations of intraoperative cerebral ischemia cited above undoubtedly were made in
patients with MAP above 50mmHg and an average cerebral blood flow above 20ml/
100g/min.
Suppose that in many patients cerebral vasculature is compromised in some brain
regions more than others and that local hemodynamics and autoregulation may vary
between brain regions due to factors other than total perfusion volume, MAP, and
temperature. Even though CPB parameters remain constant, rCBF in different local
brain regions might occasionally fall below the volume critical for a given region,
with consequent changes in the EEG. Serial quantitative EEG (QEEG) observa-
tions might then be expected, in some patients, to show relatively abrupt onset of
QEEG signs of cerebral ischemia, which might be reversed by temporary increase of
perfusion, with reappearance in the same or a different region some time later. Fail-
ure to correct such local cerebral ischemia may produce the subtle postoperative
neurological and cognitive deficits with which we are now concerned.
The purpose of this paper is to present some evidence supporting these expecta-
tions.

Methods

SUbjects

Most of the examples presented in this paper come from observations typical of pa-
tients whom we monitored intraoperatively. These included 80 patients undergoing
CPB, 105 neuroembolizations of arteriovenous malformations, and 73 cases involv-
ing clipping of cerebral aneurysms. All surgery was performed at University Hospi-
tal, New York University Medical Center.

Electrophysiological Methods

Data Acquisition
An elastic cloth helmet containing 19 electrodes (Electrocap Corporation) was
placed on the patient's head. The electrode locations corresponded to the Interna-
tional 10/20 Electrode Placement System. Two additional electrodes were placed
medially above and laterally below the left eye to detect eye movements (EOG).
Two electrodes connected to the mastoids were linked and used as the reference for
mono polar recording. A standard EKG electrode on the right shoulder served as
ground. Contact between the electrodes in the helmet and the scalp was achieved by
Monitoring Brain Function During Cardiovascular Surgery 407

injecting an electrolyte gel through the center of each disk-shaped electrode. EOG,
reference, and ground electrodes were attached using EKG electrodes or EEG elec-
trode paste. All electrode impedances were kept below 50000 and were checked
regularly throughout surgical procedures.
Data were gathered using a Brain State Analyzer (BSA) (Cadwell Laboratories,
Kennewick, WA). Nineteen monopolar and one EOG channel were recorded simul-
taneously under control of the microprocessor in the BSA. An on-line automatic
artifact detection algorithm was used. Artifact thresholds were defined separately for
each of the 20 channels. Data were gathered in 2.5-s segments and displayed on the
screen of the BSA video monitor as a 20-channel white tracing. Activity on any chan-
nel which exceeded its artifact threshold was displayed in red. Each segment of data
was also subject to conventional examination by the BSA operator who could choose
to override the automatic artifacting algorithm for any segment.

Data Analysis
A set of artifact-free EEG samples will be referred to as a "session." Sessions usually
contained 60 s of data. Where data were especially stationary, as at low temperatures,
sessions might contain as little as 15 s of data. Data from each session were analyzed
as soon as acquired. Analysis consisted of subjecting the data to spectral analysis
using the fast Fourier transform [34]. A number of quantitative features were then
extracted, including the absolute (uv2 ) and relative (%) power in the L1 (1.5-3.5 Hz),
() (3.5-7.5 Hz), a (7.5-12.5 Hz), and f3 (12.5-25 Hz) frequency bands [18, 19].

Z-transformation
These quantitative descriptors were subjected to various logarithmic transformations
which have been previously demonstrated to achieve Gaussian distributions for large
samples of these features extracted from normally functioning healthy persons. Gaus-
sian distributions are a prerequisite if QEEG data are to be evaluated using para-
metric statistics [11, 18, 19].
Age-regression equations describing the mean value and standard deviation of dis-
tributions of these QEEG descriptors obtained from a large healthy population aged
6-90 years had previously been constructed for 8 bipolar (F7TiFgT4, T3TS/T4T6,
C3~/C4~' and P301/P402 in the 10/20 system) and 19 monopolar derivations and
were stored in the BSA [19]. Each individual feature derived from a patient was Z-
transformed relative to the appropriate normative equations. Figure 1 describes the
Z-transformation. A Z-score was calculated by obtaining the difference between the
observed feature value and the age-appropriate normal value predicted by the equa-
tion and dividing that difference by the corresponding standard deviation. The Z-
score can be directly translated to the probability that the patient's value falls within
the normal distribution, with a Z-value of 1.96 corresponding to P<0.05 [34]. We
refer to this as "neurometric QEEG" analysis.
Similarly, successive QEEG samples from the same individual can be compared
to an appropriate initial sample ("self-norm"). Intraoperatively, in patients with
preoperative evidence of compromised cerebral vasculature, it is particularly advan-
408 E. R. John et al.

A-I -I 0
S.D.
1 1

i! - SUBJECT MEAN VALUE

VALUE OF SAMPLE
STANDARD DEVIATION
OF SAMPLE
Fig.t. Z-score transformation derived by calculating
the difference between an observed QEEG feature
i! ,.. PROBABILITY THAT SUBJECT VALUE value and the age-appropriate normal value and divid-
LIES WITHIN 'NORMAL! RANGE ing by the corresponding standard deviations

tageous to Z-transform data relative to the anesthetized, pre-bypass self-norm. Some-

times, inspection of relative power maps without Z-transformation suffices to reveal
cerebral ischemia, but we prefer to scale our maps relative to some known baseline
reference. Neurometric scaling is an invaluable aid in interpreting the clinical signifi-
cance of QEEG observations.

Topographical Mapping
For any selected feature (e.g., absolute or relative power in a particular frequency
band, Z-transformed value, or significance of change of any feature between two ses-
sions), a topographical map could be computed which displayed the distribution of
values of that feature across the head. To construct such maps, the values of the
feature to be mapped were entered in the position on a head diagram which cor-
responded to the anatomical location of the electrodes from which each value had
been extracted, with interpolation of values at intermediate positions. The map con-
tained 40000 pixels. The value of each pixel was computed as the sum of the feature
values at the four electrodes (EI-E4) nearest to the pixel, attenuated by the inverse
square of the distance from the pixel to each electrode, i.e.,

Value (El) + Value (E2) + Value (E3) + Value (E4)

Value (pixel) = d2 (El) d 2 (E2) d 2 (E3) d2 (E4)

This interpolation algorithm was selected as a reasonable approximation to the

values of the field at any point inside the quadrilateral bounded by any four elec-
trodes. A color code convention for actual values was adopted using a "heat scale"
so that the higher the value at any point the hotter it would appear. The colors corre-
sponding to the heat scale ranged from brown through shades of red, orange, and
yellow to white. To show statistical significance in neurometric maps, the color code
ranged from red for values significantly greater than the normative data to blue for
values less than the norm. The appropriate scales were displayed in a color key on
the screen next to each head map.
Monitoring Brain Function During Cardiovascular Surgery 409

Results
Relationship Between Relative (%) L1 Power and rCBF

Many previous workers, using qualitative evaluations, have concluded that there is a
relationship between the EEG and rCBF, such that slow wave shifts reflect cerebral
ischemia. Using QEEG, we have obtained evidence which demonstrates that not
only does a slow wave shift occur with cerebral ischemia, but that decreased cerebral
blood flow is particularly well reflected by an increase in the relative power in the L1
band.

Topographical Map of Relative (j Power in Confirmed Stroke

Figure 2 (see page 410) presents a topographical QEEG map recorded from a patient
in the recovery room following a coronary artery bypass operation. In all maps pre-
sented in this paper, the head is viewed from above with face upward. This patient
had not been monitored intraoperatively and was found to be left hemiplegic after
recovery from anesthesia. A CT scan, taken after the EEG recording, revealed oc-
clusion of the right middle cerebral artery.
The map shows the percentage of the total power in each region which was in the
.1 frequency band. Relative .1 power appears to be particularly high in the anterior
half of the right hemisphere. This is reflected by the bright yellow quadrant which
corresponds to the perfusion field of the right middle cerebral artery.

Differential .1 and () Shifts After Interruption of Blood Flow

In one particularly informative case, we were monitoring a patient at 27°C shortly
after going on bypass, when circumstances unexpectedly required very low blood
flow for several minutes.
The left side of Fig. 3 shows the Z-values for relative .1 and () power in the parieto-
occipital and central regions over the right hemisphere before, during, and after the
brief episode of minimal blood flow. Before this episode, significantly increased Z-
values for e, together with Z-values within the normal range for .1, had been ob-
served in both the parieto-occipital and central regions. This pattern is normally ob-
served during CPB and characterizes many different anesthetic regimens. Upon
severe diminution of blood flow, the EEG spectrum changed drastically. Because of
the great potential utility of the information available from this rare circumstance,
successive 15-s EEG samples (six 2.5-s segments) were analyzed separately.
Within 15 s after blood flow was decreased to 500 ml/min, a very marked increase
in relative .1 power occurred while relative () power decreased dramatically. The Z-
value of this .1 shift was much more significant for the parieto-occipital than for the
central regions, suggesting a relatively greater sensitivity to cerebral ischemia in
anatomical regions of the "watershed" or "boundary perfusion" area. Within 15 s
after flow was resumed to 5500ml, after having remained at 500ml for 5 min and
15 s, .1 diminished and () increased so that the relative power in each band in both
brain regions returned to approximately the same level seen before this episode.
The right side of this figure shows the Z-values for absolute .1 and () power in the
parieto-occipital and central regions observed during the same episode. Notice that
410 E . R . John et al.

.
Fig. 2 Fig.S

-
-~-,.- - I!!ig -

-. -
~ ,,.-
~

.-.-,-.
. ,---.
.. I

....- -~-'-.---
'.~--" ~ ~''-- .= -;:-~-.
- -- - -
~ ~

! - ",--.. __
- -~~ --=
~ - ~~-.;-- ~. !!! - ~ --li~-
- - . - ... - -
.. - ~

$ --1-: -~-~ =~- =l -

:l ."." ~= - 21 ' ' '' 11'

Fig. 7

Fig. 2. Topographic QEEG map of relative power in the delta frequency range obtained from a pa-
tient who was left hemiplegic due to right middle cerebral artery occlusion confirmed by CT scan

Fig. S. Change in relative delta power seen 15 s after the inflation of a balloon catheter in the left
carotid artery during a neuroembolization procedure

Fig. 7. Changes in Z-transformed delta, theta, alpha, and beta relative to the anesthetized self-norm
during a CPB procedure. Changes seen correspond to the effects of an 8% reduction in blood flow
at a constant mean arterial pressure of 65 mmHg

Fig. SA-F. Topographic maps of relative delta power in a patient with CT-scan-documented infarcts
in the left posterior temporal and parieto-occipital regions. These maps were obtained pre-aortic
valve replacement (A) , 23-min post bypass, BP = 75 mmHg (B) , after BP was raised to 170mmHg
(C), after a drop in BP to 35 mm Hg (D), after vasopressor injection increased BP to 80 mm Hg (E),
and 14 days postoperatively (F)

Fig. 9A, B. Changes in delta activity comparing preoperative and postoperative values in a typical
monitored (A) and a typical nonmonitored (B) CPB patient. Note the Z scale for A is ± 3.5 and for
B ±7.0
Monitoring Brain Function During Cardiovascular Surgery 411

A,B

C,D

E,F

A B
Fig. 9
412 E. R. John et al.

RELATIVE POWER ABSOLUTE POWER

6
5 §-6

r-
P402
4
3
2 . b-<><>..o..i
0-

w -1
c::
0 -2
u
CJl

N
3
2
Fig. 3. The left side shows Z-values for
relative delta and theta over the right
pariet%ccipital and central regions
-1 during ePB surgery before, during,
-2 and after a brief episode of minimal
blood flow. The right side shows Z-
values for absolute delta and theta
power during the same periods of time

the effects of the ischemic episode are much less obvious in absolute than in relative
power measures. This is due to the greater variance of absolute power measures
under these circumstances, which lessens the statistical significance of observed shifts.
These data confirm the observation of a slow wave shift in the EEG correlated
with cerebral ischemia reported by other workers and indicate that relative L1 power
may be an EEG feature which is more sensitive to regional ischemia or hypoxia than
the absolute power measures which are commonly used. Since no marked postopera-
tive neurological deficits were noted, these data also suggest that, under hypothermic
conditions, brief periods of severe cerebral ischemia can be tolerated.

Pump Function
The most casual experience with electrophysiological monitoring during CPB sur-
gery reveals that dramatic changes occur in both the EEG and the EP whenever
changes in flow rate or blood temperature take place. How is one to distinguish be-
tween statistically significant and clinically significant changes? It might be possible
to predict changes from biophysical considerations and to identify clinically signifi-
cant changes as those greater than expected with particular CPB parameters. The
expected percent change in EEG or evoked potential (EP) measures can be calcu-
lated with the following equation:

0.2 (VN~VP) + 0.5 (TNl~Tp)

Monitoring Brain Function During Cardiovascular Surgery 413

Where VN is normal flow, Vp present flow, TN normal blood temperature, and Tp

present blood temperature. This pump function is based upon our initial set of obser-
vations in CPB. Brain electrical activity is sustained by glucose metabolism. The
pump function predicts that brain metabolic activity will decrease proportional to the
percentage of normal blood flow which is available (first term in the function) and to
the well-known QlO rule which describes the metabolic consequences of temperature
change (second term in the function). This function corresponds well to some of our
observations. The factors of proportionality may differ in various brain regions and
among different features of brain electrical activity. Accordingly, we decided that it
would be worthwhile to try to construct pump functions using regression methods,
also considering effects of electrode position, mean arterial pressure, and anesthetic.

Brainstem Auditory Evoked Response Changes and the Pump Function

Systematic changes in the latency of evoked potentials can be observed during CPB.
The brainstem auditory evoked response (BAER) has five positive peaks. In the top
of Fig. 4, the absolute latency values of peaks P1-P5 , N4 and N6 are plotted as they
were observed during a CPB procedure. They appear as a set of nested curves.

PEAK LATENCY (msec) .--__ Pi

16 P2
<>---<J
.......... P3
3l 14
l>----6 P4
.......... P5
'>""--"1 N4
.5. 12 -'---9 Ne
>-
~ 10
lLJ
~ 8
-'
::.:::
<[
6
lLJ
~ 4

11'20 '30 '40 ,50 12,00 '10 '20 '30 '40 '50 13'00
RELATIVE CHANGE IN PEAK LATENCY (%)
100

0:: 80 +-+ BLOOD TEM~ CORRECTION

lLJ
3:
<>--<> STATE COEFFICIENT
gOO
en
.,e4O

20
- O.......ItCiI!;;1II:."/
0::
~-20
~
.,e
11 ,20 '30 '40 '50 12'00 ,10 ,20 ,30 ,40 ,50 13'00
TIME (hr'min)
Fig.4. Relationships between changes in BAER absolute peak latency values, percent change in
BAER initial peak latency values, and the value of the pump function computed as a function of
CPB parameter changes
414 E.R.John et al.

In the bottom of the figure, the five curves have been replotted after rescaling
from absolute latency to percent change from initial latency. The five curves now be-
come superimposed. The heavy black curve represents the value of the pump func-
tion computed as CPB parameters changed during the operation. The change in
peak latency lags the pump function, as should be expected, but corresponds well to
the predicted value.

Anatomical Distribution of Changes in L1 Activity

During Neuroembolization of Cerebral Vessels
Neuroembolization of cerebral vessels offers a unique opportunity to observe the
QEEG consequences of reversible local occlusion of a cerebral vessel by a balloon
catheter. Typically, a slow wave shift in the EEG is observed within a few seconds
after inflation of the balloon whose position is angiographically confirmed.
Figure 5 (see p. 410) illustrates the immediate increase in relative L1 power seen
within 15 s after inflation of a balloon catheter in the left carotid artery. Paradoxi-
cally, the shift in relative L1 power excess occurs primarily in contralateral brain re-
gions. This shift is a reflection of "cerebral steal" and shows that hemodynamic com-
pensation can lead to paradoxical ischemia remote from the most compromised cere-
bral vessel. We have observed similar paradoxical ischemia after test clamping of
branches of cerebral arteries, suggesting backfilling or failure in autoregulation.

Relative Sensitivity of 133Xe, QEEG, and Conventional EEG

as Indices of Cerebral Ischemia
In a collaborative study with Jonkman et al. [20], we studied 93 patients 14 days after
a cerebrovascular insufficiency. These patients were divided into four groups: com-
pleted strokes (CS), persisting neurological symptoms (PNS) , reversible ischemic
neurological deficits (RIND), and transient ischemic attacks (TIA). All patients
were evaluated using 133Xe to measure rCBF and a multivariate QEEG feature to
measure electrical indications of ischemia. Figure 6 shows that the multivariate
QEEG feature (overall shift in EEG frequency spectrum across the whole head) was
consistently more sensitive to abnormality in these patients than 133Xe. The superior
sensitivity of the QEEG was especially noteworthy in the asymptomatic RIND and
TIA groups. Members of these groups seldom displayed any focal abnormalities, but
rather revealed a distributed abnormality suggesting that hemodynamic compensa-
tion of local vascular compromise led to a generalized suboptimal cerebral perfusion.
This distributed abnormality could be well detected by the multivariate QEEG fea-
ture but was poorly reflected in 10cal 133 Xe measures of rCBF.

Sensitivity of QEEG to Small Reductions in Cerebral Blood Flow

Not only does QEEG appear to be more sensitive than 133Xe as an index of cerebral
ischemia, as just demonstrated, but the decrease in flow necessary to detect EEG
changes is much smaller than previously believed.
Figure 7 (see p. 410) is a photo of the face of a new intraoperative monitoring
device which we are evaluating, called CIMON (cardiovascular intraoperative moni-
Monitoring Brain Function During Cardiovascular Surgery 415

PERSISTING REVERSED TRA NSIENT ~

COMPLET ED ISCHEMIC AVERAGE ~
STROKE NEUROLOGICAL NEUROLOGICAL
SYMPTOMS DEFECIT
ISCHEM IC
ATTACK ABNORMAL $

100 -
-.J
<l
~ 80 -
o
~ 60 - -
<l
~ 40 -

20 -
o~ ~ ~- ~ ~~ ~
OEEG reaF OEEG reaF OEEG reaF OEEG reBF
(n'll) ( n ' 4 3) (n'15) ( n · 25 )

Fig. 6. Percentage of abnormal QEEG and rCBF findings in patients with either completed strokes,
persisting neurological symptoms, reversible ischemic neurological deficits, or transient ischemic
attacks. Also shown is the average percentage of abnormal findings across patient groups for
QEEG, rCBF, and a conventional EEG interpretation. The percentage of abnormal QEEG findings
in a sample of normals is also presented

tor , Cadwell Laboratories, Kennewick , WA) . Instead of the familiar interpolated

topographical map, this device provides a picture of the history of the EEG fre-
quency spectrum in each lead across a period of observation. In brief, successive 2.5-s
EEG segments are subjected to frequency analysis using the fast Fourier transform
(FFT). A sliding window is constructed which averages N such segments to achieve
statistical reliability. In this example, the sliding window is 15 s long and updates
every 2.5 s. Instead of the conventional compressed spectral array usually used to
represent such data , the average absolute or relative power in the .1, (), a, and f3
bands are computed for each sliding window, and can be Z-transformed relative to
population norms or to some previous state of the patient, called a self-norm. The
statistical significance of the deviation between each sliding window and the refer-
ence norm is color coded separately for each frequency band. This display is like a
compressed spectral array viewed from above and divided into four frequency bands,
with the significance of the deviation from the reference norms color coded for four
segments of a line.
For each rectangle in Fig. 7, four columns are displayed, corresponding to the ,1,
(), a, and f3 relative power in each electrode, Z-transformed with respect to the anes-
thetized self-norm of this patient, obtained on bypass at full flow (3800mllmin) at
27°C. The oldest data in each column are at the top and the most recent at the bot-
tom. At the two times indicated by the arrows, flow was reduced about 8%, from
3800mllmin to 3500mllmin at a constant mean arterial pressure of 65mmHg. Each
time , flow was restored to 3800 mllmin after 60 seconds. The increase in relative .1
power can be seen in many leads as a band of color shifting toward the orange/yellow,
indicating that properly scaled QEEG measures can detect as little as an 8% drop in
rCBF.
416 E.R.John et al.

Evidence of Impaired Cerebral Circulation in Patients

with Atherosclerosis of Coronary Arteries

The incidence of chronic cerebral ischemia was studied in 11 coronary artery bypass
patients for whom preopertive EEG examinations had been obtained before any pre-
medications had been administered. The patients ranged from 44 to 79 years of age.
Using age-regression norms, Z-scores for relative L1 power were computed for eight
bipolar derivations in each patient.
In six of these patients, no Z-scores significant at the 0.01 level were found in the
relative L1 power values for any of the eight bipolar brain regions. In the other five
patients, Z-scores for relative L1 power values were found to be significant at the 0.01
level or higher. In three, the abnormal L1 activity was focal, occurring in only one re-
gion; for the remaining two, abnormal values were found to be diffusely distributed
across four to eight regions. Only a total of three significant Z-scores at the P < 0.01
level would have been expected by chance in a sample of this size. These data suggest
that atherosclerosis causes diminution of flow in cerebral vasculature as well as coro-
nary vasculature in a significant proportion of patients undergoing heart surgery.
Such patients are reasonably to be considered selectively at risk for postoperative
neurological impairment because significant cerebral ischemia during hypoperfusion
is more likely to occur in brain regions where partial stenosis of major vessels already
has caused a reduction in blood flow.

Intraoperative Reversal of Signs of Cerebral Ischemia

This patient was a 66-year-old female with a history of rheumatic fever and a myo-
cardial infarct 4 years earlier. Since then, the patient had been in chronic congestive
heart failure and atrial fibrillation. Six months earlier, the patient had a stroke with
CAT scan documentation of infarct lesions in the left posterior temporal and parieto-
occipital regions. Although no paresis remained, the patient nonetheless still had dif-
ficulty in reading comprehension and remembering recent events. A cardiac catheter-
ization revealed a 70-mm gradient across the aortic valve calcifications. Diagnoses of
aortic and mitral valve stenoses were made. The patient underwent an aortic valve
replacement and an open mitral commissurotomy.
Preoperatively, the patient's EEG was evaluated with the BSA. The relative L1
power map showed significant signs of ischemia in several brain regions, as can be
seen in the head map in Fig. 8A (see p. 411). These regions included the left pos-
terior temporal (Ts), bilateral occipital (01. O 2), and left parietal (P3), concordant
with the previously CT-scan-documented infarct lesions.
Continuous intraoperative EEG monitoring was performed. The patient was put
on cardiopulmonary bypass at 4400 mllmin and cooled to 27°C. Twenty-three minutes
after going on bypass the mean arterial pressure (MAP) was 75 mmHg. Severe L1 excess
developed, most markedly on the anterior left hemisphere, but extending into right
anterior hemispheric regions (Fig. 8B, p. 411). It was noteworthy that the regions
displaying signs of ischemic distress were not those compromised by the prior CVA,
suggesting complex hemodynamic changes consequent to the earlier stroke. Flow
rate of 4400 ml/min was maintained. After these persistent ischemic signs appeared,
injections of vasopressor were administered. Successive injections were required to
Monitoring Brain Function During Cardiovascular Surgery 417

raise MAP to 170mmHg. At this elevated blood pressure, all signs of cerebral ischemia
were successfully reversed (Fig. 8C, p. 411). During the next hour, MAP fluctuations
occurred, and signs of cerebral ischemia including L1 excess and paroxysmal activity
appeared sporadically and disappeared spontaneously, although flow rate remained
constant at 4400 mllmin. Fifty-eight minutes after the last injection of vasopressor,
the perfusate temperature was raised to 31°C. MAP dropped from 220 to 35 mmHg
during the next 40 min, perhaps reflecting generalized vasodilation. Slow wave activ-
ity increased and generalized paroxysmal activity appeared (Fig. 8D, p. 411), now
manifested in a quite different region on the right hemisphere. Vasopressor injec-
tions raised MAP to 80. Prompt disappearance of paroxysmal activity and of the L1
excess indicated reversal of the cerebral ischemia (Fig. 8E, p. 411). The patient was
removed from CPB uneventfully. Upon testing 14 days postoperatively, all relative
L1 power values were within narmallimits (Fig.8F, see p. 411). The preoperative
ischemic sequelae to the CVA had disappeared. After discharge, reading compre-
hension improved slowly but steadily.

Comparison of Intervened Vs Nonintervened CPB Patients

In a subset of the patients whom we studied, it was possible to obtain QEEG mea-
sures before and 7-10 days after CPB, in the absence of any medication. In one sub-
group of these patients, continuous intraoperative EEG monitoring had been per-
formed, with modification of flow rate or blood pressure to reverse QEEG signs of
cerebral ischemia which persisted for several minutes. In the other subgroup, no
intraoperative monitoring or compensatory maneuvers occurred.
Table 1 presents the t tests for changes in the mean Z-value of relative L1 power
across all leads, comparing the preoperative with the postoperative QEEG examina-
tions of these two groups of CPB patients. The mean shift in L1 far the "monitored"
and intervened group is very close to zero, indicating that no significant postopera-
tive deficit in rCBF occurred in members of this group. However, for the nonmoni-
tared and nonintervened group, the mean increase in L1 is significant at more than
the P:5 0.01 level, presumably reflecting a diffuse postoperative decrease in rCBF.
Figure 9A (p. 411) shows the map of changes in L1 activity obtained by comparing
postoperative with preoperative QEEG in a typical intervened CPB patient. Diffuse

Table 1. Change in mean Z value L1 relative power

across all leads before and after CPB surgery in pa-
tients monitored and not monitored via QEEG dur-
ing the surgical procedure

Relative Absolute
power power

Intervened -0.01 -0.24

Nonintervened +2.77 +2.33
10.70 8.03
P SO.OO1 SO.OOl
418 E.R.John et al.

blue color indicates better brain perfusion following surgery. Figure 9B similarly il-
lustrates QEEG changes in a typical nonintervened CPB patient. Diffuse red color
indicates poorer brain perfusion postoperatively.
Although, unfortunately, no prospective behavioral measures were available for
these patients, the high correlation demonstrated in other studies to exist between
increases in slow wave, decreased rCBF, and behavioral impairment supports our
belief that continuous QEEG monitoring and reversal of intraoperative signs of isch-
emia should greatly diminish postoperative behavioral deficits [40, 42]. Prospective
studies of behavioral as well as QEEG changes, with and without intervention to
reverse intraoperative QEEG signs of cerebral ischemia, are currently in progress to
determine whether or not this inference is correct.

Discussion

The numerous previous studies cited in the introduction provided substantial support
for the belief that decreased regional cerebral blood flow, below a critical level of
flow, was reflected by a slow wave shift in the EEG. Reversal of these signs of cere-
bral ischemia prevented postoperative neurological deficits, which often appeared if
the cerebral ischemia indicated by those EEG signs was allowed to persist.
This paper has added several new observations to this body of data. Quantitative
analysis in instances when cerebral blood flow was known to be almost zero showed
that increases in the Z-values of relative L1 power were particularly well correlated
with cerebral blood flow. By use of normative age-regression equations, it was found
that the incidence of cerebral ischemia, as reflected in significant positive Z-values
for relative L1 power, was markedly higher in many candidates for coronary artery
bypass surgery than in normal age-matched controls. This suggests that coronary
artery disease may often be a manifestation of widespread vascular compromise ex-
tending to cerebral vessels. Numerous intraoperative observations revealed that
electrical signs of cerebral ischemia appeared repeatedly during many CPB proce-
dures, with the anatomical pattern of distress typically shifting in successive episodes
within the same patient. These QEEG indices of ischemia could be readily reversed
by increasing flow or mean arterial pressure. Comparison of patients with and with-
out intraoperative monitoring and reversal of signs of cerebral ischemia showed a
widespread postoperative increase in slow EEG activity in only the nonmonitored,
nonintervened patients.
Preoperative histories of cerebrovascular disease, or abnormal relative L1 power
data obtained when preoperative evaluations are referred to normative age-matched
norms, can alert the attending staff to the need for greater precaution. QEEG signs
of cerebral ischemia during continuous intraoperative monitoring indicate the im-
minence of cerebral damage. The alerted surgeon can then implement corrective
measures, whose adequacy can be confirmed by the return of the brain waves to nor-
mal. Intraoperative cerebral damage may thus be substantially reduced.
With respect to the issue of whether micro embolization or hypoperfusion is the
major cause of intraoperative cerebral injury, our data show a direct cause and effect
relationship, with an abnormal pattern reflecting imminent injury appearing with de-
Monitoring Brain Function During Cardiovascular Surgery 419

creased MAP or perfusion, rapid restoration of a normal pattern when MAP or per-
fusion are raised, and gradual reappearance of abnormal brain activity (often in a
different region) if perfusion or MAP are again diminished. This pattern of distress
and recovery is highly consistent with the interpretation that hypoperfusion rather
than micro embolization is the major current cause of CPB-associated cerebral in-
jury. Why should emboli be dislodged reliably by increased flow and why should
debris be released to produce emboli when flow is reduced?
Our observations in the 80 cases we have monitored indicate that hypoperfusion
can occur either when MAP or perfusion rates drop below a critical level. The litera-
ture indicates that the critical level of cerebral blood flow to maintain normal brain
electrical activity is about 20 mlllOO g/min, which corresponds to a mean arterial
pressure of about 50mmHg. Our data indicate that regional blood flow can vary sig-
nificantly among brain regions, and within a single region at different times, even
though total flow rate or mean arterial pressures are apparently safe. These variables
should be considered as relatively independent, and both must be maintained within
safe limits. Signs of cerebral ischemia reliably appear when MAP is allowed to drop
below 55mmHg. However, regional susceptibility may vary markedly because of
cerebrovascular disease. Protection of local brain regions against injury by a priori
selection of systemic CPB parameters would appear to provide less protection than
continuous intraoperative monitoring of the EEG in relation to MAP and flow rate.
The observations reported here are characteristic of patients undergoing coro-
nary artery bypass as well as aortic valve replacement. The patient used as a case
history in this paper had suffered a CVA prior to surgery and received valve replace-
ment. We chose this example because a preoperative neurological deficit disappeared
following aortic valve replacement. This raises the intriguing possibility that valve re-
placement improves marginal cerebral perfusion and may help reverse long-standing
functional deficits caused by cerebral ischemia.

References

1. Aberg T, Ronquist G, Tyden H, Brunnkvist S, Hultman J, Bergstrom K, Lilja A (1984) Adverse

effects on the brain in cardiac operations as assessed by biochemical, psychometric and radio-
logic methods. J Thorac Cardiovasc Surg 87: 99-105
2. Adelman MH, Jacobson E (1960) Electroencephalography in cardiac surgery. Am J Cardiol26:
763-772
3. Baker JD, Gluecklich B, Watson CW, Marcus E, Kamat V, Callow AD (1975) An evaluation
of electroencephalographic monitoring for carotid study. Surgery 78: 787 -794
4. Beebe HG, Pearson JM, Coatsworth JJ (1978) Comparison of carotid artery stump pressure and
EEG monitoring in carotid endarterectomy. Ann Surg 44: 655
5. Branthwaite MA (1975) Prevention of neurological damage during open heart surgery. Thorax
30:258-261
6. Breuer AC, Furlan AJ, Hanson MR, Lederman RJ, Loop FD, Cosgrove DM, Greenstreet RL,
Estafanous FG (1983) Central nervous system complications of coronary artery bypass graft sur-
gery: prospective analysis of 421 patients. Stroke 14: 682-687
7. Brierley JB, Prior PF, Calverley J, Jackson SJ, Brown AW (1980) The pathogenesis ofischaemic
neuronal damage along the cerebral arterial boundary zones in papio anubis. Brain 103: 929-965
8. Chiappa KH, Burke SR, Young RR (1979) Results of electroencephalographic monitoring dur-
ing 367 carotid endarterectomies: use of dedicated minicomputer. Stroke 10: 381-388
420 E. R. John et al.

9. Chiappa KH, Young RR (1977) Carotid endarterectomy monitoring with a dedicated mini-
computer. Electroencephalogr Clin Neurophysiol43: 518
10. Chiappa KH, Young RR, Jones TIl, Crowell RM, Ojemann RG (1978) The usefulness of EEG
monitoring during carotid endarterectomy. Neurology 28:341-347
11. Gasser T, Bacher P, Mocks J (1982) Transformation towards the normal distribution of broad
band spectral parameters of the EEG. Electroencephalogr Clin Neurophysiol 53 : 119-124
12. Gonzalez-Scarano F, Hurtig HI (1981) Neurologic complications of coronary artery bypass
grafting: case-control study. Neurology 31: 1032-1035
13. Grundy BL, Sanderson AC, Webster MW, Richey ET, Procopio P, Karanjia PN (1981a) Hemi-
paresis following carotid endarterectomy: comparison of monitoring methods. Anesthesiology
55:46-52
14. Grundy BL, Webster MW, Nelson P, Sanderson AC, Karanjia PN, Troost BT (1981b) Brain
monitoring during carotid endarterectomy. Anesthesiology 55 : A 12
15. Harris EJ, Brown WH, Pavy RN, Anderson WW, Stone DW (1967) Continuous electroen-
cephalographic monitoring during carotid artery endarterectomy. Surgery 62: 441-447
16. Heiss WD, Waltz AG, Hayakawa T (1975) Neuronal functions and local blood flow during
experimental cerebral ischemia. In: Harper AM, Jennett WB, Miller JD, Rowan JO (eds)
Blood flow and metabolism in the brain. Churchill Livingstone, Edinburgh, pp 27-28
17. Horton DA, Fine RD, Lethlean AK, Hicks RG (1974) The virtues of continuous EEG monitor-
ing during carotid endarterectomy. Aust NZ J Med 4: 32-40
18. John ER, Prichep L, Ahn H, Easton P, Fridman J, Kaye H (1983) Neurometric evaluation of
cognitive dysfunctions and neurological disorders in children. Prog Neurobiol 21 : 239-290
19. John ER, Prichep L, Easton P (1988) Normative data banks and neurometrics: Basic concepts,
current status and clinical applications. In: Remond A, Lopes da Silva F (eds) Computer analy-
sis of EEG and other neurophysiological variables: clinical applications. EEG Handbook, vol 3
(in press)
20. Jonkman EJ, Poortvliet DCJ, Veering MM, deWeerd AW, John ER (1985) The use of neuro-
metrics in the study of patients with cerebral ischemia. Electroencephalogr Clin Neurophysiol
61:333-341
21. Juneja I, Flynn RE, Berger RL (1972) The arterial venous pressures and the electroencephalo-
gram during open heart surgery. Acta Neurol Scand 48 : 163-168
22. Kolkka R, Hilberman M (1980) Neurologic dysfunction following cardiac operation with low-
flow, low-pressure cardiopulmonary bypass. J Thorac Cardiovasc Surg 79: 432-437
23. Lee MC, Geiger J, Nicoloff D, Klassen AC, Resch JA (1979) Cerebrovascular complications
associated with coronary artery bypass (CAB) procedure. Stroke 10: 13-20
24. Loop FD, Cosgrove DM, Lytle BW, Thurer RL, Simpfendorfer C, Taylor PC, Proudfit WL
(1979) An 11 year evolution of coronary arterial surgery (1967-1978) Ann Surg 190: 444-455
25. Malone M, Prior P, Scholtz CL (1981) Brain disease after cardiopulmonary by-pass: correlations
between neurophysiological and neuropathological findings. J Neurol Neurosurg Psychiatry 44:
924-931
26. Marshall BM, Loughead WM (1969) The use of electroencephalographic monitoring during
carotid endarterectomy, as an indicator for the application of a temporary by-pass. Can Anaes-
thesiol SocJ 16:331-335
27. Matsumoto GH, Baker JD, Watson CW, Gleucklich B, Callow AD (1976) EEG surveillance as
a means of extending operability in high risk carotid endarterectomy. Stroke 7: 554-559
28. McKay RD, Sundt TM Jr, Michenfelder JD, Gronert GA, Messick JM, Sharbrough FW, Piep-
gras DG (1976) Internal carotid artery stump pressure and cerebral blood flow during carotid
endarterectomy: modification by halothane, enflurane and innovar. Anesthesiology 45: 390-399
29. Muraoka R, Yokota M, Aoshima M, Kyoku I, Nomoto S, Kobayashi A, Nakano H, Veda K,
Saito A, Rojo H (1981) Subclinical changes in brain morphology following cardiac operations as
reflected by computed tomographic scans of the brain. J Thorac Cardiovasc Surg 81: 364-369
30. Myers RR, Stockard JJ, Saidman U (1977) Monitoring of cerebral perfusion during anesthesia
by time-compressed Fourier analysis of the electroencephalogram. Stroke 8: 331-337
31. Ojemann RG, Crowell RM, Roberson GR, Fisher CM (1975) Surgical treatment of extracranial
carotid occlusive disease. Clin Neurosurg 22:214-263
32. Perez-Borja C, Meyer JS (1965) Electroencephalographic monitoring during reconstructive sur-
gery of the neck vessels. Electroencephalogr Clin Neurophysiol 18: 162-169
Monitoring Brain Function During Cardiovascular Surgery 421

33. Phillips MR, Johnson WC, Scott RM, Vollman RW, Levine H, Nabseth DC (1979) Carotid
endarterectomy in the presence of contralateral carotid occlusion: the role of EEG and intra-
luminal shunting. Arch Surg 114: 1232-1239
34. Rabiner LR, Gold B (1975) Theory and application of digital signal processing. Prentice-Hall,
Englewood Cliffs
35. Rampil II, Holzer JA, Quest DO, Rosenbaum SH, Correll JW (1983) Prognostic value of com-
puterized EEG analysis during carotid endarterectomy. Anesthesiol Analg 62: 186-192
36. Russell RWR, Bharucha N (1978) The recognition and prevention of border zone cerebral
ischaemia during cardiac surgery. Q J Med 47: 303-323
37. Sachdev NS, Carter CC, Swank RL, Blachly PH (1967) Relationship between post-cardiotomy
delirium, clinical neurological changes and EEG abnormalities. J Thorac Cardiovasc Surg 54:
557-563
38. Salerno TA, Lince DP, White DN, Lynn RB, Charrette EJP (1978) Monitoring of electroen-
cephalogram during open-heart surgery. J Thorac Cardiovasc Surg 76: 97-105
39. Sharbrough FW, Messick JM, Sundt TM (1973) Correlation of continuous electroencephalo-
grams with cerebral blood flow measurements during carotid endarterectomy. Stroke 4: 674-683
40. Smith PLC, Treasure T, Newman SP, Joseph P, Ell PJ, Schneidau A, Harrison MJG (1986)
Cerebral consequences of cardiopulmonary bypass. Lancet I: 823-825
41. Sotaniemi KA (1980a) Brain damage and neurological outcome after open-heart surgery. J
Neurol Neurosurg Psychiatry 43: 127-135
42. Sotaniemi KA (1980b) Clinical and prognostic correlates of EEG in open-heart surgery patients.
J Neurol Neurosurg Psychiatry 43: 941-947
43. Stockard JJ, Bickford RG, Schauble JF (1973) Pressure dependent cerebral ischemia during
cardiopulmonary bypass. Neurology 23: 521-529
44. Sundt TM, Houser DW, Sharbrough FW (1977) Carotid endarterectomy: results, complications
and monitoring techniques. In: Thompson RA, Green JR (eds) Advances in neurology, vol 16.
Raven, New York, pp 97-119
45. Sundt TM, Sharbrough FW, Anderson RE, Michenfelder JD (1974) Cerebral blood flow mea-
surements and electroencephalograms during carotid endarterectomy. J Neurosurg 41: 310-320
46. Sundt TM, Sharbrough FW, Piepgras DG, Kearns TP, Messick JM, O'Fallon WM (1981) Corre-
lation of cerebral blood flow and electroencephalographic changes during carotid endarterctomy.
Mayo Clin Proc 56:533-543
47. Sundt TM, Sharbrough FW, Trautmann JC, Gronert GA (1975) Monitoring techniques for
carotid endarterectomy. Clin Neurosurg 22: 199-213
48. Trojaborg W, Boysen G (1973) Relation between EEG, regional cerebral blood flow and inter-
nal carotid artery pressure during carotid endarterectomy. Electroencephalogr Clin Neurophysiol
34: 61-69
49. Turnipseed WD, Berkoff AA, Belzer FO (1980) Postoperative stroke in cardiac and peripheral
vascular disease. Ann Surg 2: 365-368
50. Witoszka MM, Tamura H, Indeglia R, Hopkins RW, Simeone FA (1973) Electroencephalo-
graphic changes and cerebral complications in open-heart surgery. J Thorac Cardiovasc Surg
66:855-864
Chapter 34
Cerebral Blood Flow During Cardiopulmonary Bypass
L. HENRIKSEN

Introduction

Since the introduction of cardiopulmonary bypass in 1953 many improvements have

been made. By the end of the 1970s, the general mortality rate for open heart sur-
gery had fallen to about 2%-3%. In the last few years further improvements have
been reported, and the present mortality rate is probably lower: in patients who had
coronary venous bypass grafting, the mortality rates can be below 1%. As the mor-
tality rates are so low, they are of little use in evaluating the markedly different tech-
niques for extracorporeal circulation. Other criteria thus have to be considered.
The patients almost always show signs of neuropsychological dysfunction follow-
ing bypass: only its severity, duration, and whether or not the changes are reversible
are open to discussion. The incidence of CNS dysfunction is around 30% -60% in
most studies (depending on the sensitivity of the tests), being highest in prospective
investigations undertaken soon after surgery. Furthermore, it was shown that cere-
bral blood flow (CBF) was reduced following cardiac surgery [17], and a flow pattern
was found which resembled the "delayed" hypoperfusion which follows known epi-
sodes of ischemia [30]. During bypass markedly abnormal CBF values have been ob-
served indicating that the optimal bypass technique has yet to be evolved [13,15,16,
18-20].
The brain's homeostatic mechanisms are undoubtedly under great strain during
cardiopulmonary bypass. The type and depth of anesthesia undergo changes, and
hematocrit is usually lowered to around 25% , sometimes to even less. Core temper-
ature is reduced to below 30°C, with brain temperature being even lower. Perfusion
pressures below the lower limit of CBF autoregulation are often reached, with low
pressure even being accepted for long periods as a state of controlled shock. The
extracorporeal circulation is usually nonpulsatile, and microemboli (bubbles and
particulate matter) are unavoidable even with membrane oxygenators and extensive
arterial filtration [19]. For these reasons the assumption that cerebral perfusion is
adequate during such conditions must be seriously questioned.
The influence and possible pathogenetic importance of these various factors are
not easily determined and separated. The present review will discuss the major fac-
tors affecting CBF during bypass and consider their possible clinical implications.
Cerebral Blood Flow During Cardiopulmonary Bypass 423

Regulation of Cerebral Blood Flow

Autoregulatiou

Since the regulation of CBP and the concept of autoregulation are well described in
anesthetized and awake man [2, 3, 23-25, 27], I will confine the present account to
some key features of pathophysiological importance during extracorporeal circula-
tion.
Only limited information has been published about CBP regulation during bypass
[5,6,8, 11, 13, 16,20,32]. Regulation of CBP, mainly subserved by precapillary re-
sistance vessels, occurs following endogenous metabolic alterations and exogenous
influences which threaten to upset the coupling between metabolism and CBP. The
regulation is therefore both localized, with a tight coupling to neuronal activity, and
more general, with changes in cerebrovascular resistance in response to altered sub-
strate supply, e.g., during hypoxia, hypoglycemia, hypothermia, and hypotension.
The coupling between regional CBP and metabolism and neuronal activity has been
demonstrated by means of dynamic emission tomography [12, 28, 29, 31]. There are
no reasons to believe that the coupling between flow and metabolism and its re-
lationship to neuronal activity is different during hypothermic cardiopulmonary by-
pass as compared with normothermia without bypass. Of importance, however,
is the more general regulation of CBP to changes in blood pressure, temperature,
hematocrit, PaC02, and Pa02. The importance of these factors has recently been de-
scribed [20]. Their effects on CBP during bypass are summarized in Pigs. 1-6.
Cerebral autoregulation maintains a constant CBP within wide limits of systemic
blood pressure (50-150mmHg) by dilatation or constriction of intracerebral resis-
tance vessels. Outside these pressure limits, alterations in perfusion pressure result
in changes in CBP. The classically defined lower limit of CBP autoregulation is not
the maximum point of cerebral vascular dilatation, but it reflects rather the point
where CBP becomes pressure dependent. The lower limit of CBP autoregulation
varies considerably between patients and may be positioned at a much higher pres-
sure in hypertensive patients. The cerebral vascular response to changes in perfusion
pressure can be described as consisting of two resistances in series: the extraparen-
chymal vessels being influenced by the autonomic nervous system, while the intra-
parenchymal vessels are regulated by intrinsic mechanisms. Before maximal auto-
regulatory vasodilation of the intraparenchymal vessels is achieved, it might be ex-
pected that any vasoconstriction of larger extraparenchymal vessels would be met by
compensatory vasodilation of intraparenchymal vessels, thereby maintaining CBP
constant. Once the lower limit of autoregulation has been exceeded (maximal auto-
regulatory dilatation of intraparenchymal vessels achieved), then any constriction of
the extraparenchymal vessels would tend to further reduce CBP. Therefore, a partial
or total blockade of the sympathetic nervous system reduces or abolishes the effect
of extraparenchymal vessels and shifts the lower limit of autoregulation toward lower
blood pressure.
CBP autoregulation is of course important, but not too much emphasis should be
placed on the lower limit of autoregulation. This is, in any case, a variable point
which depends on the resting systemic pressure of the patient and on the means of
424 L. Henriksen

Figs. 1-4. CBF autoregulatory curves which

CEREBRAL BLOOD FLOW show the response to changes in MABP at
mI1(100g·min)
90 different levels of Hct, PaC02, Pa02, and
temperature. At low Hct and Pa02 the auto-
80
regulation is shifted up and rightward, but
70
without impairment of the CBF autoregula-
Htc =20·/.
tion as noted at high levels of PaC02 where
60 Htc = 25"10 the autoregulation is almost abolished. CBF
Htc=30"lo
is reduced with falling temperature, but the
50
autoregulation is unaffected. CBF, cerebral
40 blood flow (mIIlOO g . min); MABP, mean
arterial blood pressure (mmHg); Het, he-
30 matocrit (%); PaCOZ, plasma carbon di-
oxide tension (mmHg); PaOZ, plasma
20
MEAN ARTERIAL oxygen tension (mmHg). (Modified from
BLOOD PRESSURE Henriksen [20])
1 30 40 50 60 70 80 90 mmHg

CEREBRAL BLOOD FLOW

ml/(lOOg·min)

110 Low PC02:(40mmHg

Medium PC02:40-50mmHg
100 High PC02: >50mmHg

90

80

70

60

50

40

30

20
MEAN ARTERIAL
BLOoo PRESSURE
2 30 40 50 60 70 80 90 100 mmHg

inducing the hypotension. The worst situation is hemorrhagic hypotension because

this causes sympathetic activation with constriction of extracerebral inflow tract ves-
sels. With drug-induced hypotension the lower limit is pushed toward lower pressure
levels [4]. The situation would be most serious in hypertensive patients in whom the
lower limit of autoregulation is shifted toward higher pressure levels. The modest
falls in CBF which occur at pressures just below the lower limit of autoregulation
need not have any special clinical or biological significance because the oxygen ex-
traction can normally be increased: CBF can normally be halved by hyperventilation
without producing significant brain dysfunction. The important clinically relevant
points on the autoregulatory curve lie some way below the lower limit and are those
at which the various indicators of ischemic brain dysfunction appear.
Cerebral Blood Flow During Cardiopulmonary Bypass 425

Figs. 3, 4. Legends on opposite page

CEREBRAl BLOOD FLOW
100 ml/(lOOg·min)
HIGH PC~-
90 LOW pco2 o----o

80
Pa02 = 78 mmHg
70 Pa~ =134 mnYig

60

50 P~=349mmHg

40

30

20

10
MEAN ARTERIAL
BLOOD PRESSURE
O+-~~.-~~.-~~.-~~.-~ ____-.
3 10 20 30 40 50 60 70 80 90 100 mm Hg

CEREBRAL BLOOD FlOW

90 ml/(100g·min)
hypotherm ia<>--o
80 rewarm ing ........

CBF Response to Complete or Incomplete Ischemia

The ischemia that may occur during bypass is the combination of incomplete isch-
emia arising from episodes of reduced blood pressure, focal complete ischemia from
microemboli (gaseous and particulate emboli), and transient capillary stasis. The im-
pact of the severity and the duration of such transient ischemia should be considered
in relation to changes in CBF and oxygen supply. The principal difference between
complete and incomplete ischemia is that the latter allows some substrate supply.
However, whether the ischemia is complete or incomplete, restoration of adequate
perfusion leads initially to hyperemia, later followed by secondary hypoperfusion.
At tissue levels a mismatch between CBF and metabolism results. As the postisch-
426 L. Henriksen

emic metabolic rate is initially low, the tissue first passes through a stage of transient
hyperoxia ("luxury" perfusion). Later, as metabolism increases, a state of relative
tissue hypoxia ("misery" perfusion) may follow. To produce brain damage when
arterial oxygenation is normal, cerebral perfusion pressure must be reduced to 25
mmHg for at least 15 min. This period can probably be further extended during
hypothermic bypass, where total circulatory arrest for about 1 h or more has been
used. However, before oligemic ischemia leads to brain damage, the tissue will prob-
ably have been through a series of pathophysiological events which alone or in com-
bination may show disturbance of brain homeostasis. The impairment of CBF auto-
regulation and the cerebral hyperemia which occur after blood pressure is restored
are well documented. Depending on the extent and severity of the ischemia, the
hyperemia may be short-lasting (minutes) or can last for days or weeks. The luxury
perfusion stems from an acute cerebral metabolic acidosis that accompanies hypoxia
and is related to a local tissue lactic acidosis. The critical cerebral perfusion pressure
below which loss of autoregulation occurs in the posthypotensive period lies between
30 and 40 mm Hg, a value later confirmed in man during bypass [20]. It has also been
shown that oligemia with a CBF reduced to approximately 40% of normal seems to
be a flow threshold below which brain edema begins. This cortical flow threshold for
the formation of edema is just above the threshold for the loss of sensory-evoked
potential amplitude (16-18 ml/lOO g . min) and seems to be about twice the level at
which homeostasis of extracellular K+ and Ca2 + is completely disrupted. Thus, water
accumulation begins before the integrity of the ionic pumps in the cell membrane is
completely lost. There is also a time threshold for edema formation; in animal experi-
ments at normothermia it seems to be around 30 min [1], but during hypothermia it
can probably be further extended.
In the view of the above findings, it can be seriously doubted whether hypoten-
sive periods should be allowed at all during bypass if they can in any way be avoided.
Hypotensive episodes undoubtedly impair CBF autoregulation during part or the
whole of bypass, and the impairment may even extend into the postoperative period.

Effect of Hemodilution on CBF

Hemodilution is important during bypass because of the necessity of saving blood

and preventing complications due to the homologous blood syndrome. With reduced
hemoglobin oxygen delivery capacity is reduced, and the cerebral milieu maintains
the local concentration of O2 necessary for metabolism by either increasing CBF or
increasing the oxygen extraction. Kawashima and co-workers [21] and Michenfelder
and Theye [26] found a compromised O 2 consumption at hematocrit (Hct) levels
below 20% , indicating a "threshold" concentration of hemoglobin below which local
tissue metabolism may be compromised.
It is well known that high Hct or protein content decreases CBF and that hemo-
dilution or a reduction in protein concentration increases CBF [14]. During hemo-
dilution both oxygen binding capacity and the viscosity of the blood are reduced.
There has been some debate as to whether the CBF changes were due to changes in
the viscosity or in the oxygen carrying capacity of the blood. Both factors playa role,
but probably independently.
Cerebral Blood Flow During Cardiopulmonary Bypass 427

A decrease in arterial oxygen saturation, either through hypoxia or hemodilu-

tion, increased CBF during all ranges from hypo- to hyperventilation in anesthetized
dogs [7]. Koster and co-workers [22] reported that the increase in CBF resulting
from hemodilution during hypothermia was balanced by the reduced O 2 carrying
capacity such that O 2 delivery was unaltered. CBF autoregulation is not impaired
although the autoregulation curve is shifted to higher flow levels and to the right on
the blood pressure axis with decreasing Hct or Pa02 (Figs. 1-4).
The optimal Hct for oxygen delivery in healthy individuals is around 30%, and
with the possible threshold concentration of approximately 20%, the optimal Hct
during hypothermic bypass is probably somewhere between these two values. There
is no evidence to suggest that 25% is the optimal Hct, but around this value it would
probably be advisable to add blood to the perfusate to prevent further hemodilution.

Effect on CBF Following Changes in CO2

The effect of PaC02 on CBF is possibly mediated through periarteriolar changes in

pH. Periarteriolar extracellular fluid is a bicarbonate-containing fluid with few pro-
tein or other nonbicarbonate buffers. The pH at this site depends on the tension of
the freely diffusible CO 2 and the local bicarbonate concentration.
During cardiopulmonary bypass the periarteriolar environment may change as
temperature falls, becoming more alkaline and containing less CO 2, which results in
vasoconstriction. The frequent and often progressive systemic acidosis which follows
most cardiac operations may only affect CBF insignificantly because the blood-brain
barrier (BBB) prevents acute systemic changes in H+ and HCO"3 from influencing
the cerebral microcirculation. More chronic changes through metabolism and respi-
ration are compensated for by changes in extracellular bicarbonate, but this adapta-
tion is of little importance for the acutely induced changes during bypass. On the
other hand, CO 2 crosses the BBB freely, and changes in PaC02 cause rapid changes
in CBF.
In normal awake or anesthetized man not undergoing cardiopulmonary bypass,
the CBF-PaC02 relationship is well established, the CBF change/1 mmHg being ap-
proximately 3%-4%. During bypass the CBF-PaC02 relationship was normal (4%/
1 mmHg) during all periods of nonpulsatile bypass (Fig. 5) [20]. The CBF-PaC02 re-
sponsiveness was unaffected by Hct and temperature. Only the level of blood pres-
sure influenced the CO 2 response (Fig. 6).
Although mean arterial blood pressure (MABP) can be altered over a fairly wide
range without affecting cortical blood flow in hypo- or normocapnic conditions,
there is a passive pressure-flow relationship in hypercapnia indicating abolished auto-
regulation. This was first demonstrated in dogs [9, 10] and has recently been con-
firmed in man during cardiopulmonary bypass [20]. The explanation is probably that
vessels already maximally or nearly maximally dilated by hypercapnia are unable to
dilate further in response to a lowered blood pressure. As discussed in detail below,
changes in CBF caused by variations in PaC02 are probably neither harmful nor
entail a risk for brain cells, whereas combination of a high PaC02 and hypotension
could probably provoke brain damage which otherwise would not have occurred. To
illustrate this situation, consider a patient with a localized relative ischemia. Within
428 L. Henriksen

CEREBRAL BLOOD FLOW

ml,1100g ·min)
1 Before
100 2 Hypothermia
3 Rewarming
4 Normothermia 3.
90 5 After
4.
80

70

60

50

40

30 Fig. 5. Normal relationship between CBF

and PaC0 2 was found during all five periods
20 of surgery, there being no difference in sen-
CARBON DIOXIDE sitivity between the various periods. (Modi-
10~~'-~~-'~~-T~-T-.~m_m~~~_ fied from Henriksen [20])
10 20 30 40 50 60 70 80

CEREBRAL BLOOD FLOW

mll(100g-min)

90
MABP.68
80

70

;#
MMABP=48
60

50 ,~~."
40

30
Fig. 6. Sensitivity of CBF to changes in PaC02 was
20 markedly influenced by the blood pressure: the
lower the blood pressure, the less the sensitivity to
10
CARBON DIOXIDE CO2 • MABP, mean arterial blood pressure
mmHg
(mmHg). (Modified from Henriksen [20])
10 20 30 40 50 60 70 80

the local ischemic area, the blood vessels will probably be dilated under the influence
of elevated tissue PaC02 and reduced tissue Pa02' This dilatation may well compen-
sate and maintain a normal regional CBF; even if the compensatory increase in CBF
is inadequate, the substrate supply may still be sufficient because the arteriovenous
extraction is increased. However, superimposing the effect of focal ischemia (micro-
emboli), or a further aggravation of the hypoxic oligemia, one must consider a wor-
sening of the intracerebral steal with implications for both selective vulnerability and
recovery. This is especially so if such circumstances are complicated by hypercapnia,
Cerebral Blood Flow During Cardiopulmonary Bypass 429

Fig. 7. Regional CBF changes in

seven patients studied before and
8 days after extracorporeal cir-
culation. Note the delayed hypo-
perfusion in areas known to be
especially vulnerable to ischemic
events (arterial boundary zones)

causing an abolished autoregulation and further aggravation of abnormal CBF hetero-

geneity. This effect was demonstrated by Haggendal and Wins6 [7], who noted that
the reactive increase in CBF following hypoxic stress was more sufficient to maintain
"critical" tissue oxygen tension during hypocapnia than during hypercapnia.
The optimal level of PaC0 2 during bypass has been debated since the start of
extracorporeal circulation (Fig. 7). It seems however reasonable to state that high
PaC02 during bypass is clearly of causative importance in the hyperperfusion syn-
drome that follows bypass conducted with the addition of CO 2 . The author thus
advocated low PaC02 during bypass, the optimal level of which is probably around
25 mmHg at actual temperature.

Conclusion

During bypass microembolization is unavoidable; however, to minimize the formation

of emboli, membrane oxygenators are probably an advantage, and arterial filtration
with small pore size filters can significantly reduce the number of circulating emboli.
Low cerebral blood flow (CBF) just before and during bypass will also reduce the
number of emboli which may enter the brain capillaries. This can be achieved by
selecting an anesthetic technique which results in the lowest dissociation between
CBF and metabolism, e.g., the combination of diazepam and fentanyl. Carbon
dioxide should be kept low, preferably around 25-30mmHg (at actual temperature)
before and during the bypass. There is probably nothing to gain from reducing
hematocrit below 25% . Autoregulation is preserved during bypass with a mean
lower limit of autoregulation of around 50-60 mm Hg; pressure levels below this
must therefore be considered unacceptable. In the hyperoxygenated-hypocarbic
state even lower pressures seem to be tolerated before the lower limit of CBF auto-
regulation is exceeded.
430 L. Henriksen

Acknowledgments. This review is based on earlier published papers on this topic. The studies reported
were supported by the Foundation of 1870, the Danish Heart Foundation, the Johann and Hanne
Weimann Foundation, and the Danish Hospital Foundation for Medical Research (Region of
Copenhagen, the Faroe Islands, and Greenland).

References

1. Bell BA, Symon L, Branston NM (1985) CBF and time thresholds for the formation of ischemic
cerebral edema, and effect of reperfusion in baboons. J Neurosurg 62: 31-41
2. Betz E (1972) Cerebral blood flow: its measurements and regulation. Physiol Rev 52: 595-630
3. Edvinson L, MacKenzie ET (1976) Amine mechanisms in cerebral circulation. Pharmacol Rev
28:275-348
4. Fitch W, Ferguson GG, Sengupta D, Garibi J, Harper AM (1976) Autoregulation of cerebral
blood flow during controlled hypotension in baboons. J Neurol Neurosurg Psychiatry 39: 1014-
1022
5. Fox LS, Blackstone EH, Kirklin JW, Steward RW, Samuelson PN (1982) Relationship of whole
body oxygen consumption to perfusion flow rate during hypothermic cardiopulmonary bypass.
J Thorac Cardiovasc Surg 83: 239-248
6. Govier AV, Reves JG, McKay RD, Karp RB, Zorn GL, Morawetz RB, Smith LR, Adams M,
Freeman A (1984) Factors and their influence on regional cerebral blood flow during non-
pulsatile cardiopulmonary bypass. Ann Thorac Surg 38: 592-600
7. Haggendal E, Winso I (1975) The influence of arterial carbon dioxide tension on the cerebro-
vascular response to arterial hypoxia and to haemodilution. Acta Anaesthesiol Scand 19: 134-
145
8. Halley MM, Reemtsma K, Creech 0 (1958) Cerebral blood flow, metabolism, and brain volume
in extracorporeal circulation. J Thorac Surg 56: 506-518
9. Harper AM, Glass HI (1965) Effects of alterations in the arterial CO2 tension on the blood flow
through the cerebral cortex at normal and low arterial blood pressure. J Neurol Neurosurg
Psychiatry 28: 449-452
10. Harper AM (1966) Autoregulation of cerebral blood flow: influence of the arterial blood pres-
sure on the blood flow through the cerebral cortex. J Neurol Neurosurg Psychiatry 29: 398-403
11. Held K, Gottstein U, Niedermayer W (1969) CBF in non-pulsatile perfusion. In: Brock M,
Fieschi C, Ingvar DH, Lassen NA, Schiirmann K (eds) Cerebral blood flow. Clinical experimen-
tal results. Springer, Berlin Heidelberg New York, pp 94-95
12. Henriksen L, Paulson OB, Lassen NA (1981) Visual cortex activation recorded by dynamic
emission computed tomography of inhaled 133Xenon. Eur J Nucl Med 6: 487-489
13. Henriksen L, Hjelms E, Rygg IH, Skovsted P, Lindeburgh T (1981) Cerebral blood flow mea-
sured in patients during open heart surgery using intraarterially injected 133Xenon. J Cereb
Blood Flow Metab 1: 532-534
14. Henriksen L, Paulson OB, Smith RJ (1981) Cerebral blood flow following normovolemic hemo-
dilution in patients with high hematocrit. Ann Neurol 9: 454-457
15. Henriksen L (1982) Cerebral blood flow in patients before, during and after open-heart surgery.
Acta Neurol Scand [Suppl 90] 65: 168-169
16. Henriksen L, Hjelms E, Lindeburgh T (1983) Cerebral blood flow measured in man by intra-
arterially injected 133Xe: evidence suggestive of intraoperative microembolism. J Thorac Cardio-
vasc Surg 86: 202-208
17. Henriksen L (1984) Cerebral blood flow following cardiac operations. Evidence suggestive of
intraoperative microembolism. Lancet 1: 816-822
18. Henriksen L, Barry DI, Rygg IH, Skovsted P (1986) Cerebral blood flow during early cardio-
pulmonary bypass in man. Effect of procaine in cardioplegic solutions. Thorac Cardiovasc Surg
34:116-123
19. Henriksen L, Hjelms E (1986) The effect of arterial filtration on cerebral blood flow during
cardiopulmonary bypass in man. Thorax 41 : 386-395
Cerebral Blood Flow During Cardiopulmonary Bypass 431

20. Henriksen L (1986) Brain luxury perfusion during cardiopulmonary bypass in humans. A study
of the CBF response to changes in CO 2, 02, and blood pressure. J Cereb Blood Flow Metab 6:
366-378
21. Kawashima Y, Yamoto Z, Manabe H (1974) Safe limits of hemodilution in cardiopulmonary
bypass. Surgery 76: 391-397
22. Koster JK, Van de Vanter SH, Bean J, Collins JJ, Cohn LH (1976) Effect of hemodilution and
profound hypothermic circulatory arrest on blood flow and oxygen consumption of the brain.
Surg Forum 27: 235-237
23. Kuschinsky W, Wahl M (1978) Local chemical and neurogenic regulation of cerebral vascular
resistance. Physiol Rev 58: 656-689
24. Lassen NA (1974) Control of cerebral circulation in health and disease. Circ Res 34: 749-760
25. Lassen NA, Christensen MS (1976) Physiology of cerebral blood flow. Br J Anaesthesiol48:
719-734
26. Michenfelder JC, Theye RA (1969) The effects of profound hypocapnia and dilutional anemia
on canine cerebral metabolism and blood flow. Anesthesiology 31: 449-457
27. Owman CH, Edvinson L, Hardebo JE (1978) Pharmacological in vitro analysis of amine-
mediated vasomotor functions in the intracranial and extracranial vascular beds. Blood Vessels
15:128-147
28. Phelps ME, Maziotta JC, Kuhl DE, Nuwer M, Packwood J, Metter J, Engel J (1981) Tomo-
graphic mapping of human cerebral metabolism: visual stimulation and deprivation. Neurology
31:517-529
29. Phelps ME, Mazziotta JC, Huang S (1982) Study of cerebral function with positron computed
tomography. J Cereb Blood Flow Metab 2: 113-162
30. Siesjo BK (1984) Cerebral circulation and metabolism. J Neurosurg 60: 883-908
31. Sokoloff L (1981) Localization of functional activity in the central nervous system by measure-
ments of glucose utilization with radioactive deoxyglucose. J Cereb Blood Flow Metab 1: 7-36
32. Wollman H, Stephen GW, Clement AJ, Danielson GK (1966) Cerebral blood flow in man dur-
ing extracorporeal circulation. J Thorac Cardiovasc Surg 52: 558-564
Chapter 35

Cerebral Hemodynamics
During Nonpulsatile Cardiopulmonary Bypass
T.LUNDAR

Introduction

Cerebral damage during cardiac operations remains among the most serious and
devastating complications of open heart surgery. A prospective study from Cleve-
land Clinic demonstrated a stroke risk of 5% during standard aortocoronary bypass
operations: 2% of the patients were left with a substantial permanent cerebral deficit
[5] and another 11 % of the patients demonstrated clinical symptoms and signs of dif-
fuse encephalopathy. Enormous numbers of cardiopulmonary bypass (CPB) proce-
dures are now being done worldwide, but our knowledge of cerebral pathophysio-
logical events during and after these operations remains scanty. Early experimental
studies using electromagnetic carotid artery flowmetry indicated that the carotid flow
fluctuated passively with the flow from the heart-lung machine during nonpulsatile
CPB [4, 7].
Early reports on cerebral perfusion during CPB in humans were based on cere-
bral arteriovenous O 2 differences. As the coupling between cerebral blood flow
(CBF) and metabolism seems lost during nonpulsatile CPB [3], such indirect calcula-
tions of CBF is hardly of any value. Recent reports using xenon washout techniques
have yielded conflicting results. While Henriksen and co-workers [8] found a 67% in-
crease of CBF during nonpulsatile CPB, Govier et al. [6] reported a 55% reduction.
The regimen for CPB was not identical, yet such large differences are difficult to ex-
plain. In a small clinical study using direct electromagnetic flowmetry on the internal
carotid artery, we have previously demonstrated enhanced but pressure-passive
cerebral perfusion during moderately hypothermic nonpulsatile CPB [15].
Monitoring of the middle cerebral artery (MCA) flow velocity by transcranial
Doppler (TCD) technique [1] offers new opportunities in the study of cerebral perfu-
sion. The TCD method has proved to be of value in the management of patients with
subarachnoid hemorrhage [2] as well as in patients with precerebral and intracranial
occlusive artery disease [11, 12].
The diameter of the proximal MCA is supposed to be relatively constant [9, 22]
as regulation of cerebrovascular resistance acts on the arteriolar level, both in auto-
regulation of CBF during changes in cerebral perfusion pressure (CPP = mean arte-
rial blood pressure - mean intracranial pressure) as well as during carbon dioxide-
induced changes in cerebral perfusion. Comparative observations on MCA flow ve-
locity and electromagnetically recorded internal carotid artery flow have revealed
that changes in MCA flow velocity reflect concomitant changes in cerebral perfu-
sion. A recent study reports the effect of changes in PC02 on MCA flow velocity in
healthy volunteers [21]. The demonstrated classic CO2 effect on cerebral circulation
Cerebral Hemodynamics During Nonpuisatile Cardiopulmonary Bypass 433

further supports that observed changes in MCA flow velocities are related to similar
changes in MCA volume flow. A recent study confirms that changes in MCA blood
velocity reflect volume flow changes recorded electromagnetically [13].
When the transcranial Doppler technology became available in 1982 [1], the
method was adopted in our cerebral monitoring in cardiac surgery. We have for sev-
eral years performed extensive cerebral monitoring in highly selected cases, sup-
posed to be at high risk for cerebral complications during and after cardiac surgery.
Such risk factors include previous stroke, anticipated CPB time exceeding 2h, mul-
tiple valvular replacement, revalvular replacement, combined valvular replacement
and aortocoronary bypass procedures, and aortic arch aneurysmectomies.
The monitoring included recording of arterial blood pressure (BP) and central
venous pressure (CVP).
Cerebral electrical activity was recorded by a cerebral function monitor (CFM).
In adult patients the epidural intracranial pressure (EDP) was recorded by a mini-
transducer implanted through a burrhole in the right frontal region in advance of the
surgical procedure [14].
Continuous recording of MCA flow velocity throughout the surgical procedure
demonstrated changes in MCA flow velocity at the inception of CPB. As the diameter
of the MCA is unknown in the individual patient, comparison of the flow velocities
between the patients has no meaning. Compared with the flow velocity recorded
during the last 5 min before CPB (pre-bypass level) changes expressed as percent of
pre-bypass level will reflect concomitant changes in cerebral perfusion in the indi-
vidual patient.

Cerebral Perfusion During Moderately Hypothermic Cardiopulmonary Bypass

in Adults

Continuous recording of MCA flow velocity throughout moderately hypothermic

(28°-32°C) nonpulsatile CPB in 28 adult patients revealed markedly increased flow
velocities. At the introduction of CPB, MCA flow velocity increased in spite of rapid
reduction in the CPP. Introduction of the clear priming solution in adults caused a
rapid hemodilution. A typical inception of CPB is demonstrated in Fig. 1. During
steady-state low flow, low-pressure (1.51/m2 body surface/min) CPB at 28°-32°C,
most of these adult patients demonstrated MCA flow velocities in the range of
150%-250% of the pre-bypass value. In spite of the increased flow velocity values,
the flow velocities changed in a pressure-passive manner with the CPP as demon-
strated in Fig. 2. The pressure-dependent changes in MCA flow velocity take place
irrespective of the cause of changes in CPP. Figure 3 demonstrates three different
causes of changed CPP: venous obstruction, increased flow from the heart-lung
machine, and vasopressor infusion. In this patient a wide range of CPP values was
observed within a short time. The corresponding MCA flow velocity value of each
CPP observation is plotted in Fig. 4, demonstrating that the cerebral autoregulation
was not in effect. Such plots are only meaningful when other factors with profound
effect on the cerebral perfusion such as temperature, hematocrit, and PaC0 2 can be
kept constant.
434 T. Lundar

CFM100~
)JV
10 ~~ CFM
100 0 35 36 31 19 19 23 21 22 21 22 22 I 22 Hct
MCA
flow . MCA
velocity 50

80
BP Illll~.-1IJ,I BP
EOP
40
mm Hg ~,,_._ _........
;.....w
! I""''''.'j!'''''''\f-.'''''
,. ''''' ~ EDP
t t t
• b c
o
~--~----~----~----~--~. minutes

Fig. I. Recording of MCA flow velocity, BP, EDP, and CFM at the introduction of cardiopulmonary
bypass in a patient. Repeated hematocrit (Hef) values are also given to illustrate the progress and
stabilization of hemodilution. As the priming solution was introduced (arrow a), the MCA flow
velocity increased rapidly. A biphasic response of the CFM activity was observed. The well-known
drop in BP seen at the introduction of bypass (arrow b) caused the dip in CPP (BP-EDP) as well as
in MCA flow velocity. As BP and CPP were restored (arrow e), MCA flow velocity once again in-
creased, to about twice the initial (pre-bypass) value (Lundar et al. [16])

60
BP 40 BP
EDP
EDP
20
mm Hg
0 t

, ' ml nut ••
•

Fig. 2. Records of MCA flow velocity, BP, and EDP during steady-state CPB in a patient. A sponta-
neous fall in CPP caused profound reduction in MCA flow velocity (Lundar et al. [16])
Cerebral Hemodynamics During Nonpulsatile Cardiopulmonary Bypass 435

CF M 100
)JV
10

MC A 150 0
fl ow
veloc it y
cm/s

o
BP
E DP . ~
40 ~'''''' r ..... .
J'II1,.~.;Jt" ••: .~

....
mmHg ~rY!, . ED P
\"""" ,~

•
o
•a •b c
~~~~--------~~~~ ________ minu t e s

Fig. 3. Records of BP, EDP, MCA flow velocity, and CFM activity during an 18-min period of CPB
in a patient. From a low CP state of 23 mmHg, a short-term obstruction of the venous return caused
an abrupt increase in the central venous pressure and ED P, and a CPP reduction to about 5 mm Hg
(arrow a) . The corresponding fall in MCA flow velocity is clearly seen. Administration of 1 mg
metaraminol (arrow b) and increased pump flow from 1.8 to 2.111min/m2 (arrow c) caused marked
increase in CPP and MCA flow velocity. The marked changes in the intracranial pressure (EDP) are
clearly demonstrated (Lundar et al. [17])

12 0

100
/'
MeA / -A
80
flow /'
. . ./ '
ve loc it y

.•..
em/ s 60 Fig. 4. Plot of corresponding CPP and MCA
" ./' flow velocity values from the record period in
40 Fig. 3. During this period an unusually wide
range of CPP values were observed (5-48
20
mmHg) . The linear relationship indicates a
pressure-passive, nonautoregulatory situation
' - ---:;";.----::!'=-- - - ' - - --'-__- ' C P P
(Lundar et al. [17])
10 20 30 40 50
mm H9

In spite of the impaired autoregulation, the cerebral CO 2 reactivity was well pre-
served in these adult patients. The CO 2 reactivity was easily tested during CPB as
demonstrated in Fig. 5. The PaC0 2 was rapidly reduced by instant increase in the gas
flow to the membrane oxygenator. The cerebral CO2 reactivity during moderately
hypothermic, nonpulsatile CPB was in the range of 2.5%-4.5%/mmHg, which is
considered normal. The CO 2 reactivity is, however, somewhat dependent on the
CPP. The CO 2 reactivity is reduced below CPP levels of about 35mmHg, but ap-
pears to be partially retained down to CPP levels of about 20mmHg. The dissocia-
tion between cerebral autoregulation and CO 2 reactivity as well as the dependency
of the CO 2 reactivity on the CPP is described in detail elsewhere [17, 18].
436 T. Lundar

CFM 100f
10 "". _\. 4..:... . ~... ....... "\". ~.-.,,... . ..,.'.. ....... ~............ ....,. .,. . ........ .. -.
)IV
o

MeA 100 [41':l1li'1"

fl ow
vel ocity 0
cm / s Fig. 5. Records of BP, EDP, MCA flow
velocity, and CFM activity during a COz-
reactivity test. Gas flow to the membrane
oxygenator was increased from 1.0 to 10.0
IImin (arrow a) and reversed to 1.0llmin

]~~
BP (arrow b) . Arterial CO2 tension changed
EOP abruptly from 40.7 to 20.9mmHg. The
mm Hg
marked reduction in MCA flow velocity is
clearly seen. Cerebral perfusion pressure
+ + (CPP) increased during the test from 32 to
44mmHg due to increased BP and simul-
a b
'--_ _--'-_ _----', m i n ut e s taneously reduced EDP (Lundar et al. [18])

Cerebral Perfusion During Hypothermic Cardiopulmonary Bypass

in Children

The regimen for CPB in children in our institution implies a blood-containing prime
solution, which is precooled with subsequent rapid cooling of the child at the incep-
tion of CPB. At the introduction of CPB, a short-lasting increase in MCA flow ve-
locity was observed followed by reduced flow velocities during the progressive cool-
ing and fall in CPP. During steady-state bypass at 20°C MCA flow velocities in 12
children studied ranged 40%-105% of the pre-bypass value observed at CPP (BP-
CVP) in the range of 15-25 mmHg. The hemodilution ratio (pre-bypass hematocrit
divided by hematocrit during steady-state CPB) was about 1.5 in these children com-
pared with 2.0 in the adults. A typical recording from the pediatric series is demon-
strated in Fig. 6. Stepwise changes in the output from the heart-lung machine caused
concomitant changes in the CPP as well as in the MCA flow velocity as demonstrated
in Fig. 7. The pressure-passive changes in flow velocity are evident.

Cerebral Hemodynamics During Cardiopulmonary Bypass

Our recordings in adults during non pulsatile CPP demonstrated a pressure-passive

perfusion with well-preserved CO 2 reactivity. This conclusion was based on electro-
magnetic flowmetry on the internal carotid artery as well as on middle cerebral
artery flow velocity recordings [15-18]. In spite of the pressure-passive characteris-
tics of the cerebral vasculature during nonpulsatile CPB at 28°-30°C, the perfusion
was clearly increased compared with pre-bypass levels. In the pediatric series [20] ,
the cerebral perfusion was pressure passive, but reduced to about 50% of the pre-
bypass level.
Cerebral Hemodynamics During Nonpulsatile Cardiopulmonary Bypass 437

CFM
)'V

MCA 1001.........
flow velocity
cm/s ~"'''''I_''''''''. _______.''___________
o

BP
8
t

,
t
b c

CVP
mm Hg
~ __~____~____~__________~____~__~
-,---- mlnut ••

Fig. 6. Records of BP, CVP, MCA flow velocity, and CFM activity during introduction of CPB in a
child. A transitory increase in MCA flow velocity was observed at the introduction of the priming
solution (arrow a). The progressive reduction in CPP (BP-CVP) caused a marked reduction of MCA
flow velocity (arrow b). At steady-state CPB, MCA flow velocity stabilized at about half the pre-
bypass value (c). The effect of the rapid cooling to about 20°C on the CFM activity is clearly demon-
strated

CFM 100[
.
JAV 10
0 . -• _____~~. J

Pump flow
I/mln :[~
MCA
flow
velocity
cm/s

A B : C D

r
BP
Ie T. :
'Of ~
----
cVP
mm Hg
~
0
minutes

Fig. 7. Records of BP, CVP, MCA flow velocity, and CFM activity during CPB at 20°C in a child.
Repeated changes of 1 min duration of output from the heart-lung machine (A, B, C, D) caused con-
comitant changes in CPP and MCA flow velocity, indicating a pressure-passive, nonregulatory system
438 T.Lundar

Our conclusion with respect to impaired or lost autoregulation during CPB is in

accordance with early experimental studies using carotid flowmetry [4, 7] and to
some degree also in accordance with the studies of Henriksen and co-workers [8]
using xenon washout techniques. Our results are, however, in sharp contrast to other
xenon studies [6, 10], which demonstrated reduced flow during CPB as well as pre-
served autoregulation. It is possible that some of the discrepancies among these
studies are caused by the difference in CPB regimen. In our institution, temperature-
corrected CO 2 values at about 38 mm Hg are used during CPB at 30°C, while Govier
et al. [6] and lohnsson and co-workers [10] advocated temperature-uncorrected CO 2
values. It is well-known that reduction of PC0 2 improves the cerebral autoregula-
tion. It is, on the other hand, difficult to discuss autoregulation based on studies
where the CPP is actually unknown.
The cerebral perfusion during CPB will therefore obviously depend on the regi-
men for CPB [19]. Cerebral microembolism during CPB appears to be one of the
major causes for cerebral damage during cardiac surgery, and CPB time exceeding
120 min has been found to give an increased risk of cerebral complications. High-
flow, high-pressure regimens for CPB imply more mechanical damage to the blood
and increased risk of microembolism. Excessive use of vasopressor during CPB is
also potentially harmful to the brain. The facilitated cerebral perfusion during CPB
seems at least partly caused by the hemodilution. Low-flow, low-pressure regimens
for CPB appear more appropriate than high-flow, high-pressure regimens, provided
that the cerebral perfusion can meet the metabolic demands at the actual tempera-
ture. The regimens for CPB differ vastly among leading centers for cardiac surgery,
indicating that the optimal regimen for CPB is not yet settled. Transcranial Doppler
appears to be a suitable tool in future research for optimal CPB regimens to obtain
maximal brain protection during cardiac surgery. The method furthermore provides
a useful monitoring method to secure the cerebral perfusion during CPB procedures
on an individual base in high-risk patients.

References

1. Aaslid R, Markwalder TM, Nomes H (1982) Noninvasive transcranial Doppler ultrasound

recording of flow velocity in basal cerebral arteries. J Neurosurg 57: 769-774
2. Aaslid R, Huber P, Nomes H (1984) Evaluation of cerebrovascular spasm with transcranial
Doppler ultrasound. J Neurosurg 60:37-41
3. Andersen K, Waaben J, Husum B, Voldby B, Bl1!dker A, Hansen AJ, Gjedde A (1985) Non-
pulsatile cardiopulmonary bypass disrupts flow-metabolism couple in the brain. J Thorac Cardio-
vasc Surg 90: 570-579
4. Ankeney JL, Viles PH (1961) The effect of total body perfusion on carotid blood flow. Surgery
49:209-214
5. Furlan AJ, Breuer AC (1984) Central nervous system complications of open heart surgery.
Stroke 15: 912-915
6. Govier AV, Reeves JG, McKay RD, Karp RB, Zorn GL, Morawetz RB, Smith LR, Adams M,
Freeman AM (1984) Factors and their influence on regional cerebral blood flow during non-
pulsatile cardiopulmonary bypass. Ann Thorac Surg 38: 592-599
7. Halley MM, Reemtsma K, Creech 0 (1958) Cerebral blood flow, metabolism and brain volume
in extracorporeal circulation. J Thorac Surg 36: 506-518
Cerebral Hemodynamics During Nonpulsatile Cardiopulmonary Bypass 439

8. Henrikson L, Hjelrns E, Lindeburgh T (1983) Brain hyperperfusion during cardiac operations:

cerebral blood flow measured in man by intra-arterial injection of xenon 133: evidence sugges-
tive of intraoperative microembolism. J Thorac Cardiovasc Surg 86: 202-208
9. Huber P, Handa J (1967) Effect of contrast material, hypercapnia, hyperventilation, hypertonic
glucose and papaverine on the diameter of the cerebral arteries: angiographic determination in
man. Invest Radiol2: 17-32
10. Johnsson P, Messeter K, Ryding E, Nordstrom L, Stahl E (1987) Cerebral blood flow and auto-
regulation during hypothermic cardiopulmonary bypass. Ann Thorac Surg 43: 386-390
11. Lindegaard KF, Bakke SJ, Grolimund P, Aaslid R, Huber P, Nornes H (1985) Assessment of
intracranial hemodynamics in carotid artery disease by transcranial Doppler ultrasound. J Neuro-
surg 63: 890-898
12. Lindegaard KF, Bakke SJ, Aaslid R, Nornes H (1986) Doppler diagnosis of intracranial artery
occlusive disorders. J Neurol Neurosurg Psychiatry 49: 510-518
13. Lindegaard KF, Lundar T, Wiberg J, Sjfl!berg D, Aaslid R, Nornes H (1987) Variations in middle
cerebral artery blood flow investigated with noninvasive transcranial blood velocity measure-
ments. Stroke 18: 1025-1030
14. Lundar T, Frfl!ysaker T, Nornes H, Lilleaasen P (1982) Aspects of cerebral perfusion in open-
heart surgery. Scand J Thorac Cardiovasc Surg 16:217-222
15. Lundar T, Frfl!ysaker T, Lindegaard KF, Wiberg J, Lindberg H, Rostad H, Nornes H (1985)
Some observations on cerebral perfusion during cardiopulmonary bypass. Ann Thorac Surg 39:
318-323
16. Lundar T, Lindegaard KF, Frfl!ysaker T, Aaslid R, Wiberg J, Nornes H (1985) Cerebral per-
fusion during nonpulsatile cardiopulmonary bypass. Ann Thorac Surg 40: 144-150
17. Lundar T, Lindegaard KF, Frfl!ysaker T, Aaslid R, Grip A, Nornes H (1985) Dissociation be-
tween cerebral autoregulation and carbon dioxide reactivity during nonpulsatile cardiopulmonary
bypass. Ann Thorac Surg 40: 582-587
18. Lundar T, Lindegaard KF, Frfl!ysaker T, Grip A, Bergman M, Am-Holen E, Nornes H (1986)
Cerebral carbon dioxide reactivity during nonpulsatile cardiopulmonary bypass. Ann Thorac
Surg 41 :525-530
19. Lundar T (1986) Cerebral perfusion during cardiac surgery. Lancet 2: 457
20. Lundar T, Lindberg H, Lindegaard KF, Tjfl!nneland S, Rian R, BfI! G, Nornes H (1987) Cerebral
perfusion during major cardiac surgery in children. Pediatr Cardiol 8: 161-165
21. Markwalder TM, Grolimund P, Seiler RW, Roth F, Aaslid R (1984) Dependency of blood flow
velocity in the middle cerebral artery on end-tidal carbon dioxide partial pressure - A trans-
cranial ultrasound Doppler study. J Cereb Blood How Metab 4: 368-372
22. Strandgaard S, Paulson OB (1984) Cerebral autoregulation. Stroke 15: 413-416
Chapter 36
Cerebral Outcome After Open Heart Surgery:
A Long-term Multidimensional Follow-up of Valvular
Replacement Patients
K. SOTANIEMI

Due to improvement in surgical and anesthesiological techniques and in perfusion

equipment during the past two decades, severe cerebral complications related to
open-heart surgery have become rare [17]. Despite all advances, however, central
nervous system (CNS) dysfunction is still one of the major threats in connection with
cardiac operations. Depending on methodological differences, the incidence of CNS
dysfunction has usually varied from 23% to 53% [4, 16, 22, 31, 32] in prospective
studies carried out by neurologists after valvular replacement surgery, while con-
siderably lower values have been reported in retrospective studies and in studies per-
formed by nonneurologists. Thus far, the lowest reported incidence of clinical signs
of CNS dysfunction complicating valvular surgery is 23% [16] when studies applying
properly designed neurological assessments are considered. Tenfold differences be-
tween neurological and surgical evaluations of the CNS outcome have been reported
[27]. Cerebral disorders are not limited to valvular operations; a recent prospective
study on coronary artery bypass surgery [21] revealed neurological complications in
64% of the patients 24 h after the operation.
During the 20-year-old history of open heart surgery, the incidence, nature, and
determinants of cerebral disorders occasionally induced at operations have been
thoroughly scrutinized [2, 5, 15, 17,26]. However, the neurological reports on the
cerebral outcome have been limited to a description of the immediate postoperative
results, the follow-up times having usually only been the first few postoperative
weeks or months [7, 9, 10]. The available long-term studies have concentrated on
cardiological and surgical aspects [3, 4, 8,18,19]. The resultant lack of knowledge of
the neurological outcome and higher cerebral functions includes not only the clinical
aspects but also any subclinical measures. For instance, little is known about the
changes in intellectual functions more than 6 months after operation, and no long-
term electroencephalographical (EEG) studies have been published.
Since CNS disorders related to operative procedures require continuous recon-
sideration and since the available information concerning the long-term neurological
outcome is scarce, the present study was designed in order to elucidate the problems
faced by patients some years after surgery. This study is a continuation of a previous
series of investigations on pre- and postoperative clinical [22], EEG [23], quantita-
tive EEG (QEEG) [24], and neuropsychological [25] aspects of valvular replacement
up to 1 year postoperatively. Multidimensional follow-up has now been extended to
5 years.
Cerebral Outcome After Open Heart Surgery 441

Table 1. Patients

n Valve operation
Aortic Mitral Multiple

Preoperative 65 44 16 5
Deaths during the 1st postoperative year 5 2 3 0
Available at 1 year 60 42 13 5
Deaths during the following years 2 2 2 0
Not available for investigations 3 2 1 0
Available for follow-up at 5 years 55 40 10 5

At follow-up there were 43 men and 13 women, ages ranging from 21 to 71 years (mean 46 ± 10 years)

Patients and Operative Procedures

Of the 65 patients who had undergone valvular replacement and a I-year multi-
dimensional follow-up, 55 were available for reassessment 5 years postoperatively
(Table 1). The mean follow-up time was 5.1 ± 0.3 years (range 4 years 6 months to
5 years 9 months). There were 42 men and 13 women, their age range at last follow-
up being 21-71 years (mean 46 ± 10 years). Aortic replacement had been performed
in 40 cases, mitral in ten, and multiple replacement (aortic and mitral) in the remain-
ing ten cases. The Bjork-Shiley tilting-disc prosthesis had been used in all cases.
All had been operated on using the Rygg-Kyvsgard's bubble oxygenator, moderate
hemodilution, moderate hypothermia, and nonpulsatile flow during the cardiopul-
monary bypass. All received standard pre-, intra- and postoperative care, careful
cardiological surveillance, and continuous anticoagulant treatment after the opera-
tion. The cardiological outcome was assessed to be optimal or highly satisfactory in
all except one patient, who had cardiac failure unrelated to the prosthesis.
According to their postoperative clinical outcome (evaluated 10 days after the
operation), the patients were divided into two groups: (1) those without any detect-
able clinical signs of CNS dysfunction (noncomplicated group, n = 30), and (2) those
with clinical abnormalities regardless of the degree of severity (CNS complicated
group, n = 25).

Investigations

All patients underwent a complete clinical-neurological investigation performed by

the author. EEG recordings were carried out using a 16-channel machine, standard
conditions, and four different electrode montages, the recording time being at least
30 min. The EEG evaluation was performed using conventional methods and the
same grading system as in the short-term EEG report on the series [23]. The QEEG
was performed using fast Fourier transform and the same parameters as described in
the previously reported QEEG study [24].
442 K. Sotaniemi

The neuropsychological examination was carried out using a battery of seven sub-
tests described in detail in the I-year postoperative report [25]: (1) color naming
(time and errors as variables), color vision and reading fluency (a modification of the
Stroop color test [12]); (2) learning (a pure verbal test); (3) triangles (psychomotor
speed); (4) maze time (planning ability and frontal lobe functions); (5) symmetric
drawing; (6) recognition (5 and 6 are tests with strong visual components and thus
measure mainly the functions of the nondominant hemisphere); and (7) digits back-
wards (attention and audioverbal memory). The test scores ranged from 0 to 10. The
overall neuropsychological performance was measured using the sum core of the
seven subtests. Forty-four patients completed the psychometric test follow-up, the
remaining 11 patients were excluded because of their incomplete knowledge of Fin-
nish (their native language being other than Finnish).

Results

Clinical Aspects. Table 2 shows the summary of the long-term clinical findings. The
number of patients showing clinical abnormalities attributable to the operation was
five (9%),5 years later. Five patients reported transient ischemic attacks (TlA) dur-
ing the 5 postoperative years (four further patients had experienced TlA before
operation) but none had suffered from more severe cerebrovascular disorders (in
contrast to four cases of stroke before the operation, Table 3). The vocational status
of the patients is shown in Table 4.

EEG. In the 55 patients, the postoperative prevalence of abnormal EEG was 45%,
in contrast to 25% after 5 years. The respective values for the noncomplicated and
the eNS complicated groups were 43% vs. 20% and 48% vs. 32%. The prevalence
of a bilaterally abnormal EEG after 5 years was 13% in the former and 24% in the
latter patient group. Left-hemisphere disturbances dominated before and after the
operation. Theta-range abnormalities were the most common disturbances (Fig. 1).

Table 2. Presence of clinical signs of CNS dysfunction induced at

operation (55 patients)

Investigation time n CNS signs

Nature n
10 days postoperatively 16 Motor/sensory 12
Cognitive 4
1 year postoperatively 4 Motor 2
Cognitive 2
5 years postoperatively 5 Motor 2
Cognitive 3"

a One patient displayed renewed signs after having recovered during

the first year.
Cerebral Outcome After Open Heart Surgery 443

Table 3. Clinical data (55 patients)

Pre- 1st post- 2-5 post-

opera- operative operative
tion (%) year(%) years (%)
Survival 94 89
Deaths
CNScause 3 0
Cardiological 2 2
Stroke 4 0 0
TIA 4 1 5
Syncope 6 0 0
Seizures 1 2 2
Complications of
anticoagulant therapy 0 0 2

Table 4. Vocational outcome (55 patients)

Atoper- 1 year 5 years

ation (%) (%)
(%)

Able to work 62 55 47
Retired 4 7 11

Sick leave
Cardiological cause 33 30 31
Neurological cause 2 9 11
Neurological % all 5 24 26

MAIN EEG ABNORMALITY

N OF
ABN. EEG
50
%
o THETA
~ SHARP W.
40
• DELTA
30

20

10

PRE-OP. 1 YEAR 5 YEARS

Fig. I. Nature of the main EEG abnormality before operation, and 1 and 5 years postoperatively
444 K. Sotaniemi

Table 5. Follow-up of psychometric test scores

Test Patient Preoperative Score change

group score
Pre- Preoperation Difference
operation to 5 years between the
to 1 year NC and CC
Abso- Pvalue
Absolute lute Pre- At
change change opera- S years
tively

Digits NC 5.8± 1.1 +0.4 0 NS i NS NS

backwards CC 5.5 ± 1.1 -0.3 - 0.1 NS 1
Triangles NC 5.1± 2.3 +0.2 + 0.6 <0.05 i NS NS
CC 5.4± 2.7 -0.5 - 0.1 NS 1
Learning NC 6.2± 1.8 +0.7 + 2.4 <0.001 i NS NS
CC 5.8± 2.7 +0.1 + 0.9 NS i
Recognition NC 5.5± 1.8 +1.2 + 2.5 <0.001 i NS <0.05
CC 4.9± 2.2 +0.7 + 1.3 <0.05 i
Maze time NC 6.1 ± 2.8 -0.1 + 2.7 <0.001 i NS <0.01
CC 6.0± 2.7 -0.5 + 0.5 <0.05 i
Color-text NC 6.0± 3.4 +1.3 + 2.0 <0.01 i NS NS
time CC 4.8± 3.6 +0.3 + 1.9 <0.01 i
Color-text NC 9.1 ± 1.4 -0.2 + 0.6 NS i <0.01 NS
errors CC 6.9± 3.1 +0.4 + 2.6 <0.001 i
Symmetric NC 6.1 ± 1.8 +1.2 + 1.4 <0.01 i NS <0.01
drawing CC 7.0± 3.3 0 -1.7 <0.01 1
Performance- NC 40.6± 10.7 +5.7 +11.9 <0.001 i NS <0.01
sum score CC 39.1 ± 13.5 +0.2 + 4.0 <0.05 i

NC, noncomplicated; CC, complicated; NS, not significant; i, higher than the preoperative value;
L lower than the preoperative value

Neuropsychological Tests. Table 5 provides a summary of the psychometric perfor-

mance before and after surgery. Preoperatively, the only difference between the
clinical groups was performance in the color test (number of errors). Five years after
the operation this difference could not be found any more; instead the eNS compli-
cated group had statistically significant lower scores than the noncomplicated group
in three of the subtests (recognition, symmetric drawing, and maze time). The latter
group had a rise in the scores of six subtests while the eNS-complicated group ex-
perienced a rise in four and fall in three subtests. In contrast to an equal performance
with regard to sum score values before operation, the eNS complicated group had a
significantly lower value after 5 years (Fig. 2, Table 5). It is worth mentioning that
the sum score difference between the two groups first reached statistical significance
after 5 years. The eNS-complicated group experienced a score fall after the first year
while the noncomplicated group continued its rising course (Fig. 2).
Cerebral Outcome After Open Heart Surgery 445

PERFORMANCE
SCORE
CHANGE

12
11 •• ~ NON- COMPLIC.
10 o
g.
Fig. 2. Follow-up of the neuropsycho-
a logical performance sum score of
the patient groups with (CNS com-
plicated, dashed line) and without
(noncomplicated, solid line) neuro-
logical dysfunction associated with
operation (evaluated 10 days post-
operatively) . The ordinate indicates
the score change from the preopera-
tive value. **P<O .Ol; ***P<O.OOl
1 YEAR 5 YEARS

ABN . EEG
%
50
.5o PREOPER.
Y EARS

40

30

20

10

SHORT LONG SHORT LONG PERF. TIME

ALL (N55) NON-COMPL (N 30)

Fig.3. Percentage of abnormal EEGs after short « 2 h) and long (2: 2 h) perfusion times (n = 55).
Noncomplicated patients had no detectable CNS complications (10 days after operation) . The right
half of the figure shows that also the clinically noncomplicated patients (n = 30) showed considerable
differences in their long-term EEG outcome related to the duration of perfusion

Perfusion Time and Long- Term Outcome. Of the recognizable intraoperative factors
potentially hannful to the CNS the duration of perfusion appeared to be the most
prominent determinant of cerebral dysfunction. The patients were divided into a
short (less than 2h, n = 25) and a long (2h or more, n = 30) perfusion-time groups .
Four of the five patients with persistent clinical signs (Table 2) had long perfusion
times. As concerns EEG outcome, those with long-perfusion times experienced a
less beneficial 5-year course (Fig. 3). Significantly harmful effects of long-perfusion
time were not only seen in the CNS complicated group in which this kind of influence
would seem self-evident, but were also seen in the noncomplicated group (Fig. 3), re-
446 K. Sotaniemi

PERFORMANCE
SCORE
CHANGE
12
< 2H
.. ~
11
10
9
B
7

6 Fig. 4. Postoperative follow-up of neuro-

5 ............0, •••••••••••••••••• psychological performance score changes
4
r·····
.t>~
·..·....·..
.11 ••••••

~2H
after short « 2 h, solid line, n = 22) and
3 ", .., 0
long (~ 2 h dashed line, n = 22) perfusion
times. The difference between the two
.... ..........
,;'
groups did not reach statistical significance
until after 5 years. *P<0.05; *** P<0.OO1
1 YEAR 5 YEARS

PERFORMANCE
SCORE
CHANGE
10
9 x X AORTIC
B
..-..
.J*> MITRAL

. ............
7
6 /

.. .'
5
4
'
3
2 ...... ......$
•••••$-•••
.........
2 1 YEAR 5YEARS
MONTHS
INVESTIG.TIME
Fig. S. Follow-up of the neuropsychological performance outcome of the aortic (n = 40, solid line)
and mitral (n = 10, dotted line) patients. *P<0.05; **P<0.01; ***P<0.OO1

flecting subclinical involvement despite normal clinical findings. In QEEG, which

will not be dealt with in detail in this context, noncomplicated patients experienced
a 4% fall in mean-frequency values after short-perfusion times and a 10% fall after
long-perfusion times (10 days after the operation), the values for those with clinical
eNS signs being 10% and 24%, respectively. After 5 years those with short-perfu-
sion times had usually recovered to preoperative values, while those with long-per-
fusion times still showed a 6%-16% decline in mean frequency values. The effects
of long-perfusion times were particularly obvious in the psychometric-performance
score outcome (Fig. 4). After short-perfusion times the sum score showed a constant
rise after repeated testing, while those with long-perfusion times could not maintain
Cerebral Outcome After Open Heart Surgery 447

Table 6. Age and neuropsychological performance

Age at n Performance score

operation
(years) Preoperation Score Change ofthe
change by preoperative value
5 years (%)

15-29 All 7 50.7 8.6 + 7.7 +15.6

NC 4 49.0 11.2 +12.0 +24.5
CC 3 53.7 6.0 +1.3 + 2.4
30-44 All 17 40.5 12.6 +12.6 +29.6
NC 8 48.4 6.0 + 8.6 +17.8
CC 9 41.7 9.4 + 7.0 +16.8
45-65 All 20 33.8 11.5 + 6.8 +16.7
NC 11 35.6 7.3 +10.6 +29.8
CC 9 31.6 15.0 + 2.2 + 6.9

NC, noncomplicated; CC, complicated

their level. The perfusion-time groups first differed significantly from each other
after 5 years (Fig. 4).

Valvular Group and Postoperative Outcome. After 5 years, EEG improvement was
evident in 25% (ten out of 40) of the aortic, in 60% (six out of ten) of the mitral, and
in 40% (two out of five) of the aortic and mitral patients. The aortic group experi-
enced improvement during the first year while the mitral and aortic and mitral
groups did not show essential improvement until later. Analysis of the neuropsycho-
logical outcome revealed a corresponding difference also in the psychometric scores
(Fig.5). The mitral patients (and the multiple replacement patients) were much
slower in their score rise than the aortic patients).

Age and Neuropsychological Results. Table 6 shows the effect of age on psychometric
performance. Preoperatively, there was a negative correlation between age and sum
score. The long-term outcome seemed to be most beneficial in the age group 45-65
years; both the younger and the older group had showed a smaller rise in their sum
score by 5 years.

Discussion

A great majority (75%) of the postoperative CNS disorders were evidenced by focal
motor and/or sensory signs, in accordance with previous neurological reports [5, 6,
30,31]. Most of these deficits resolved and the remaining ones improved during the
first postoperative year [22]. Two patients had persistent motor disorders 5 years
later (Table 2), but there was no practical functional invalidity in either. Thus, the
overall course of the motor/sensory deficits seems to be very benign. The prevalence
of early postoperative CNS dysfunction in this series was 29%, which compares
448 K. Sotaniemi

favorably with the previously reported values [17]. The first postoperative year
seems to determine the neurological recovery (Table 2) and the ability to return to
work (Table 4). Corresponding long-term follow-up studies are not available for
comparisons concerning the clinical course, but the observed vocational outcome is
in agreement with previous findings [4, 11, 13, 18, 19].
In contrast to the recovery from motor and sensory deficits, the intellectual dys-
function appeared much more persistent (Table 2). After 5 years, two of the four pa-
tients who had experienced operation-induced higher cerebral function abnormalities
suffered from cognitive impairment still interfering with daily activities and the qual-
ity of life. Another patient who had displayed postoperative cognitive dysfunction
attributable to nondominant parietal involvement, but who had fully recovered dur-
ing the first year, showed renewed signs and symptoms of dysfunction after 5 years.
In this patient there was no clinical or neuroradiological evidence of cerebrovascular
accident (CVA) or other recognizable cause of renewed impairment. Thus, the oc-
currence of operation-induced involvement in the higher cerebral functions is much
more important in the postoperative long-term course of the patient than the occur-
rence of motor complications. As severe postoperative CNS complications have now
become rare [5, 17,26], intellectual impairments seem to be the major concern for
the patients' well-being.
The preoperative prevalence of CVA was 14% and stroke and TA 7% (four out
of 55, Table 3). In all of the stroke cases, the event had occurred within 18 months
prior to operation. The 5 year results make a favorable contrast to the preoperative
prevalence of CVA; there was no stroke. However, five patients reported TIA, ex-
ceeding the number of TIA patients during the last preoperative year. Precerebral
angiograms (performed in four of the five TIA patients) revealed no abnormalities.
Thus, despite continuous and controlled anticoagulant treatment the valvular pros-
thesis is the most probable source of emboli [3]. It is worth mentioning that only one
of the five TIA patients had experienced attacks before operation. Hence, as valve
replacement seems to improve the cerebrovascular prognosis, at the same time it
constitutes new threats of CVA, requiring continuous consideration.
Syncopal attacks were relatively common (11%) before but were absent after
operation (Table 3). This phenomenon reflects improvement in cerebral circulatory
conditions achieved with correction of prolonged cardiac disturbance. The disap-
pearance of syncopal attacks is of particular interest as preoperative syncope is one
of the predictors of postoperative CNS dysfunction [22, 26].
The achieved improvement in cerebral state was also clearly seen in the long-
term EEG course. Prevalence of abnormal EEG after 5 years (25%) was remarkably
lower than the preoperative value (45%). The improvement was mainly due to de-
crease of slow wave abnormalities, the dominant activity measured both in conven-
tional EEG and in QEEG remaining virtually unchanged. Improvement in EEG has
also been previously noted in short-term studies [1] but little is known of the late
EEG outcome. The beneficial EEG results are encouraging in developing operative
treatment; the negative consequences of the surgical damage seem to be outweighed
by the overall advantages. However, the long-term EEG course was closely corre-
lated with the preoperative occurrence of clinical complications, which is in agree-
ment with the short-term results [1-3]. This emphasizes the importance of careful
clinical evaluation as the basis for the assessment of CNS outcome [17, 26].
Cerebral Outcome After Open Heart Surgery 449

A close correspondence could be established also between the long-term neuro-

psychological outcome and the clinical course. Patients who had clinical signs of eNS
involvement after operation clearly had a less benign outcome than those whose
operation had been uncomplicated (Table 5, Fig. 2). At the 5-year re-evaluation, the
most prominent difference between the clinical groups was the finding that the eNS
complicated group, although having shown good recovery during the first postopera-
tive year, could not maintain the score level later on (Fig. 2). This caused a signifi-
cant difference between the performance scores of the clinical groups after 5 years
(Table 5). The late decline in the complicated group was mainly due to the poor out-
come of the older age groups (Table 6).
Evident differences in the course of subtest scores were seen between the two
clinical groups (Table 5). The subtests measuring mainly nondominant hemisphere
functions (recognition and symmetric drawing) differentiated the two groups most
clearly from each other. This is in accordance with previous information concerning
the interhemispheric differences in cardiac patients [5, 14,20].
The nondominant hemisphere seems to be somewhat more vulnerable than the
dominant hemisphere, but its capacity for recovery seems to be better than that of
the dominant hemisphere [26]. Some EEG findings [23 and 24] support this conclu-
sion based on clinical and neuropsychological data, but the background of this inter-
hemisphere disparity is not known.
The phenomenon of rising neuropsychological test scores is somewhat contro-
versial (Fig. 2, Table 5). Previous studies have completely denied the possibility that
psychometric performance could improve after correction of prolonged circulatory
and/or metabolic disturbance [9, 32] and the rising scores have been attributed to the
learning effect when the same tests are repeated. The design of the present study
does not provide evidence verifying real enhancement of cognitive functions but the
possibility of a true improvement should not be completely excluded. In addition to
the considerable decrease in various eNS symptoms and signs [28], the present and
previous EEG and QEEG findings [23 and 24] and some neuropsychological findings
[29] in the present patient series, might suggest that some real improvement of per-
formance could be achieved. If this could be ascertained, much more attention should
be paid to the optimal timing of operative correction in severe valvular disease.
The neuropsychological results show that the presence of operative clinical com-
plications also entails an increase in the risk of subclinical dysfunction. What is im-
portant is that these consequences are apparent not only immediately after operation
but also in the long-term outcome. Thus intraoperative occurrences may, even after
an uneventful operation, be reflected in the quality of life years after surgery.
Besides the neuropsychological aspects, late results of operative complications
can also be seen when the effects of extracorporeal perfusion time are considered.
The harmfulness of long perfusion time has been clearly shown in previous clinical,
EEG, and neuropsychological short follow-up reports [5, 10,24]. The long-term ef-
fects of long perfusion time on EEG (Fig. 3) and neuropsychological performance
demand that we pay increasing attention to the whole entity of intraoperative condi-
tions and standards. Perfusion time is only one measure of the operative damage but
reflects an increased risk for all the potentially harmful factors operating cumula-
tively during extracorporeal circulation [5, 6, 17]. It is particularly worth emphasiz-
ing that the consequences of long perfusion times were seen not only in the patients
450 K. Sotaniemi

with postoperative clinical manifestations of eNS involvement but also in the out-
come of those who had no evident complication. The fact that the difference be-
tween the short and long perfusion time groups even accentuated with time, without
any obvious concomitant neurological or cardiological cause, makes the question of
preventing preoperative cerebral complications more challenging in the future.

Conclusions

According to the present 5-year follow-up, postoperative neurological recovery from

cerebral consequences of prolonged cardiac disorder, from strains associated with
operation, and from occasionally induced clinical and subclinical eNS dysfunction
seems to be a much longer process than has been so far assumed. eNS outcome after
cardiac surgery cannot be adequately assessed without a long-term follow-up using a
multidimensional approach designed to consider not only clinical but also subclinical
levels of cerebral functions. Also multiparameter evaluation is essential in improving
our knowledge of assessing the risk of operation and in a better understanding of the
correct timing of operative treatment from the eNS point of view.
Intraoperative occurrences may entail consequences which are not necessarily
apparent until long after the procedure. Recognition of the many problems a cardiac
patient faces not only before but particularly after operation requires multidiscipli-
nary consideration and cooperation. Now as the major technical and surgical prob-
lems in cardiac surgery are in the process of being solved, the subclinical measures
should be the standards of development and cerebral safety.

References

1. Arfel G, Walter S (1977) Aspects EEG au stade chirurgical d'anesthesies de longue duree en
chirurgie cardiaque effets des variation thermiques. Rev EEG NeurophysioI7:45-46
2. Barash PG (1980) Cardiopulmonary bypass and postoperative neurological dysfunction. Am
Heart J 99: 675-682
3. Bjork VO, Henze A (1979) 10 years' experience with the Bjork-Shiley tilting disk valve. J Thorac
Cardiovasc Surg 78: 331-342
4. Blachly PH, Blachly PJ (1968) Vocational and emotional status of 263 patients after open-heart
surgery. Circulation 38: 524-532
5. Branthwaite MA (1972) Neurological damage related to open-heart surgery. Thorax 27:748-
753
6. Branthwaite MA (1975) Prevention of neurological damage during open-heart surgery. Thorax
30:258-261
7. Egerton N, Kay JH (1964) Psychological disturbances associated with open-heart surgery. Br J
Psychiat 110: 433-439
8. Edmiston WA, Harrison EC, Batista E, Sarma R, Kay JH, Lau FYK (1980) Clinical experience
with the Kay-Shiley mitral valve prosthesis: an eleven-year follow-up study. Scand J Thor Cardio-
vasc Surg 14:241-247
9. Frank KA, Heller SS, Kornfeld DS, Maim JR (1982) Long-term effects of open-heart surgery
on intellectual functioning. J Thorac Cardiovasc Surg 64: 811-815
10. Gilberstadt H, Sako Y (1967) Intellectual and personality changes following open-heart surgery.
Arch Gen Psychiat 16:210-214
Cerebral Outcome After Open Heart Surgery 451

11. Gilman S (1965) Cerebral disorders after open-heart operations. N Engl J Med 272: 489-498
12. Golden CJ (1976) Identification of brain disorders by the Stroop color and word test. J Clin
PsychoI32:654-658
13. Heck GA, Wright CB, Doty DB, Rossi NP (1979) Combined multi-valve procedures. A five-
year experience with 125 patients. Ann Thorac Surg 27: 320-326
14. Javid H, Tufo HM, Najafi H, Dye WS, Hunter JA, Julian OC (1969) Neurologic abnormalities
following open-heart surgery. J Thorac Cardiovasc Surg 69: 816-819
15. Kolkka R, Hilberman M (1980) Neurologic dysfunction following cardiac operation with low-
flow, low-pressure cardiopulmonary bypass. J Thorac Cardiovasc Surg 70: 432-437
16. Lee WH, Brady MP, Rowe JM (1971) Effects of extracorporeal circulation upon behaviour,
personality and brain function. Ann Surg 173: 1013-1023
17. Pedley TA, Emerson RG (1984) Neurological complications of cardiac surgery. In: Matthews
WB, Glaser GH (eds) Recent advances in clinical neurology. Churchill Livingstone, Edinburgh,
pp 159-178
18. Ross JK, Diwell AE, Marsh J (1978) Wessex cardiac surgery follow-up survey. The quality of
life after operation. Thorax 33: 3-9
19. Ross JK, Monro JL, Diwell AE (1981) The quality of life after cardiac surgery. Br Med J 282:
451-453
20. Sachdev NS, Carter CC, Swank RL, Blachly PH (1967) Relationship between post-cardiotomy
delirium, clinical neurological changes and EEG abnormality. J Thorac Cardiovasc Surg 54:
557-563
21. Smith PL, Treasure T, Newman SP, Joseph P, Ell P, Schneinaus A, Harrison MJG (1986) Cere-
bral consequences of cardiopulmonary bypass. Lancet i: 823-825
22. Sotaniemi KA (1980) Brain damage and neurological outcome after open-heart surgery. J
Neurol Neurosurg Psychiat 43: 127-135
23. Sotaniemi KA (1980) Clinical and prognostic correlates ofEEG in open-heart surgery. J Neurol
Neurosurg Psychiat 43: 941-947
24. Sotaniemi KA, Sulg I, Hokkanen TE (1980) Quantitative EEG as a measure of cerebral dys-
function before and after open-heart surgery. Electroenceph Clin Neurophysiol50: 81-95
25. Sotaniemi KA, Juolasmaa A, Hokkanen TE (1981) Neuropsychological outcome after open-
heart surgery. Arch NeuroI38:2-8
26. Sotaniemi KA (1982) Neurological complications of open-heart surgery. In: Silverstein A (ed)
Neurological complications of therapy. Selected topics. Futura, New York, pp 293-315
27. Sotaniemi KA (1983) Cerebral outcome after extracorporeal circulation. Comparison between
prospective and retrospective aspects. Arch Neurol40: 75-77
28. Sotaniemi KA (1985) Five-year neurological and EEG outcome after open-heart surgery. J
Neurol Neurosurg Psychiat 48: 569-575
29. Sotaniemi KA, Mononen H, Hokkanen TE (1986) Long-term cerebral outcome after open-
heart surgery. A five-year neuropsychological follow-up study. Stroke 17:41-46
30. Stockard JJ, Bickford RG, Myers RR, Aung MH, Dilley RB, Schauble JF (1974) Hypotension
induced changes in cerebral function during open-heart surgery. Stroke 5: 730-746
31. Tufo HM, Ostfeld AM, Shekelle R (1970) Central nervous system dysfunction following open-
heart surgery. JAm Med Assoc 212: 1333-1340
32. Aberg T, Kihlgren M (1974) Effect of open-heart surgery on intellectual function. Scand J
Thorac Cardiovasc Surg, Suppl15, pp 1-63
Chapter 37
Cerebral Protection During Open Heart Surgery:
Clinical, Psychometric, Enzymological,
and Radiological Data
T.ABERG

In 1971 [1], we made our first investigation on cerebral injury during open heart sur-
gery. We found that the incidence of obvious neurological complications was 15%.
Three percent of the patients died with or from cerebral injury. We made psycho-
metric test measurements and found that the scores were moderately to markedly
reduced, even after an uncomplicated operation (Fig. I}.
There were certain discernible risk factors, especially diagnosis (Fig. 2), time of
perfusion (Fig. 3), and degree of valvular calcification (Table I).
We interpreted these findings as suggestive of microembolization as one main
mechanism behind the intellectual impairment.
There were two sources for these emboli: the pump oxygenator and the operative
site. Late postoperative data suggested that there was a marked reversibility in the
intellectual impairment and that this reversibility occurred mainly during the first
2 months.
We drew the conclusions that:
1. Operations during extracorporeal circulation (ECC) bring about impairment in
intellectual function.

2 3 , 5 6
Scores 11 111 II 1lI II 11 Scores II 11 II 111 II 1lI
6 8

, 4

0 0

-, -,
-8 -8

-12 -12

Fig.1. Test results in comparison group (x) (lung operations), open heart surgical group without
(., n = 99) and with obvious neurological complications (e, n = 14). II, postoperative test after
1 week; III, after 2 months; 1-6, different psychometric tests (Aberg [1])
 ,
Cerebral Injury and Heart Surgery 453

1 2 3 5 6 SS
Scores 0 UI 11 III 11 /11 Scores 11 m II /11 D JI[ Scores 11 ill
8 8 20

" " 10

0 0 0

-" -4 -10

-8 -8 -20

Fig.2. Open heart surgery patients without neurological complications tested with various psycho-
metric tests. Division according to perfusion time: < 90 min (e); 9O-140min (0); > 140 min (.&).
SS, sum of the six subtests (Aberg [1])

Scores II /11

-2

Fig. 3. Open heart surgery patients without neurological compli-

cations (n = 99). Results in figure identification test divided into
different diagnoses. Con, congenital heart disease; MS, MI, mitral
stenosis, insufficiency; AS, AI, aortic stenosis, insufficiency;
Ao + Mi, double valve disease; GTS, general thoracic surgery
-B (comparison group) (Aberg [1])
454 T.Aberg

Table 1. Patients undergoing valve operations and influence of valve calcification on cerebral com-
plications

Patients Mean Mortality Cerebral Psychometric

(n) age (n) (%) complications score
(n) (%)

Patients without
valve calcifications 44 52.1 2 9.1 3 6.8 - 3.9
Patients with
light calcifications 14 53.5 4 14.3 2 14.3 5.2
Patients with
severe calcifications 43 55.5 4 9.1 10 23.9 -23.7

Total 101 53.8 10 9.9 15 14.8

Table 2. Cerebrospinal fluid adenylate kinase (UII) before and after open heart surgery and in con-
trol patients

n Mean SD Range

Before open heart surgery 20 0.017 0.005 0.015-0.030

After open heart surgery 36 0.056 0.040 0.015-0.158
Control (lung surgery) 8 0.008 0.008 o -0.015

2. This intellectual impairment probably reflects cerebral injury.

3. Most of the injuries seem reversible but,
4. Permanent injuries do occur.
Since then, we have made several investigations with the aim of reducing cerebral
injury during open heart surgery [2-6, 8, 9]. We have followed the principle of trying
to reduce or eliminate microemboli at the source.
In 1976 [2] and 1978 [3], we obtained markedly improved results. The incidence
of neurological complications decreased to 3.5% and 1.4%, and psychometric test
scores improved. During this period, the degree of calcification disappeared as a risk
factor. Perfusion time as a risk factor decreased in importance in 1976 and disap-
peared in 1978.
By this time, we had become disillusioned by the use of psychometric results only
in evaluating brain protection. Psychometric methods are cumbersome and expen-
sive and have a large variance. We started looking for other methods and included
clinical, psychometric, biochemical, and computer tomographic data in our investi-
gations. As a biochemical marker of brain cell injury we have used adenylate kinase
(AK) in cerebrospinal fluid (CSF) [4,5,6]. AK in CSF has been found to be a sensi-
tive marker of brain cell injury. It is an intracellular, mainly cytoplasmic enzyme [7].
An ordinary operation without extracorporeal circulation did not evoke release
of AK into CSF, whereas open heart operations evoked a significant increase [4]
(Table 2). There was no evidence of damage to the blood-brain barrier. The increase
in CSF AK showed a weak but statistically significant correlation to the psychometric
scores [5].
Cerebral Injury and Heart Surgery 455

Thus, there seemed to be a causal relationship between brain cell injury during
bypass and release of AK into CSF. In 22% of the patients operated on during cardio-
pulmonary bypass, a considerable increase in CSF AK was seen, in 33% a moderate
increase, and in 46% none or just a trivial increase was denoted [6].
Computed tomography (CT) of the brain was performed pre- and postopera-
tively in 75 patients. Two of these had cerebral infarctions as judged from CT, de-
spite an essentially normal postoperative state clinically.
There was no correlation between indices of brain injury and patient diagnosis or
length of perfusion [5, 6] (possibly due to restrictions of range, as the most difficult
and long-lasting operations were excluded from the study). We concluded that there
was little evidence that a continuous discharge of microembolic material from the
heart-lung machine is a major factor in the etiology of the brain injuries. For this
interpretation speaks the fact that 46% of the patients had no or only trivial degrees
of CSF AK release and that there was no correlation between psychometric scores
or CSF AK and the duration of surgery. The correlation between impaired psycho-
metric results and length of perfusion did exist during the early 1970s [1, 2]. A global
ischemic brain injury due to multiple, diffusely distributed microemboli thus seems
less likely as the major cause of the changes observed. The random way in which
changes in psychometric scores, CSF AK, and CT scans seem to be distributed, to-
gether with the lack of correlation between CSF AK and both the length of perfusion
and the changes of various postoperative biochemical parameters, are most likely
due to focal brain injury. The data fit with the interpretation that small emboli, not
giving changes large enough to be detected on the CT scan, are the prevailing cause
of the intellectual disturbance. Occasionally somewhat larger particles are respon-
sible. This concept points more to the operative field and the surgeon as the source
of small emboli than to the extracorporeal circuit. For this interpretation also speaks
the known difficulties in removing intracardiac air and debris prior to reperfusion
and weaning from bypass.
Other factors predisposing for cerebral injury are the patient's age and degree of
cerebral arteriosclerosis. In this series we have not been able to pinpoint any specific
detail in the routines around the operation as the cause of injury.
Of the various particulate emboli that may be harmful to the brain, platelet
aggregates have been suggested as important. The effect of the potent thrombocyte
inhibitor prostacyclin (PGI2) was studied in a blind, randomized, placebo-controlled
study of 103 patients [8]. Prostacyclin was given in a dose of 12.5 ng/kg/min. The
presence of a drug effect was checked by observing its hypotensive properties and
by measuring platelet count and activity. No difference was found in brain injury
parameters between the two groups. Thus, there was no evidence that thrombocyte
aggregation plays a major role in the pathogenesis of brain injury following open
heart surgery.
We also looked at clinical correlates to the brain injury parameters (Table 3). We
divided the patients into three groups, one with obvious cerebral complications, one
with subtle cerebral signs (mostly tiredness), and one with no clinical cerebral signs.
It was found that both CSF AK and psychometric results correlated well with this
grouping [6].
The proportion of patients suffering from postoperative brain injury is dependent
upon the method by which they are measured (Table 4). Thus, we found a span be-
456 T.Aberg

Table 3. Mean activity of cerebrospinal fluid adenylate kinase (CSF AK, U/I) and assessment by
psychometric tests (SS3) following open heart surgery compared with simple clinical correlates in 101
patients

n % CSFAK SS3
Obvious cerebral complications 4 3.8 0.209' -15
Subtle cerebral signs 16 15.5 0.122 -12
No cerebral signs 81 80.5 0.055 - 3.4

, These figures refer to measurements in only one patient.

Table 4. Proportion of patients with cerebral in-

jury following open heart surgery as assessed by
various methods

Clinical observation
Obvious clinical complications 3.8%
Subtle clinical complications 15.5%

Laboratory methods
Psychometric investigation 60%
CSF AK determination 55%
Computed tomography 2.7%

tween 2.7% (computed tomography), 3.8% (obvious clinical complications), 55%

(CSF AK), and 60% (psychometry) [6]. We therefore have to improve our bedside
diagnostic sensitivity to those injuries. Daily recognition and acknowledgment of this
brain damage are a prerequisite for the surgeon to improve the protection of the
brain during open heart surgery. In a recent publication, Shaw et al. [9] were able
to find a similarly high (61%) incidence of subtle signs of perioperative neurological
injury.
From these studies the following general statements can be made:
1. Open heart surgery more or less regularly brings about brain cell injury and can
therefore be used as a brain injury model.
2. Perioperative brain injury can be at least semiquantitatively measured by psycho-
metric, biochemical, and morphological means.
3. There are often subtle clinical correlates to even minor psychometric and bio-
chemical signs of brain cell injury.
4. Although the neurological complication rate has declined markedly during the
last 15 years (in our materials from 15% to 3%), no improvement has occurred
during the last 5 years as judged from our standard postoperative psychometric
scores.
5. Although the microembolism theory may have been valid 10-15 years ago, mod-
em extracorporeal circulation seems to have improved to a point where a con-
tinuous discharge of microemboli during the whole period of perfusion can be
dismissed as an important etiological factor.
Cerebral Injury and Heart Surgery 457

6. The cause of the brain cell injury is still not settled; indirect evidence points to
emboli of small to intermediate size coming from the operative field, sometimes
large enough to cause injuries visible on computed tomography, but most often
causing injuries below the resolution power of that method.
7. Postoperative brain injury is permanent in some cases. From a functional point of
view (psychometry) there is no evidence that all injuries are necessarily perma-
nent as many patients recover 2 months and 1 year postoperatively. From a bio-
chemical point of view, it is theoretically possible that temporary membrane leak-
age from a reversible injured cell is the cause of raised CSF AK. The issue of
reversibility is therefore not settled.
8. Continued efforts to reduce the incidence and extent of postoperative brain in-
jury should be made along at least two lines:
a) Removal of emboli at the source. As the operative field now seems to be the
major source, surgical awareness should be heightened and specific methods
worked out. Avoidance of multiple clamping of the ascending aorta and in-
creased use of the internal mammary artery as bypass graft are two such sug-
gestions.
b) Realizing that it is impossible to effectively remove all intracardiac and intra-
aortic debris (artherosclerosis, air, fat) at the end of bypass and that some em-
bolic material always will be expelled into carotid arteries, one principle to
persue would be to improve the ability of the brain cell to withstand an anoxic
(embolic) event until the embolus has been removed (platelets, fat, oxygen)
or adequate collateral circulation has taken over.
9. Our studies have hitherto not included patients with extremely long anticipated
bypass time. It is possible - and indeed probable - that the previous statements
are not pertinent to perfusions beyond 3 or 4 h.

References

1. Aberg T (1974) Effect of open heart surgery on intellectual function. Scand J Thorac Cardiovasc
Surg [Suppl] 15: 1-63
2. Aberg T, Kihlgren M (1977) Cerebral protection during open-heart surgery. Thorax 32:525-533
3. Aberg T, Kihlgren M, Jonsson L, Stjernlof K, LOnn U, Rystedt T, Tyden H, Westerholm CJ,
Taube A (1982) Improved cerebral protection during open-heart surgery. A psychometric investi-
gation on 339 patients. In: Becker R, Katz J, Polonius MJ, Speidel H (eds) Psychopathological
and neurological dysfunctions following open-heart surgery. Springer, Berlin Heidelberg New York,
pp 343-351
4. Aberg T, Ronquist G, Tyden H, Ahlund P, Bergstrom K (1982) Release of adenylate kinase into
cerebrospinal fluid during open-heart surgery and its relation to postoperative intellectual func-
tion. Lancet 1: 1139-1142
5. Aberg T, Ronquist G, Tyden H, Brunnkvist S, Hultman J, Bergstrom K, Lilja A (1984) Adverse
effects on the brain in cardiac operations as assessed by biochemical, psychometric and radiologic
methods. J Thorac Cardiovasc Surg 87: 99-105
6. Aberg T, Ronquist G, Tyden H, Brunnkvist S, Bergstrom K (1987) Cerebral damage during
open-heart surgery. Scand J Thorac Cardiovasc Surg 21: 159-163
7. Ronquist G, Terent A (1982) Cerebrospinal fluid markers of disturbed brain cell metabolism.
Prog Neurobiol18: 167-180
8. Aberg et al. (in press)
9. Shaw PJ, Bates D, Cartlidge NEF, Heaviside D, Julian DG, Shaw DA (1985) Early neurological
complications of coronary artery bypass surgery. Br Med J 291 : 1384-1387
SUbject Index

Acclimatization 328 cAMP 65,67,87,105

Acetazolamide 331,339 Amplitude 145
Acetylcholine 10, 64, 112, 129 Anesthesia and brain monitor 389
Acetylcholine release 64 Anesthetics 381, 392
Acid-base balance 257, 328 Anesthetic depth 376
Aconitine 129 Aneurysm 170
Acrometron 145 Angioplasty 251
Acrophase 144, 145, 264 Animal hypnosis 187
Action potential duration 30,38,39,41,55, Annihilation 108
58,83, 117 - of circus movement reentry 117
Activator calcium 4, 13 - of oscillatory activity 116
Acute ischemic LV failure Anoxia 242
see Heart failure Antiarrhythmic 42, 47, 85
Acute phase reactant 44 Antiarrhythmic agents 100
Acute respiratory failure 357 Antiarrhythmic therapy 177, 179
Adams-Stokes syndrome 205 Antibodies 318
Adenosine 62 Anticoagulation treatment 228,441,443
- effects on respiration 68 Anticonvulsants 381
Adenosine-catecholamine antagonism 66 Antidepressants 291, 381
Adenosine triphosphate (ATP) 62,67,239 Antioxidant 246
Adenosine-vagal interaction 63 Antiplatelet-aggregating agents 228
Adenylate cyclase 65 Anxiety 186
Adenylate kinase 454, 456 Aortic blood pressure 37, 40
Adrenal cortex 159 Aortic stenosis 166, 213
Adrenal medulla 21 Apnea 187
a-Adrenergic blockade 85,100 see also Sleep apnea
,B-Adrenergic blockade 38, 84, 100, 128 Arousal response 269
Adrenergic hyperactivity 99 Arrhythmias 98, 108, 120, 131, 165, 194
a-Adrenergic receptors 9,12,86 - ischaemia induced 29,32,89
,B-Adrenergic receptors 9,86 - during sleep 209,266,270
a-Adrenergic stimulation 85,103,104 Arrhythmogenesis 79,83,100,167,170,283
,B-Adrenergic stimulation 83,104 Arterial diseases 290
Adrenoceptors Arterial filtration 429
see Adrenergic receptors Arteritis 290
Age-regression equations 418 Arthritis 318
Air embolus 390 Asystole 194, 205, 270
Ajmaline 299 Ataxia 288, 315
Akinesia 231 Atenolol 38
Alcohol 283,290 Atmosphere pressure 327
Alcuronium 395 ATP 51,62,67,239,249
Algorithm 408 ATP-vagal interaction 64
Alveolar oxygen 330 Atrial arrhythmias 111, 120
Ambulatory cardiovascular monitoring 143 - vagal-mediated 65
Ambulatory ECG/EEG recordings 200, 202 Atrial fibrillation 128, 134, 138, 225
Amiodarone 100, 300 Atrial flutter 128
460 Subject Index

Atrial natriuretic factor (ANF) 207 Bypass time 457

Atrial pacing 126
Atrial tachycardias 128 Ca2+ -ATPase 249
Atrioventricular (AV) block 37,40,42,135, Ca2+ Current
205,296 see Second inward current
Atrioventricular (AV) node 70 Ca2+ Homeostasis 57
Atropine 128, 135, 189, 192 Ca2+ Ions 57
ATXII 58 Calcium 86, 88, 249
Australian Therapeutic Trial 152 Calcium antagonists 3,13,100,258
Autoimmune reaction 320 - cerebrovascular effects 3
Automaticity 83,85, 105, 112, 114 - coronary vascular effects 8
Autonomic nervous system 70, 77, 108, 120, Calcium channels 7, 8, 9
128,139 Calcium channel blockers
Autonomic neural control 62, 69 see Calcium antagonists
Autonomic reflexes 67 Calcium influx 7, 13,88,398
Autoregulation 361,378,397,423,424,433 Calcium stores 10, 58
Autosomal dominant 236, 312 Capnogram 391
AV Conduction disturbances 42,111 Cardiac arrest 185,255,294
AVNodal block Cardiac arrhythmias
see Atrioventricular (AV) block see Arrhythmias
AV Nodal conduction 37, 38, 40, 42, 122 Cardiac contraction 57
AV Nodal function curve 123, 126 Cardiac failure
AV Nodal refractory period 37, 38, 42, 126 see Heart failure
Azotemia 379 Cardiac ischemia
see Myocardial ischemia
Barbiturates 258, 383 Cardiac myocytes 55,87
Barometric pressure 330 Cardiac output 329,348
Baroreceptor 22, 80 Cardiac pacemakers
Baroreceptor reflex 80 see Pacemaker
Behavioral intervention 143 Cardiac pain 29, 80
Benzodiazepines 383 Cardiac surgery CNS outcome 440, 450
Birth asphyxia 242 - surgical problems 450
Blood-brain barrier 427 Cardiac sympathetic nerves 38, 70, 79
Blood pressure 24, 40, 142, 180 Cardiac transplant 126, 139
Blood pressure recording 148 Cardiocardiac reflex 79,99
Blood pressure screening 143 Cardiomyopathy 302, 304
Blood volume 336 Cardioplegia 251
Body mass index 156 Cardiopulmonary bypass 405,422,432,441
Bradycardia 135,165,189,267,348 - in children 436
Brain activity abnormal 419 Cardiopulmonary resuscitation 255
Brain activity monitor 393 Cardiovascular collapse 45
Brain damage 255 Cardiovascular disease risk 154
Brain injury parameters 455 Cardiovascular mortality 84
Brain monitoring 384 Cardiovascular reflexes 79, 99
Brain State Analyzer 407 Carnitine 87
Brain temperature 256 Carotid chemoreceptors
Breathholding attacks 191 see Chemoreceptors
Breathholding spells 208, 211 Carotid sinus massage 120,128
Breathing problems during sleep 266 Carotid sinus syndrome 210
Bulboreticular formation 191 Carpal tunnel syndrome 321
Bundle-branch block 138, 300 Cataplexy 208
Bupivacaine 37,38,43,45 Catecholamines 20,24, 65, 82, 84, 102
- plasma levels 45 Catecholamine release 20,24,82
- toxicity 46 Central autonomic regulation 203
Buprenorphine 32 Central nervous system (CNS) control 80
Burst-suppression pattern 382,397 Central nervous system (CNS) dysfunction
Butyrophenones 383 see Cerebral dysfunction
Subject Index 461

Central venous pressure 433 Conduction disturbances 42, 111, 179,268

Cerebellar syndrome 295 Congestive heart failure 225
Cerebral blood flow 4,205,226,338,358,377, Contractility 24, 54, 248, 368
414,423,424 Convulsions 44
- critical level 419 Convulsive syncope 2oo
Cerebral CO 2 reactivity 435 Cor pulmonale 277
Cerebral dysfunction 379,422,440,442,447, Core temperature 422
452 Coronary arteriography 174
Cerebral emboli 206, 227 Coronary artery bypass 416, 440
Cerebral function monitor 388 Coronary artery disease 166,418
Cerebral hemodynamics 406, 433, 436 Coronary artery spasm 11, 171
Cerebral hypoperfusion 180,418 Coronary circulation 102
Cerebral infarct 226 Coronary vessels 3,8
Cerebral ischemia 7,241,255,397,412,425 Cortisone 259
- indices 414 Cosine functions 147
- QEEG signs 406, 418 Cycle length alternation 127
Cerebral metabolic acidosis 426 Cytochrome 306, 315
Cerebral monitoring 433
Cerebral oxygen requirement 205 DC Extracellular electro grams 104
Cerebral perfusion 358,389,395,396,423,432 Death during sleep 269
Cerebral protection 255,452 Decompression 350
Cerebral steal 414 Decompression sickness 351,353
Cerebral vasospasm 6, 13 Decompression tables 351
Cerebral vessels 3 Deferoxamine 247
Cerebrospinal fluid 454, 456 Dehydration 332
Cerebrospinal fluid proteins 295 Delayed afterdepolarizations 89, 105
Cerebrovascular accident 448 Delayed ischemic deterioration 6
Cerebrovascular resistance 362, 423 Delta receptor 29
Chemoreceptors 64,68,79,99 Delta receptor stimulation 24
Chemoreceptor reflex 68 Dementia 315
Cheyne-Stokes respiration 328, 331 Denervated hearts 126
Chloralose anesthesia 63,99 Depolarization 56, 113
Choline acetyl transferase 64 Depolarizing clamp 56
Cholinergic blockade 65 Dextran 257
Chronic paroxysmal hemicrania 136 Diabetes 157, 258
Chronobiology 142 Diacylglycerol 88
Chronodesms 146 Digitalis 213, 291
Circadian amplitude 144 Digitalis intoxication 291
Circadian rhythms 142, 144,263 Digitoxin 292
Circulatory shock 26 Digoxin 292
Circus movement reentry 114 Diphosphoglycerate 328
Class III antiarrhythmic 100 Diphteria 288
Clots 231 Dipyridamole 68
Cluster headache 131 Dispersion of refractoriness 104, 169
CO 2 Retention 348 Diuretics 331
Coenzyme-Q 305 Dive reflex 189
Cognitive dysfunction 442, 448 Diving 343
Coma 393 - long-term effects 353
Compliance intracranial 358, 361 DNA 306
- myocardial 364 Dominant hemisphere 449
Compressed spectral array 385 Doppler technique 237
Compression 344 - transcranial 432
Compression strategies 347 Doppler ultrasound 351
Computed tomography (CT) 227,416,455 Down's syndrome 288
Computer model 124 DPI201-106 59
Computer program 150 Drop attacks 200,214
Computer-prepared profile 160 Drugs 291
462 Subject Index

Dynorphins 20 Excitable cellular membranes 47

Dyskinesia 98,231 Excitation-contraction coupling 51
Dysphagia 311 Exercise testing 174
Extracellular K+ 7,82,83
ECG Recording long-term 131,167,173,177 Extracellular Mg2+ 11
Echocardiography 230 Extracorporeal circulation 388,449,452
Edema 249 Extraventricular constraint 364
- cerebral 331,398,426
- pulmonary 331,358 False-positive diagnoses 152, 153, 177
EEG 376,442 Family history 155, 157
- amplitude ratio 392 Fast Fourier transform 380,385,407,415,441
- burst-suppression pattern 382,397 Fear 186
- computerized monitoring 385 Fear paralysis reflex 185, 194
- continuous intraoperative monitoring 416, - in animals 188
417 Fentanyl 383
- at different CBF 378, 405 FFA-lowering effect 38,42,43
- in diving 344 Fibrosis 303
- drug effects 380 Flight or fight 193
- reciprocal pattern 388, 397 Flow velocity middle cerebral artery 432
- single variables 379, 385 Flowmetry 432,436
Ejection fraction 176 Fluid balance 333
Electric shock 189 Forskolin 66
Electrical excitation 51 Framingham Study 226
Electrical stimulation of the brain 203 Frank-Starling mechanism 369
Electroconversion 233 Free fatty acids 37,42,43
Electromyogram (EMG) 265,313,388,391, Free radical 244, 259
394 Frostbite 334
Electron microscopy
see Ultrastructure GABAergic mechanisms 344
Electron transfer 250 Gas bubbles 351,353
Electrooculogram 265 Gas uptake 350
Electrophysiological study 123, 297 Giga-ohm seal 52
Electrophysiological techniques 177 Glossopharyngeal syncope 211
Emboli 225,425 Glutathione peroxidase 245
Embolic sources 230 Glycerol 241
Emotion 81,149,167,193 Glycogen 303,305
Encephalopathy 280, 316 Glycogen storage disease 288
Endocarditis 234, 321 Guillain-Barre syndrome 212, 320
Endorphins 21 Guinea pig papillary muscle 54
,B-Endorphin 21
Endothelial cell injury 319 Halothane 382
Endothelial necrosis 232 Headache 131,319
Endothelial-derived relaxant factor 7, 11 Heart failure 59,167,169,225,294
Endothelium 250 Heart rate during sleep 263,266
Endotoxin 26 Heart rate monitoring 144
End-tidal CO 2 390 Heliox dive 348
Enkephalins 20,24 Helium 343
Eosinophil cells 159 Helium tremors 344
Epidural intracranial pressure 433 Helplessness 192, 193, 194
Epilepsy 199,207 Hematocrit 333, 336
Erythropoiesis 329 Hematological changes 329
Erythropoietic response 337 Hemicrania 131
Esophageal pressure 274, 368 Hemodilution 333,337,426,436
Ether 382 Hemoglobin 256, 329
Evoked potential 412,426 Hemorrhagic hypotension 424
Evoked responses 375, 396, 413 Heparin 230
Excitability 59, 81 Hereditary diseases 289,317
Subject Index 463

Heredoataxias 288, 289 Intracellular acidosis 258

High altitude physiology 327 Intracerebral conduction 346
High-energy phosphates 241 Intracerebral steal 214, 428
High-frequency components 168 Intracranial injuries 362
High-pressure nervous syndrome 344 Intracranial pressure (ICP) 357, 391, 392
High right atrial electrogram 122, 175, 298 Intrathoracic pressure 275, 361
High risk surgery 375 Intraventricular conduction disturbance 300
His bundle electrography 37,122,175,297, Ischemia of the brain
298 see Cerebral ischemia
Histamine 10,65,66 Ischemia of the heart
Histamine release 22 see Myocardial ischemia
Hodgkin-Huxley kinetics 47 Isoflurane 382
Holter ECG recording 134, 264, 297
Homocysteine 67 Jervell and Lange-Nielsen syndrome 210
Hopelessness 192, 193, 194
Human ventricular tissue 112, 113 K+ Currents
Hydrogen 343, 349 see Potassium currents
Hydrogen desaturation technique 359 Kappa receptors 22, 32
Hydrogen peroxide 244 Kappa receptor stimulation 25
Hydroxyl radical 244 Kearns-Sayre syndrome 293,314
Hyperbaric 343
Hyperbaric index 156 Lactate 305
Hypercapnia 275, 396, 427 Late potentials 169, 174
Hyperexcitability 346 Libman-Sacks endocarditis 322
Hyperoxygenated-hypocarbic state 429 Lidocaine 43,44,47,100
Hyperpolarization 112 Ligand-binding techniques 22, 87
Hyperpolarizing current 116 Lipid peroxides 245
Hypertensive patients 424 Lipoxygenase 246
Hyperventilation 12,214,215,257,328,330, Liquid pressure 365,366
390,424 Liver cirrhosis 379
Hypobaric 327 Local anesthetics 37, 53
Hypoglycaemia 27, 258 - cardiac effects 37
Hyponatremia 6,291, 396 - cardiovascular toxicity 45
Hypoperfusion 26, 180, 418 - cerebral effects 37
Hypothalamus 191 - cerebral toxicity 44
Hypothermia 256,334,423,427,433,441 - medullary effects 47
Hypoxemia 275,277,278 - systemic effects 43
Hypoxia 327,357,387,388,396 Long Q-T syndrome 166,210
Hypoxic vasoconstriction 276 Longchain acy1carnitines 87
Hysterical fainting 215 Low-flow, low-pressure regiments 438
Low perfusion states 20
Immature brain EEG 376 Luft's disease 314
Immunological diseases 290 Lung 43
Impaired cerebral circulation 416 see also Pulmonary
Impotence 273 Luxury perfusion 378, 426
Improving survival 28, 84 LV Contractility 368
Inactivation gate 53 LV End-diastolic pressure 365
Indomethacin 136 LV Function curve 370
Infarct size 169 Lysophosphatidylcholine 84
Inorganic phosphate 239
Inosine 241 Magnetic resonance imaging
Inositol phosphates 88 see Nuclear magnetic resonance
Intellectual impairment 273, 282, 452 Mannitol 257
Interdive period 352 Maternal inheritance 316
Internal rhythm 159 Mathematical analysis 148
Interstitial fibrosis 168 Maximal oxygen uptake 330
Interstitial fluid pressure 362 Maximum rate of depolarization 54
464 Subject Index

Mechanical ventilation 357 Myocardial biopsies 303

Mechanoreceptors 79, 98 Myocardial ischemia 11,79,82,105,114,115,
MELAS 294, 314 240
Membrane channels 52 Myocardial scintigraphy 174
Membrane currents 52, 56, 59 Myocarditis 288, 321
Membrane oxygenator 422,429,436 Myoclonus 315
Membrane permeability 51 Myofibrillar degeneration 195
Membrane potential 46, 51, 56, 112 Myogenic tone 4, 9
Memory 273,335,353 Myopathy 312, 319
Meningococci 288 Myositis 320
Mental changes 335 Na+
Mental impairment
see also Sodium
see Intellectual impairment Na+ Current 54
Mental retardation 295
Na+ Current inactivation 53,59
Mental stress
Na/Ca Exchange 57,58
see Stress
Na/K-ATPase 57,249
Meperidine 381
Naloxone 25,26,259
Mepivacaine 43
- antiarrhythmic effect 30
Meptazinol 27,31 - in shock 28
- antiarrhythmic effect 30
Narcolepsy 199,208
MERRF 294, 314
Narcotic effect 349
MESOR 145, 150, 264
Neurological complications 405,443,452
Metabolic acidosis 27, 257
Neurometric maps 408
Metabolic diseases 291
Neurometric scaling 408
Metabolic substrate 43
Neuromonitoring 375,398
Metabolism 239,257,305,413
Neuromuscular block 392, 394
Methohexitone 395
Neuromuscular relaxants 391
Microelectrode techniques 52,54, 113
Neuronal death 397
Microemboli 171,418,422,425,438,452,456
Neuropsychological examination 442
Micturition syncope 212
Neuropsychological performance 445,446,447
Migraine 12, 199,319
Neuropsychological tests 444
Mild hypertension 152
Neurotransmission 68,69,346
Mitochondria 249,302,313
Neurotoxins 51
Mitochondrial function 249, 398 Newborns 155
Mitochondrial myopathies 293,314
Nifedipine 5,9, 11, 12, 100
Mitral stenosis 166, 228
Night vision 335
Mitral valve 234
Nitrogen 343,347,349
Mitral valve prolapse 236
Nitroprusside 384
M-mode criteria 236
Nonpulsatile flow 432,441
Modulated receptor hypothesis 46, 47
Noradrenaline 4, 20, 24, 81
Monge's disease 332
Norepinephrine
Monoamines 193
see Noradrenaline
Monophasic action potential 39
Noxious stimulation 193
Morphine 23, 41, 381
NREM sleep 268, 273
- electrophysiological effects 23
Nuclear magnetic resonance (NMR) 239,354
Mt. Everest expedition 335
Nucleus tractus solitarius 80, 191, 203
Mountain sickness acute 330
- chronic 332 Obesity 278
Mu receptors 22, 30 Obstructive apnea
Mu receptor stimulation 23 see Sleep apnea
Multivariable monitoring 202, 375 Occupational disease 280
Muscarinic receptors 10 Ocular muscular dystrophy 311
Muscle biopsies 294,311,312,320 Ocular myopathy 293
Muscle relaxants 384 Ocular pressure 216
Muscular dystrophies 289 Oculopharyngeal muscular dystrophy 311
Muscular weakness 321 Office hypertension 152
Myelin 313 Open heart surgery 233,379,440,452
SUbject Index 465

Ophthalmoplegia 311 Phentolamine 103

Opiate antagonists cardiovascular effects 25 Phlebotomy 332
Opioid pep tides 20 Phosphatidylinositol 88
- in myocardial ischaemia 29 Phosphoinositide metabolism 89
- in shock 26, 28 Phospholipids 53
- in stress 29 Phosphorus metabolism 242
Opioid receptors 41 Physiological tolerance 375, 396
- classification 22 Pineal gland 159
Opioid receptor agonists cardiovascular effects Plasma volume 336
23;24 Platelet 171,455,457
Organic solvents 280 Play opossum 188
Orienting reaction 190 Pleural surface pressure 366
Orthostatic tests 216 Poisoning 195
Oxygen 244,256,267,327,349 Poliomyelitis 289
Oxygen consumption 329 Polycythemia 332
- cerebral 377 Polysomnography 273, 280
- myocardial 42 Population rhythm 156
Oxygen radicals 244,246,248 Positive end-expiratory pressure (PEEP) 357,
Oxygen saturation 266,269,281,427 362,364,368
Oximetry 267,274,281 Postgraduate education 292
Postoperative brain injury 447,448,454,457
Pacemaker 70, 206, 294, 306 Postprandial hypotension 212
Pacemaker activity 114 Postural hypotension 212
Pacemaker current 113 Potassium currents 54, 86, 113
Painters 281 Prazosin 12,85, 100, 101
Paracrystalline inclusions 316 Pregnancy 45
Parasympathetic Preload 364, 370
see also Vagal Premature atrial beats 123
Parasympathetic nerves 117, 203 Premature ventricular beats 134, 170
Parasympathetic overactivity 140 Pressure-related health index 154
Parasympathetic system 268 Presynaptic receptors 21
Parkinson's disease 212 Programmed electrical stimulation 175
Paroxysmal AF 227 Progressive external ophthalmoplegia 293,311
Paroxysmal hemicrania 136 Prolactin 215
Partial pressure 349 Propranolol 67, 100, 103, 128
Patch clamp technique 52 Prostacyclin 455
Pentobarbital anesthesia 37,63 Prostaglandin 5, 10,70,258
Pentoxifylline 334 Prosthetic valves 234
Perfused brain 241 Protein binding 44
Perfused heart 240,248 Pseudoseizures 214
Perfusion pressure 389,390 Pseudosyncope 215
Perfusion time 445,449 Psychiatric patients 377
Pericardial catheter 364 Psychometric tests 405, 444, 452, 454
Pericardial pressure 365,366,370 Psychopharmacology 380
Pericardial surface pressure 369 Psychotic 347
Pericarditis 288, 321 Ptosis 296,312
Pericardium 365 Pulmonary arterial hypertension 277
Period 145 Pulmonary arterial pressure 275,302
Peripheral blood circulation 338 Pulmonary hemodynamics 272, 275, 277
Peripheral nerve conduction 313,346 Pulmonary hypertension 166
Peripheral neuropathies 319 Pulmonary J receptors 22
Peripheral vessels 5 Pump function 412
Pertussis toxin 66 Purine metabolism 69
Petit mal seizure 207 Purinergic nerves 69
Phase-resetting 112, 114 Purkinje cells 115
Phase-response curve 109 Purkinje muscle junction 115
Phenothiazines 291,380 Pyruvate 305
466 Subject Index

Q10 Rule 413 Silent bubbles 351

QEEG map 409,410 Single-channel recording 57
Quality of life 159,449 Sinoatrial block 135,138
Quantitative EEG (QEEG) 378,386,406,440 Sinoatrial disease (SA) 179,206
Quinidine 213 Sinus arrests 268
QRSAxis 297 Sinus pacemaker activity 63,67,70
QRS Duration 168 Sleep 25, 149,263,272,281,376
QT Interval 210,264,269,270 Sleep apnea 267, 273, 280
Sleep apnea syndrome 209,266,272,280
Ragged-red fibers 293,294,302,311 Slow inward current
Rapid-eye movement (REM) 263 see Second inward current
Rashes 318 Smoke reflex 189
Rat isolated atria 23 Snoring 272
Rate constants 47 Sobbing spasm 211
Raynaud's phenomenon 12 Sodium bicarbonate 258
Receptor analysis 87 Sodium channels 46, 47, 53, 59
Red-cell volume 336 Sodium load 58, 59
Reductase 315 Sphygmochron 149,158
Reentry 41,105,114,120,170 Status epilepticus 209
Refractory period 39,83,85,117,121,169 Stellate ganglion 80
Refsum's disease 166 Stellate stimulation 99,109, 110, 111, 121
Regional CBF 377, 429 Stellectomy 103, 104
Regional cerebral blood flow (CBF) 377, 405, Stenosed neck arteries 201
409,429 Stereoisomers 31
Regional ischemia 240,412 Steroid anesthetics 383
Reperfusion 248, 255 Stimulus-contraction coupling 4, 9
Reperfusion damage 248 Stokes-Adams attacks 294
Respiration 209 Stress 29,81,174, 186, 190, 195
Respiratory acidosis 257 Stroke 204,225,227,228,230,414,442
Respiratory alkalosis 328 Subarachnoid hemorrhage 6, 204
Respiratory disturbance index 267 Subclavian steal syndrome 214
Respiratory sinus arrhythmia 216 Subclinical dysfunction 449
Retinal bleeding 331,335 Succinylcholine 384
Retinal pigmentation 293,296 Sudden cardiac death 79,98,165,185,259
Retrograde conduction 115,117 - risk of 173
Rheumatic fever 288 Sudden death 194
Rhythmometry 157 Sudden infant death 187, 195
Right atrial pressure 369 Sudden infant death syndrome 186, 188
Right heart failure 277 Superoxide dismutase (SOD) 245
Romano-Ward syndrome 210 Superspecialization 292
Rubella 289 Supraventricular tachycardia 120, 178
Surface pressure 366
Saline 333 Sustained tachycardia 121
Sarcolemmal dammage 249 Sustained ventricular tachycardia 176
Scavengers 246, 259 Swallowing 211
Screening 143 Sympathetic afferent fibers 39
Sea anemone toxin 55 Sympathetic denervation 81
Second inward current 55,84,86 Sympathetic fibers 203
SeUarre 199,207,242 Sympathetic nervous system 79,98
Selenium 245,246 Sympathetic-parasympathetic interaction 193,
Sensory impairment 335 270
Septicaemia 27 Sympathetic reflex 121
Shock 26, 28, 195 Sympathetic stimulation 103,109, 111
Shock avoidance 190 Sympathetic tone 25,42,108, 123, 190
Sick sinus syndrome 211 Sympathetic transmitter release 20,24,69,82
Sigma receptor agonists 26 Sympatho-medullary system 28
Signal-averaged recording 168 Synaptic transmission 68,69,346
Subject Index 467

Syncope 165,172,181,199,202,448 Vacuoles 302,312

- after myocardial infarction 178 Vagal
- cardiovascular mechanisms 177 see also Parasympathetic
- differential diagnoses 199,202 Vagal activity 39,41, 194
Syncope team 217 Vagal maneuvers 128
Syphilis 287 Vagal stimulation 108, 109, 111, 185
Systemic lupus erythematosus 318 Vagal tone 42,63,120, 123, 128
Vagotomy 41
Tachycardia 124, 125, 165 Vagus index 216
Temperature-corrected CO2 438 Valsalva-like maneuver 191
Temperature variations 334 Valsalva maneuver 216,361
Temporal lobe epilepsy 209 Valve calcification 234
Tetrodotoxin 53 Valve disease 226
Theophylline 67 Valve replacement 380, 419, 440
Thoracic epidural anesthesia cardiac function Valvular replacement
37 see Valve replacement
Thrombi 225,233 Variant angina 12
Thrombocytes 353 Vasculopathy 318
Thrombo-embolic disease 230 Vasoactive agents 384
Thromboembolism 227 Vasoconstriction 12,102
- prevention and treatment 228 Vasodilatory action 70
Thrombogenesis 226 Vasospasm 3,6, 11, 171
Thrombosis 171, 334 Vasovagal syncope 210
Thromboxane 5, 13 Venoarterial reflexes 362
Tobacco smoke 192 Ventricular arrhythmias 41,89,99, 178,300
Toluene 350 - effects of opioid receptor antagonists 29
Tomography 338 Ventricular dysfunction
Topographical mapping 408 see Heart failure
TO Segments 104 Ventricular effective refractory period 39
Tracheostomy 269,277,278 Ventricular fibrillation 172, 185
Training 335 Ventricular hypertrophy 303
Transducer positions 231 Ventricular output 276
Transesophageal echocardiography 234 Ventricular premature beats 134, 170
Transient inward current 89 Ventricular tachyarrhythmias 165
Transient ischemic attacks (TIA) 167, 200, Ventricular tachycardia 83, 172
414,442,448 Verapamil 4,9,12,100,109,125,240
Transitory loss of consciousness drug-induced Vigilance 274
213 Viral infection 306
Transmural infarctions 232 Vitamin C 246
Transmural pressure 275,365 Vitamin E 246
Trauma 26 Voltage clamp technique 52,56
Trichlorethane 282 Voltage dependent 46
Triggered activity 89,105 Voodoo death 192
Triggering factors 170
Trombolysis 251
Tuberculosis 287 Wedge pressure 276, 302, 370
Tussive syncope 213 Wenckebach-type block 299
Twin studies 376 WPW Syndrome 180

Ultrastructure 248, 250, 302, 304, 313, 315 Xanthine oxidase 245,250
Unconsciousness cardiogenic 204 Xenon radio-activity 338
- neurogenic 207
- psychogenic 200
Unstable angina 171 Zero-line-crossing techniques 385
Urine output 333 Z-Transformation 407
 A. J. Furlan, Cleveland (Ed.)

The Heart and

Stroke
Exploring Mutual Cerebrovascular and
Cardiovascular Issues
1987.52 figures. XXII, 382 pages. (Clinical
Medicine and the Nervous System).
ISBN 3-540-16206-2

Contents: Introduction. - Cardioembolic Stroke. -

Cardiovascular Neurobiology. - Global Brain
Ischemia. - Stroke and Open Heart Surgery. -
Epidemiology and Risk Factors. - Subject Index.
Heart and brain interaction is an increasingly vital
area of clinical investigation. This is the most
comprehensive review of the subject available,
presented by internationally recognized authorities
in the field.
The book offers extensive coverage of cardioem-
bolic stroke, including a brand new contribution on
the mechanism of hemorrhagic infarction. Contro-
versial topics such as anticoagulation, combined
carotid and coronary surgery and screening for
silent coronary disease are covered. Also included
are a comprehensive review of the cardiovascular!
neurobiological role of the central nervous system
in hypertension and sudden death, and a practical
approach to the patient with syncope.
This integrated, topical presentation makes essen-
Springer-Verlag Berlin tial reading for neurologists, cardiologists, intemists
Heidelberg New York London and anyone caring for patients with stroke or
Paris Tokyo Hong Kong cardiac disease.
 D. Ganten, Heidelberg; D. Pfaff, New York (Eds.)

Central
Cardiovascular
Control
Basic and Clinical Aspects
1983.71 figures. V, 192 pages. (Current Topics in
Neuroendocrinology, Volume 3). ISBN 3-540-11350-9
Contents: Functional and Anatomic Aspects of Central
Nervous Cardiovascular Regulation. - Autonomic
Nervous System and Blood Pressure Control in
Normotensive and Hypertensive Conditions. - Reflex
Control of Circulation in Normotensive and Hyperten-
sive Humans. - Corticotropin-Releasing Factor: Central
Nervous System Effects on the Sympathetic Nervous
System and Cardiovascular Regulation. - Neuropepti-
des and Central Blood Pressure Regulation. - Centrally
Acting Drugs as a Tool to Study Central Mechanisms
of Blood Pressure Control. - The Blood-Brain Barrier
and its Role in the Control of Circulating Hormone
Effects on the Brain. Subject Index.
Central Cardiovascular Control: Basic and Clinical
Aspects is an up-to-date review of the presently impor-
tant issues in central cardiovascular regulation and its
implications and consequences for the pathophysiol-
ogy, pharmacology, diagnosis and treatment of hyper-
tension.
The book includes a discussion of the anatomical and
physiological organization of cardiovascular control
centers in the brain, the role of cardiovascular reflexes
for short- and long-term blood pressure control, and
mechanisms and importance of selective autonomic
innervation of peripheral end-organs of blood pressure
control. New developments concerning the role of
neurotransmitters, especially biogenic amines, neuro-
Springer-Verlag Berlin peptides and antihypertensive drugs, are discussed as is
Heidelberg New York London the possibility of circulating hormones affecting brain
Paris Tokyo HongKong mechanisms through the blood-brain barrier.