Sei sulla pagina 1di 1187

Luca Saba

Eytan Raz 

From Basics to
Advanced Concepts

1 3Reference
Neurovascular Imaging
Luca Saba • Eytan Raz

Neurovascular Imaging
From Basics to Advanced Concepts

With 799 Figures and 102 Tables

Luca Saba Eytan Raz
Department of Radiology Neurointerventional Radiology Section
AOU of Cagliari Department of Radiology
Monserrato, Cagliari NYU Langone Medical Center
Italy New York, NY

ISBN 978-1-4614-9028-9 ISBN 978-1-4614-9029-6 (eBook)

ISBN 978-1-4614-9030-2 (print and electronic bundle)
DOI 10.1007/978-1-4614-9029-6
Library of Congress Control Number: 2015956531

# Springer Science+Business Media New York 2016

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or
part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way,
and transmission or information storage and retrieval, electronic adaptation, computer software, or
by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt
from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained
herein or for any errors or omissions that may have been made.

Printed on acid-free paper

This Springer imprint is published by SpringerNature

The registered company is Springer Science+Business Media LLC New York
Luca: This book is dedicated to Tiziana
Eytan: To the women of my life, my mother Daniela and
my wife Lia

Vascular neurology, vascular neurosurgery, and interventional neuroradiology

are independent fields with dedicated training programs. Neuroimaging, and in
particular what we call “neurovascular imaging,” is a unifying factor which
can be considered the intersection of these three medical specialties. With this
book, we aim to cover thoroughly the imaging techniques, potentialities, and
present and future applications as applied to the vascular diseases of the central
nervous system from the imaging point of view.
In recent years, several advances in imaging technology have dramatically
expanded the frontiers of human investigation in the neurovascular field. We
hope that this book is a timely and complete addition to the growing body of
literature on this topic.
Our aim is to cover all aspects of neurovascular disease in a comprehensive
way. The illustrations and high quality of the images will be the key to success.
This is the first collection to be published dedicated to the field of vascular

Cagliari, Italy Luca Saba

New York, USA Eytan Raz
December 2015


Do not go where the path may lead, go instead where there is no path and
leave a trail.
Ralph Waldo Emerson (1803–1882)

The foundation of every state is the education of its youth.

Diogenes (412 BC–323 BC)

Life is like riding a bicycle. To keep your balance you must keep moving.
Albert Einstein (1879–1955)

Live as if you were to die tomorrow; learn as if you were to live forever.
Mahatma Gandhi (1869–1948)


It is not possible to overstate my gratitude to the many individuals who helped

to produce this book; their enthusiasm and dedication was unbelievable. In
particular, I would like to thank Professor Mario Piga for his help and advice. I
am deeply appreciative of the support offered by my colleagues at the Azienda
Ospedaliero-Universitaria di Cagliari. My gratitude goes to all authors who
have contributed to this project. Without their incredible work, this would not
have been possible.
I would like to express my appreciation to Springer for their professional-
ism in handling this project and for the patience. Most particularly, I would like
to thank Sandra Fabiani, Sunaina Dadhwal, Neha Thapa, and Gregory Baer for
their tireless dedication and advice during each stage in the production of this
book. Your help was wonderful and made producing this project an enjoyable
and worthwhile experience.
Finally, again, I would like to acknowledge Tiziana! Thank you, for

Cagliari, Italy Luca Saba

First and foremost, I want to thank Luca Saba for inviting me to help him in this
project and for his efforts in making this project going along the way. Luca,
you are one of the most hardworking persons I have ever met!
I want to thank Springer for the patience and enthusiasm that they applied to
this project, especially Neha Thapa and Sunaina Dadwhal. Thank you very
I also wanted to thank all the teachers I had in my medical career so far,
especially those teachers that did that without clamor, those that taught ethics
before medicine, and those that taught me a method, a way of living, and
secondarily a way of working.

New York, USA Eytan Raz


Volume 1

Part I The Basics: Embryology, Pathological Basis of

Vascular Diseases, and Imaging Techniques . . . . . . . . . . . . . . . . . . 1

1 Cerebrovascular Development and Evolution . . . . . . . . . . . . 3

Maksim Shapiro and Eytan Raz
2 Atherosclerosis: Biological and Pathological Basis . . . . . . . . 33
Michele Porcu, Eytan Raz, and Luca Saba
3 Nonatherosclerotic Vascular Disease: Biological and
Pathological Basis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Pierleone Lucatelli, Beatrice Sacconi, and Carlo Catalano
4 Essentials of Transcranial Doppler Ultrasound . . . . . . . . . . . 47
Jonathan D. Kirsch
5 DSA Imaging Protocol and Technique . . . . . . . . . . . . . . . . . . 67
Eytan Raz

Part II Carotid and Vertebral Artery: Anatomy, Diseases,

and Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

6 Anatomy of the Neck Arteries . . . . . . . . . . . . . . . . . . . . . . . . . 87

Nasim Sheikh-Bahaei, Tomasz Matys, and Jonathan H. Gillard
7 Imaging of Carotid Atherosclerosis . . . . . . . . . . . . . . . . . . . . 95
Jie Sun, Niranjan Balu, and Chun Yuan
8 Carotid Artery Dissection . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Olivier Naggara, Myriam Edjlali-Goujon, Eric Bodiguel,
Marie Pierre Gobin-Metteil, Denis Trystram, Christine
Rodriguez-Regent, Jean-Louis Mas, Jean Francois Meder, and
Catherine Oppenheim
9 Nonatherosclerotic Nondissection Diseases of
Carotid Artery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
Santosh Kumar Kannath and T.R. Kapilamoorthy

xiv Contents

10 Imaging of the Pathology of the Vertebral Arteries . . . . . . . . 163

David Chiao and Max Wintermark

11 Carotid Artery Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191

Roberto Montisci and Luca Saba

Part III The Basics of Intracranial Arterial Circulation . . . . . . . . 203

12 Intracranial Atherosclerosis . . . . . . . . . . . . . . . . . . . . . . . . . . 205

Xinyi Leng and David S. Liebeskind

13 Imaging of Spontaneous Nontraumatic Intracerebral

Hemorrhage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
Stephen Quinet and Patrick Turski

14 ICA–ECA Collaterals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293

Maxim Mokin and Adnan H. Siddiqui

15 Imaging Biomarkers of Angiogenesis and the

Microvascular Environment in Cerebral Tumors . . . . . . . . . 303
Alan Jackson, Ibrahim Djoukhadar, and David J. Coope

Part IV Stroke Imaging .................................... 327

16 Patterns of Ischemic Stroke: From Lacunar to Territorial

to Multiple Embolic to Watershed Hypotensive . . . . . . . . . . 329
Jessica Rotman and Robert Zimmerman

17 Territorial Strokes as a Tool to Learn Vascular

Territories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363
Behroze Adi Vachha and Pamela Whitney Schaefer

18 Hemorrhagic Stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383

Julius Griauzde, Elliot Dickerson, and Joseph J. Gemmete

19 Medical Therapy of Acute Stroke . . . . . . . . . . . . . . . . . . . . . . 413

Keith G. DeSousa and Albert S. Favate

20 Endovascular Treatment of Stroke . . . . . . . . . . . . . . . . . . . . . 425

Paolo Machi and Paolo Garofalo

21 Sickle Cell Disease and Stroke . . . . . . . . . . . . . . . . . . . . . . . . 439

Akifumi Fujita, Chie Asai, Yu-Ming Chang, Nadja Kadom,
Martin H. Steinberg, Naoko Saito, and Osamu Sakai

22 Hypoxic–Ischemic Encephalopathy (Preterm, Term,

and Adult) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469
Mauricio Castillo and Francisco Chiang
Contents xv

Part V Aneurysms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495

23 Aneurysmal Subarachnoid Hemorrhage . . . . . . . . . . . . . . . . 497

Frédéric Clarençon, Nader-Antoine Sourour, Vincent Degos,
Aurélien Nouet, Federico Di Maria, Eimad Shotar,
Joseph Gabrieli, Lise Le Jean, and Jacques Chiras
24 Blister Aneurysms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521
James V. Byrne and Svein Harald Mørkve
25 Surgical Treatment of Aneurysms . . . . . . . . . . . . . . . . . . . . . 535
Saul F. Morales-Valero, Shanna Fang, and Giuseppe Lanzino
26 Endovascular Treatment of Brain Aneurysms . . . . . . . . . . . 551
Luca Quilici and Edoardo Boccardi
27 Postop Evaluation of Aneurysms (Including Vasospasm) . . . 583
Michael Mayich, Brian P. Walcott, Christopher
J. Stapleton, and Daniel Thomas Ginat

Volume 2

Part VI Other CNS Vascular Diseases . . . . . . . . . . . . . . . . . . . . . . . . 603

28 Brain Arteriovenous Malformations . . . . . . . . . . . . . . . . . . . 605

Daniel Sahlein and Nathan Manning
29 Dural AVF and Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . 641
Pierre Guedin
30 Carotid Cavernous Fistula . . . . . . . . . . . . . . . . . . . . . . . . . . . 657
Timothy R. Miller, Ravishankar Shivashankar,
Gaurav Jindal, and Dheeraj Gandhi
31 Imaging of the Neurovasculitides . . . . . . . . . . . . . . . . . . . . . . 683
Adam J. Davis
32 Miscellaneous Vascular Malformations (Cavernous
Malformations, Developmental Venous Anomaly,
Capillary Telangiectasia, Sinus Pericranii) . . . . . . . . . . . . . . 719
Luiz Celso Hygino da Cruz and Cintia Elias Pires
33 Miscellaneous Vascular Diseases . . . . . . . . . . . . . . . . . . . . . . . 751
Matylda H. Machnowska and Richard I. Aviv
34 Vascular Loop Syndromes . . . . . . . . . . . . . . . . . . . . . . . . . . . 781
Divyata R. Hingwala and Kesavadas Chandrasekharan
35 Migraine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 791
Mark Kruit
xvi Contents

36 Moyamoya Disease (Spontaneous Occlusion

of the Circle of Willis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 817
Akira Yamamoto, Tomohisa Okada, and
Jun C. Takahashi
37 Brain Death Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865
M. Sawicki, Joanna Wojczal, Bozena Birkenfeld, and
Lech Cyrylowski

Part VII Veins Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 897

38 Jugular Vein Thrombosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 899

Nicholas A. Koontz, Richard H. Wiggins III, and
Lubdha M. Shah
39 Imaging of Cerebral Venous and Sinus Thrombosis . . . . . . . 935
Jennifer Linn
40 The Role of Diagnostic Imaging Techniques for Detection
of Extracranial Venous System Abnormalities Associated
with Central Nervous System Disorders . . . . . . . . . . . . . . . . 953
Kresimir Dolic and Robert Zivadinov

Part VIII Pediatrics ........................................ 981

41 Pediatric Vascular Malformations . . . . . . . . . . . . . . . . . . . . . 983

Sachin K. Pandey and Darren B. Orbach
42 Pediatric Stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1009
Daniel Tibussek, Gabrielle deVeber, and Manohar Shroff
43 Congenital Vascular Syndromes and Diseases . . . . . . . . . . . . 1041
Sarah Milla, Jennifer Vaughn, and Nilesh K. Desai

Part IX Spine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1061

44 Spinal Vascular Imaging: Technique . . . . . . . . . . . . . . . . . . . 1063

Maria Isabel Vargas, Fabrice Bing, Joanna Gariani, and
Jean-Louis Dietemann
45 Spinal Vascular Anatomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1095
Tibor Becske and Eytan Raz
46 Spinal Vascular Malformations and Treatment . . . . . . . . . . . 1105
Srinivasan Paramasivam
47 Spinal Cord Infarction and Differential Diagnosis . . . . . . . . 1125
Srikanth R. Boddu, Alessandro Cianfoni, Kyung-Wha Kim,
Mohammad Amin Banihashemi, Emanuele Pravatà,
Y. Pierre Gobin, and Athos Patsalides
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1185
About the Editors

Prof. Luca Saba received his M.D. from the University of Cagliari, Italy, in
2002. Today, he works in the A.O.U. of Cagliari. Professor Saba’s research
fields are focused on multi-detector-row computed tomography, magnetic
resonance, ultrasound, neuroradiology, and diagnostic in vascular sciences.
His works, as lead author, are more than 170 publications in high-impact-
factor, peer-reviewed journals such as American Journal of Neuroradiology,
Atherosclerosis, European Radiology, European Journal of Radiology, Acta
Radiologica, Cardiovascular and Interventional Radiology, Journal of
Computer Assisted Tomography, American Journal of Roentgenology, Neuro-
radiology, Clinical Radiology, Journal of Cardiovascular Surgery, Cerebro-
vascular Diseases, Brain Pathology, Medical Physics, and Atherosclerosis.
He is a well-known speaker and has spoken over 45 times at national and
international level conferences.
Dr. Saba has won 15 scientific and extracurricular awards during his career.
Dr. Saba has presented more than 500 papers and posters in national and
international congresses (RSNA, ESGAR, ECR, ISR, AOCR, AINR, JRS,
SIRM, AINR). He has written 21 book chapters and is Editor of 10 books in
the field of computed tomography, cardiovascular, plastic surgery, gynecolog-
ical imaging, and neurodegenerative imaging.
He is a member of the Italian Society of Radiology (SIRM), European
Society of Radiology (ESR), Radiological Society of North America (RSNA),
American Roentgen Ray Society (ARRS), and European Society of Neurora-
diology (ESNR) and serves as reviewer of 40 scientific journals.

xviii About the Editors

Eytan Raz currently works at NYU Langone Medical Center in New York
City, where he is an Assistant Professor working for the Neuroradiology and
Interventional Neuroradiology Sections. He graduated from Sapienza
University of Rome School of Medicine in 2006, where he also obtained his
Radiology Specialty in 2011. He completed his fellowships in Diagnostic
Neuroradiology and Interventional Neuroradiology at NYU Langone
Medical Center. He is the author of more than 60 publications in
high-impact-factor, peer-reviewed journals such as Radiology, AJNR, AJR,
JNIS, Neurology, and Brain Pathology. His fields of expertise include
neurovascular diseases, traumatic brain injury, and multiple sclerosis. In
2012, he was appointed Editor for the Fellows’ Portal of AJNR. He is a
reviewer for more than 10 journals and has presented more than 70 abstracts
at international conferences. He is a member of the Radiological Society of
North America (RSNA), the American Society of Neuroradiology (ASNR),
and the Italian Society of Neuroradiology (AINR).

Chie Asai Departments of Radiology, Boston University, Boston Medical

Center, Boston, MA, USA

Richard I. Aviv Division of Neuroradiology, Department of Medical Imag-

ing, School of Medicine, Sunnybrook Health Sciences Centre, University of
Toronto, Toronto, ON, Canada

Niranjan Balu Department of Radiology, University of Washington, Seattle,


Mohammad Amin Banihashemi Department of Neurological Surgery,

Weill Cornell Medical Center, New York, NY, USA

Tibor Becske Neurointerventional Radiology Section, Department of

Radiology, NYU Langone Medical Center, New York, NY, USA

Fabrice Bing Department of Radiology, University Hospital of Strasbourg,

Strasbourg, France

Bozena Birkenfeld Department of Nuclear Medicine, Pomeranian Medical

University, Szczecin, Poland

Edoardo Boccardi Ospedale Niguarda – Ca’ Granda, Milan, Italy

Srikanth R. Boddu Department of Neurological Surgery, Weill Cornell

Medical Center, New York, NY, USA

Eric Bodiguel Service de Neurologie, Faculté de Médecine, Centre

Hospitalier Sainte-Anne, Université Paris Descartes, Sorbonne Paris Cité,
Paris, France

James V. Byrne Nuffield Department of Surgical Sciences, Oxford Univer-

sity, John Radcliffe Hospital, Oxford, UK

Mauricio Castillo Division of Neuroradiology, Department of Radiology,

University of North Carolina, Chapel Hill, NC, USA

Carlo Catalano Department of Radiological, Oncological and Anatomo-

Pathological Sciences, Sapienza University of Rome, Rome, Italy

xx Contributors

Kesavadas Chandrasekharan NM Medical Centre, Mumbai, Maharashtra,

Sree Chitra Tirunal Institute for Medical Sciences and Technology, Trivan-
drum, Kerela, India
Yu-Ming Chang Departments of Radiology, Boston University, Boston
Medical Center, Boston, MA, USA
Francisco Chiang Division of Neuroradiology, Department of Radiology,
University of North Carolina, Chapel Hill, NC, USA
David Chiao Department of Radiology, School of Medicine, University of
Virginia, Charlottesville, VA, USA
Jacques Chiras Department of Interventional Neuroradiology,
Pitié-Salpêtrière Hospital, Paris VI University, Paris, France
Alessandro Cianfoni Department of Radiology, Medical University of South
Carolina, Charleston, SC, USA
Frédéric Clarençon Department of Interventional Neuroradiology,
Pitié-Salpêtrière Hospital, Paris VI University, Paris, France
David J. Coope Wolfson Molecular Imaging Centre, University of Manchester,
Withington, Manchester, UK
Lech Cyrylowski CT and MRI Clinic, “Nasz Doktor” Health Care Institu-
tion, Szczecin, Poland
MRI Clinic, “Euromedic” Health Care Institution, Szczecin, Poland
Luiz Celso Hygino da Cruz Department of Radiology, Clinics CDPI/DASA
and IRM, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil
Adam J. Davis Neuroradiology, Department of Radiology, NYU Langone
Medical Center, New York, NY, USA
Vincent Degos Department of Anesthesia and Critical Care, Neuro-intensive
Care Unit, Pitié- Salpêtrière Hospital, Paris VI University, Paris, France
Nilesh K. Desai Department of Radiology, Texas Children’s Hospital, Baylor
College of Medicine, Houston, TX, USA
Keith G. DeSousa Department of Neurology, NYU Langone Medical Cen-
ter, New York, NY, USA
Vascular Neurology, University of Miami, Miami, FL, USA
Gabrielle deVeber Children’s Stroke Program, Division of Neurology,
The Hospital for Sick Children, University of Toronto, Toronto, ON, Canada
Federico Di Maria Department of Interventional Neuroradiology, Pitié-
Salpêtrière Hospital, Paris VI University, Paris, France
Elliot Dickerson Department of Radiology, Radiology Resident, University
of Michigan, Ann Arbor, MI, USA
Contributors xxi

Jean-Louis Dietemann Department of Radiology, University Hospital of

Strasbourg, Strasbourg, France
Ibrahim Djoukhadar Wolfson Molecular Imaging Centre, University of
Manchester, Withington, Manchester, UK
Kresimir Dolic Buffalo Neuroimaging Analysis Center, Department of
Neurology, University at Buffalo, State University of New York, Buffalo,
Clinical Department of Diagnostic and Interventional Radiology, Clinical
Hospital Center Split, University of Split, Split, Croatia
Myriam Edjlali-Goujon Service de Neuroradiologie, Centre Hospitalier
Sainte-Anne, INSERM UMR894, Université Paris Descartes, Sorbonne
Paris Cité, Paris, France
Shanna Fang Mayo Medical School, Mayo Clinic, Rochester, MN, USA
Albert S. Favate NYU Langone Medical Center, New York, NY, USA
Akifumi Fujita Departments of Radiology, Boston University, Boston
Medical Center, Boston, MA, USA
Joseph Gabrieli Department of Interventional Neuroradiology,
Pitié-Salpêtrière Hospital, Paris VI University, Paris, France
Dheeraj Gandhi Department of Diagnostic Radiology, Neuroradiology,
University of Maryland Medical Center, Baltimore, MD, USA
Joanna Gariani Department of Radiology, Geneva University Hospitals,
Geneva, Switzerland
Paolo Garofalo Medicine School, University of Parma, Parma, Italy
Joseph J. Gemmete Department of Radiology, Neurosurgery, and Otolaryn-
gology, University of Michigan, Ann Arbor, MI, USA
Jonathan H. Gillard Department of Radiology, School of Clinical Medicine,
University of Cambridge, Cambridge Biomedical Campus, Cambridge, UK
Daniel Thomas Ginat Department of Radiology, University of Chicago,
Chicago, IL, USA
Y. Pierre Gobin Department of Neurological Surgery, Weill Cornell Medical
Center, New York, NY, USA
Marie Pierre Gobin-Metteil Service de Neuroradiologie, Centre Hospitalier
Sainte-Anne, INSERM UMR894, Université Paris Descartes, Sorbonne Paris
Cité, Paris, France
Julius Griauzde Department of Radiology, Radiology Resident, University
of Michigan, Ann Arbor, MI, USA
Pierre Guedin Department of Diagnostic and Therapeutic Neuroradiology,
Hôpital Foch, Suresnes, France
Divyata R. Hingwala NM Medical Centre, Mumbai, Maharashtra, India
xxii Contributors

Alan Jackson Wolfson Molecular Imaging Centre, University of Manchester,

Withington, Manchester, UK

Gaurav Jindal Department of Diagnostic Radiology, Neuroradiology,

University of Maryland Medical Center, Baltimore, MD, USA

Nadja Kadom Departments of Radiology, Boston University, Boston Med-

ical Center, Boston, MA, USA

Santosh Kumar Kannath Neurointervention Center, Department of Imag-

ing Sciences and Interventional Radiology, Neurointervention Center, Sree
Chitra Institute of Medical Sciences and Technology, Trivandrum, Kerala,

T. R. Kapilamoorthy Department of Imaging Sciences and Interventional

Radiology, Sree Chitra Institute of Medical Sciences and Technology, Trivan-
drum, Kerala, India

Kyung-Wha Kim New York Presbyterian Hospital, Weill Cornell Medical

College of Cornell University, New York, NY, USA

Jonathan D. Kirsch Department of Diagnostic Radiology, Yale University

School of Medicine, New Haven, CT, USA

Nicholas A. Koontz Department of Radiology, University of Utah Health

Sciences Center, Salt Lake City, UT, USA

Mark Kruit Department of Radiology, Leiden University Medical Center,

Leiden, The Netherlands

Giuseppe Lanzino Department of Neurologic Surgery, Mayo Clinic, Roch-

ester, MN, USA

Lise Le Jean Department of Anesthesia and Critical Care, Neuro-intensive

Care Unit, Pitié- Salpêtrière Hospital, Paris VI University, Paris, France

Xinyi Leng Department of Medicine and Therapeutics, The Chinese Univer-

sity of Hong Kong, Prince of Wales Hospital, Hong Kong, SAR, China

David S. Liebeskind UCLA Department of Neurology, UCLA, Los Angeles,


Jennifer Linn Institute of Neuroradiology, University Hospital Carl Gustav

Carus, Dresden, Germany

Pierleone Lucatelli Department of Radiological, Oncological and Anatomo-

Pathological Sciences, Sapienza University of Rome, Rome, Italy

Paolo Machi Department of Neuroradiology Unit, CHU Gui de Chauliac,

Montpellier University Hospital, Montpellier, France

Matylda H. Machnowska Division of Neuroradiology, Department of

Medical Imaging, School of Medicine, Sunnybrook Health Sciences Centre,
University of Toronto, Toronto, ON, Canada
Contributors xxiii

Nathan Manning Department of Neurological Surgery, Columbia Univer-

sity Medical Centre, New York, NY, USA
Jean-Louis Mas Service de Neurologie, Faculté de Médecine, Centre
Hospitalier Sainte-Anne, Université Paris Descartes, Sorbonne Paris Cité,
Paris, France
Tomasz Matys Department of Radiology, School of Clinical Medicine,
University of Cambridge, Cambridge Biomedical Campus, Cambridge, UK
Michael Mayich Department of Radiology, University of Chicago, Chicago,
Jean Francois Meder Service de Neuroradiologie, Centre Hospitalier
Sainte-Anne, INSERM UMR894, Université Paris Descartes, Sorbonne
Paris Cité, Paris, France
Sarah Milla Department of Radiology, Emory University School of Medi-
cine Egleston Hospital, Children’s Healthcare of Atlanta, Atlanta, GA, USA
Timothy R. Miller Department of Diagnostic Radiology, Neuroradiology,
University of Maryland Medical Center, Baltimore, MD, USA
Maxim Mokin Department of Neurosurgery, School of Medicine and
Biomedical Sciences, University at Buffalo, State University of New York,
Buffalo, NY, USA
Department of Neurosurgery, Gates Vascular Institute/Kaleida Health,
Buffalo, NY, USA
Roberto Montisci Department of Vascular Surgery, AOU of Cagliari,
Monserrato, Cagliari, Italy
Saul F. Morales-Valero Department of Neurologic Surgery, Mayo Clinic,
Rochester, MN, USA
Svein Harald Mørkve Department of Neurosurgery, Haukeland University
Hospital, Bergen, Norway
Olivier Naggara Service de Neuroradiologie, Centre Hospitalier Sainte-
Anne, INSERM UMR894, Université Paris Descartes, Sorbonne Paris Cité,
Paris, France
Service de Radiologie pédiatrique, Faculté de Médecine, Hôpital Necker
Enfants Malades, Université Paris Descartes, Sorbonne Paris Cité, Paris,
Aurélien Nouet Department of Neurosurgery, Pitié-Salpêtrière Hospital,
Paris VI University, Paris, France
Tomohisa Okada Department of Diagnostic Imaging and Nuclear Medicine,
Graduate School of Medicine, Kyoto University, Sakyo-ku, Kyoto, Japan
Catherine Oppenheim Service de Neuroradiologie, Centre Hospitalier
Sainte-Anne, INSERM UMR894, Université Paris Descartes, Sorbonne
Paris Cité, Paris, France
xxiv Contributors

Darren B. Orbach Neurointerventional Radiology, Boston Children’s

Hospital, Harvard Medical School, Boston, MA, USA

Sachin K. Pandey Neurointerventional Radiology, Brigham and Women’s

Hospital, Boston, MA, USA
Srinivasan Paramasivam Mount Sinai Roosevelt Hospital, Hyman New-
man Institute for Neurology and Neurosurgery, Centre for Endovascular
Surgery, New York, NY, USA

Athos Patsalides Department of Neurological Surgery, Weill Cornell Medi-

cal Center, New York, NY, USA

Cintia Elias Pires Department of Radiology, Clinics CDPI/DASA, National

Institute of Cardiology, Federal University of Rio de Janeiro, Rio de Janeiro,
Michele Porcu Department of Radiology, AOU of Cagliari, Monserrato,
Cagliari, Italy

Emanuele Pravatà Neuroradiology Department, Neurocenter of Italian

Switzerland, Lugano, Switzerland
Luca Quilici Ospedale Niguarda – Ca’ Granda, Milan, Italy

Stephen Quinet Department of Radiology, School of Medicine and

Public Health, Clinical Science Center, University of Wisconsin, Madison,

Eytan Raz Neurointerventional Radiology Section, Department of

Radiology, NYU Langone Medical Center, New York, NY, USA

Christine Rodriguez-Regent Service de Neuroradiologie, Centre

Hospitalier Sainte-Anne, INSERM UMR894, Université Paris Descartes,
Sorbonne Paris Cité, Paris, France

Jessica Rotman PGY-3 Diagnostic Radiology Resident, New York Presby-

terian Hospital, Weill Cornell Medical Center, New York, NY, USA

Luca Saba Department of Radiology, AOU of Cagliari, Monserrato,

Cagliari, Italy

Beatrice Sacconi Department of Radiological, Oncological and Anatomo-

Pathological Sciences, Sapienza University of Rome, Rome, Italy

Daniel Sahlein Goodman Campbell Brain and Spine, Indianapolis, IN, USA

Naoko Saito Department of Diagnostic Radiology, Saitama Medical Univer-

sity International Medical Center, Saitama, Japan

Osamu Sakai Departments of Radiology, Boston University, Boston Medi-

cal Center, Boston, MA, USA
Department of Otolaryngology – Head and Neck Surgery, Boston University,
Boston Medical Center, Boston, MA, USA
Contributors xxv

Department of Radiation Oncology, Boston University, Boston Medical Cen-

ter, Boston, MA, USA
M. Sawicki Department of Diagnostic Imaging and Interventional Radiol-
ogy, Pomeranian Medical University, Szczecin, Poland
Pamela Whitney Schaefer Neuroradiology, Massachusetts General Hospital,
Boston, MA, USA
Lubdha M. Shah Department of Radiology, University of Utah Health
Sciences Center, Salt Lake City, UT, USA
Maksim Shapiro Neurointerventional Radiology Section, Department of
Radiology, NYU Langone Medical Center, New York, NY, USA
Nasim Sheikh-Bahaei Department of Radiology, School of Clinical
Medicine, University of Cambridge, Cambridge Biomedical Campus,
Cambridge, UK
Ravishankar Shivashankar Department of Diagnostic Radiology, Neurora-
diology, University of Maryland Medical Center, Baltimore, MD, USA
Eimad Shotar Department of Interventional Neuroradiology, Pitié-
Salpêtrière Hospital, Paris VI University, Paris, France
Manohar Shroff Department of Diagnostic Imaging, The Hospital for Sick
Children, University of Toronto, Toronto, ON, Canada
Adnan H. Siddiqui Department of Neurosurgery, School of Medicine and
Biomedical Sciences, University at Buffalo, State University of New York,
Buffalo, NY, USA
Department of Neurosurgery, Gates Vascular Institute/Kaleida Health,
Buffalo, NY, USA
Department of Radiology, School of Medicine and Biomedical Sciences,
University at Buffalo, State University of New York, Buffalo, NY, USA
Toshiba Stroke and Vascular Research Center, University at Buffalo, State
University of New York, Buffalo, NY, USA
Jacobs Institute, Buffalo, NY, USA
Nader-Antoine Sourour Department of Interventional Neuroradiology,
Pitié-Salpêtrière Hospital, Paris VI University, Paris, France
Christopher J. Stapleton Department of Neurosurgery, Massachusetts
General Hospital, Boston, MA, USA
Martin H. Steinberg Departments of Medicine, Pediatrics, and Pathology
and Laboratory Medicine, Boston University, Boston Medical Center, Boston,
Jie Sun Department of Radiology, University of Washington, Seattle,
Jun C. Takahashi Department of Neurosurgery, National Cerebral and
Cardiovascular Center, Suita City, Osaka, Japan
xxvi Contributors

Daniel Tibussek Children’s Stroke Program, Division of Neurology, The

Hospital for Sick Children, University of Toronto, Toronto, ON, Canada
Department of General Pediatrics, Neonatalogy and Pediatric Cardiology,
University Children’s Hospital, Heinrich-Heine University, D€usseldorf,
Denis Trystram Service de Neuroradiologie, Centre Hospitalier Sainte-
Anne, INSERM UMR894, Université Paris Descartes, Sorbonne Paris Cité,
Paris, France
Patrick Turski Department of Radiology, School of Medicine and Public
Health, Clinical Science Center, University of Wisconsin, Madison, WI, USA
Behroze Adi Vachha Neuroradiology, Massachusetts General Hospital,
Boston, MA, USA
Maria Isabel Vargas Department of Neuroradiology, Geneva University
Hospitals, Geneva, Switzerland
Jennifer Vaughn Department of Radiology, Boston Children’s Hospital,
Boston, MA, USA
Brian P. Walcott Department of Neurosurgery, Massachusetts General
Hospital, Boston, MA, USA
Richard H. Wiggins III Department of Radiology, University of Utah
Health Sciences Center, Salt Lake City, UT, USA
Max Wintermark Department of Radiology, School of Medicine, Univer-
sity of Virginia, Charlottesville, VA, USA
Joanna Wojczal Department of Neurology, Medical University of Lublin,
Lublin, Poland
Akira Yamamoto Department of Diagnostic Imaging and Nuclear Medicine,
Graduate School of Medicine, Kyoto University, Sakyo-ku, Kyoto, Japan
Chun Yuan Department of Radiology, University of Washington, Seattle,
Robert Zimmerman Diagnostic Radiology Department, New York Presby-
terian Hospital, Weill Cornell Medical Center, New York, NY, USA
Robert Zivadinov Buffalo Neuroimaging Analysis Center, Department
of Neurology, University at Buffalo, State University of New York, Buffalo,
MRI Clinical Translational Research Center, University at Buffalo, State
University of New York, Buffalo, NY, USA
Part I
The Basics: Embryology, Pathological
Basis of Vascular Diseases, and Imaging
Cerebrovascular Development and
Evolution 1
Maksim Shapiro and Eytan Raz

Contents Abstract
Vascular Neurophylogeny . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 The comprehensive appreciation of the embryo-
The Spinal Cord . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 logical events is paramount to develop a good
The Brain: From Fishes to Humans . . . . . . . . . . . . . . . . . . 6 understanding of the cerebrovasculature in all its
Vascular Neuroembryology . . . . . . . . . . . . . . . . . . . . . . . . . 12 aspects, from the anatomy to the physiology and
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 hence the pathology. Vascularization of the brain
Stage I (Prechoroidal Stage): and spinal cord has certain fundamental similar-
Approximately 4 Weeks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Stage II (Choroidal Stage):
ities. The evolution of brain circulation attests to
Approximately 5 Weeks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 progressive recruitment of already existing vas-
Stage III: 6 Weeks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 cular networks to supply emergent cortical terri-
Stage IV: Approximately 7 Weeks . . . . . . . . . . . . . . . . . . . . 16 tories, rather than development of de novo
Stage V: Approximately 9 Weeks . . . . . . . . . . . . . . . . . . . . . 17
Stage VI: Approximately Adult Pattern . . . . . . . . . . . . . . 17
arterial solutions. This principle can be applied
to many concepts in evolutionary developmental
Practical Aspects of Neurophylogeny biology. And, since, in many aspects, “ontogeny
and Neuroembryology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Fetal Posterior Cerebral Artery . . . . . . . . . . . . . . . . . . . . . . . 18 recapitulates phylogeny” to quote Ernst Haeckel,
Accessory MCA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 father of the embryological parallelism law, the
Dominant Anterior Choroidal Artery . . . . . . . . . . . . . . . . . 19 phylogeny of the brain and spinal cord is a key
Persistent Trigeminal Artery . . . . . . . . . . . . . . . . . . . . . . . . . . 19 concept to understand in order to develop a mind
Persistent Hypoglossal Artery . . . . . . . . . . . . . . . . . . . . . . . . 19
Proatlantal Artery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 framework which becomes then useful for the
Variations in Fusion Patterns of the Basilar Artery . . . 20 approach of embryology.
Duplications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Anterior Cerebral Artery Variations . . . . . . . . . . . . . . . . . . 25 Part of the text, the concept, and the figures in this
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 chapter are used with permission from Maksim
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Shapiro, from the website

Vascular Neurophylogeny

M. Shapiro (*) • E. Raz The Spinal Cord

Neurointerventional Radiology Section, Department of
Radiology, NYU Langone Medical Center, New York, The neural tube and the spinal cord can be thought
e-mail:; of, simplistically, as a cylinder (Fig. 1). After; closure of the neural tube, nutritional support
# Springer Science+Business Media New York 2016 3
L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_15
4 M. Shapiro and E. Raz

can be provided by simple diffusion, but, at a coming from the two dorsal aortae (Fig. 2). These
certain point of body growth, dedicated vascular vessels are also called metameric: in biology,
supply to the spinal cord and adjacent tissues is metamerism is the phenomenon of having a linear
needed and is established using segmental vessels series of similar body segments; each segment
may be called somite or metamer. Each
metameric vessel supports its own block of tissue,
including the endodermal, ectodermal, and meso-
dermal tissue (Fig. 3). As diagrammatically shown
in Fig. 4, between segmental vessels longitudinal
anastomoses extending craniocaudally along the
length of the cord are then formed (Fig. 4),
establishing a primordial matrix of the adult vascu-
Spinal Cord lature. Once the longitudinal connection between
the metamers is established, there is no need for a
segmental vessel at each single level, and most
segmental vessels shrink to supply only the region
of the nerve root of the same level (Figs. 5 and 6).
The spinal cord is now supplied by an anterior
spinal artery and posteriorly by two spinal vessels
that will become the posterior spinal arteries. The
Fig. 1
segmental vessels that supply flow to the anterior
spinal artery are called radiculomedullary arteries,
while the ones that supply only the nerve root are
called radicular branches. Sometimes, one segmen-
Dorsal tal vessel supplies two nerve roots via longitudinal
Aorta anastomoses (Fig. 7). As the spinal cord continues
Folded Neural Tube to enlarge, sulco-commissural branches are
formed, which can be matched with the brain per-
forator arteries (Fig. 8).
In Figs. 9 and 10, the spinal angiographic cor-
relation can be visualized in vivo with subtracted
and unsubtracted views which demonstrate a
radiculomedullary vessel giving rise to an anterior
Fig. 2 spinal artery.

Fig. 3
To ribs,
Dorsal etc.

Segmental (a.k.a Metameric)

1 Cerebrovascular Development and Evolution 5

Fig. 4
Devloping To ribs,
Longitudinal muscle,
Anastomoses etc.

Segmental (a.k.a. Metamaric)


Fig. 5
To ribs,

Future Anterior Spinal Artery

Segmental Vessels

Fig. 6

Longitudinal vessel, future

Anterior Spinal artery

Surface pial vessel

Most segmental vessels regress and remain

purely radicular; several dominant
radiculomedullary ones remain
6 M. Shapiro and E. Raz

Fig. 7
Obviously, only To ribs,
one aorta muscle,
remains in the Anterior etc.
adult, from Spinal
which left and Artery
right segmental
vessels emerge.
Two are shown
here for

Radicular artery
Dominant radicular artery
supplying two levels

variations. The fact that newly formed vessels

Anterior are the exception, while existing pathways are
Spinal usually utilized and enlarged to accommodate
tissue needs, was recognized since the early twen-
tieth century.

The earliest vertebrates still existing today are the
jawless fishes, cyclostomes, and hagfish. The
brains of these animals possess all of the features
of vertebrates, including five divisions of the
brain: medulla, pons, midbrain, diencephalon,
Sulco-comissural Arteries and telencephalon (Fig. 11). Even though some
earlier fishes lack a cerebellum (lampreys and
Fig. 8 hagfish), its presence becomes a regular factor in
later fish evolution. Other than a primitive cere-
bellum, in fishes we see a small but effective
olfactory lobe and a primitive three-layer cortex,
The Brain: From Fishes to Humans the forerunner of the hippocampus. The “higher
center” of movement is controlled by what will
When thinking about the brain neuroembryology eventually become the present-day basal ganglia
and neurophylogeny, it is useful to conceptualize (Fig. 11). The spinal cord will remain essentially
its structure as a complex spinal cord, supplied by unchanged, supplied by segmental radiculo-
segmental vessels. Once, together with higher medullary arteries. The vertebral arteries form as
organism complexity, more complicated anatom- a result of longitudinal anastomoses between
ical structures are added on, we will see a more metameric branches of the cervical somites
sophisticated vasculature pattern that is better (Fig. 12).
understood keeping in mind the very simple In the brain two large longitudinal vessels – the
basic segmental arrangement of the spinal cord. carotids – travel up to the neck and enter the
We can think of the developing brain as a progres- cranial cavity, in a way similar to that of segmental
sively enlarging segment of the cord, requiring radiculomedullary vessels in the cord (Fig. 12).
larger surface and deep perforator vessels as key Inside the skull, the internal carotid artery gives
to understanding evolution and vascular rise to paired longitudinal arteries: one that goes
1 Cerebrovascular Development and Evolution 7

Fig. 9 Frontal view from a spinal angiogram, subtracted spinal artery (black arrow). In this particular case, the
and unsubtracted views, demonstrating a radiculo- radiculomedullary vessel arises from the T9 intercostal
medullary vessel (blue arrow) giving rise to an anterior artery (white arrow)

Fig. 10 Frontal view from

a spinal angiogram,
subtracted view. Anterior
spinal artery (red arrows)
originating from a
radiculomedullary branch
(orange arrow) arising from
the costocervical trunk
(purple arrow). This
conformation attests to the
metameric nature of the
system, where the spinal
artery can arise from
longitudinal vessels other
than the vertebral artery. A
tumor blush (yellow arrow)
from T1 hepatoma
metastasis is present, along
with several coil masses
(black arrows) deposited
after embolization of the
left-sided tumor soft tissue
8 M. Shapiro and E. Raz

up, the “cranial ramus,” and one extending cau- the forerunner of the PComm and of a small part
dally, the “caudal ramus” (Fig. 12). The “cranial of the basilar artery, which, at this point, is already
ramus” is the forerunner of the anterior cerebral fused at midline. At this point of development, a
and middle cerebral arteries. The caudal ramus is functionally significant anastomosis between the
caudal ramus and the posterior circulation system
is not activated yet, and the entire intracranial
CNS (including the brainstem and the cerebellum)
is supplied by the carotids.
The cranial ramus has two notable branches:
the medial olfactory branch, which is the future
Primitive ACA, and the lateral olfactory branch, which is
Cortex the forerunner of the Heubner and of the anterior
(mostly choroidal artery. The caudal ramus gives off a
forerunner of tectal artery which supplies the posterosuperior
hippocampus) aspect of the midbrain; from this vessel comes
off a cerebellar artery supplying the primitive
cerebellum – forerunner of the superior cerebellar
Forebrain artery.
Segmental anastomoses also exist, usually in
Midbrain very small form, between the carotid and vertebral
systems. In the human embryo, these pathways
are quite functional at early stages while they
Cerebellum regress as the vertebrobasilar system. Occasional
persistence of these channels leads to the presence
of anomalous vessels such as the persistent tri-
Cord geminal artery, the persistent hypoglossal artery,
or the proatlantal arteries, depending on the seg-
ment that failed to develop.
Fig. 11

Fig. 12 Medical Olfactory

Lateral Olfactory

Cranial Ramus

Caudal Ramus
Carotid Tectal
Fused Caudal Ramus (Basilar)

Segmental Inferior Cerebellar (PICA)

Carotido- Vertebral
1 Cerebrovascular Development and Evolution 9

Fishes Amphibians
Medial Olfactory
Lateral Olfactory
Lateral Striate
Posterior Telencephalic
(future Ant. Choroidal)
Cranial Ramus
Caudal Ramus
Posterior Choroidal
Fused Caudal Ramus (Basilar)
Inferior Cerebellar (PICA)

Segmental Carotido -
vertebral Anastomoses


Fig. 13

Amphibians cranial ramus arise multiple “perforators” supply-

The changes seen with the shift to amphibians are ing the cerebral hemispheres (Fig. 14). These
schematized in Fig. 13. The brain continues to multiple perforators will eventually become the
grow, with a larger primitive hippocampus and MCA. We understand now the concept that the
larger basal ganglia. This larger size of the very large human MCA is a relatively late
brain leads to enlargement of the supplying, phylogenetical acquisition and represents a
already existing branches and to the fusion of hypertrophied perforator branch of the anterior
smaller feeders into larger arteries (Fig. 13). cerebral artery (Fig. 14). From another perspec-
The lateral olfactory branch has enlarged with tive we can also understand how the MCA can be
additional named vessels and now consists of embryologically and phylogenetically considered
two main branches – the lateral striate artery, a branch of the ACA.
forerunner of the Heubner, of the Middle cerebral The “posterior telencephalic branch” of the
artery (MCA), and of the MCA perforators, and ICA, forerunner of the anterior choroidal, supplies
the “posterior telencephalic branch,” forerunner the posterior telencephalic structures and is
of the anterior choroidal artery (Fig. 13). On the connected anteriorly with the olfactory artery;
caudal ramus, the tectal/cerebellar complex this function will continue to be fulfilled until
assumes supply of a developing choroid plexus very late in mammals’ evolution, when the poste-
with a posterior choroidal branch. The anastomo- rior brain territory will be annexed by the “poste-
sis with the posterior circulation is still absent in rior cerebral artery.” Other important steps in
these animals. reptiles are the formation of an anterior anastomo-
sis between the medial olfactory arteries, which is
Reptiles the future Acomm, and the enlargement of the
The changes seen with the shift from amphibians cerebellum, particularly of the more caudal ver-
to reptiles are schematized in Fig. 14. From the mis, which is supplied by the distal “basilar”
lateral striate (Heubner forerunner) branch of the system.
10 M. Shapiro and E. Raz

Fig. 14 Amphibians Reptiles

Medial Olfactory
Lateral Olfactory
Lateral Striate
Lat. Striate perforators (MCA)

Posterior Telencephalic
(future Ant. Choroidal)
Posterior Choroidal

Inferior Cerebellar (PICA)

Segmental Carotido-
vertebral Anastomoses

Fig. 15 Reptiles Birds

Medial Olfactory (ACA)
Lateral Olfactory

Lat. Striate perforators

Posterior Telencephalic
(future Ant. Choroidal)

Posterior Choroidal

Inferior Cerebellar (PICA)

Segmental Carotido -
vertebral Anastomoses


Birds of its territory (Fig. 15). This large telencephalic

In the birds a discrete MCA emerges as a domi- territory of the anterior choroidal will be lost at the
nant perforator from among a collection of lateral early stages of embryogenesis. In the birds, the
striate vessels, driven by rapidly enlarging cortical vertebral system remains confined to the cord or in
volume (Fig. 15). The posterior cerebral artery some cases lower medulla/vermian region
(PCA) of birds, which is in fact the anterior cho- (Fig. 15).
roidal, continues as the dominant parieto-occipital
supply source, while the tectal artery (which is the Mammals
real PCA) extends its coverage and forms a func- The biggest change in mammals regards the
tional anastomosis with the anterior choroidal development of the caudal ramus, since the
system, setting the ground for future annexation enlargement of the cerebellum and of the posterior
1 Cerebrovascular Development and Evolution 11

Fig. 16 Birds Mammals

Lat. Striate perforators

Ant. Choroidal
Posterior Choroidal

Tectal (PCA)
Inferior Cerebellar (PICA)

Segmental Carotido -
vertebral Anastomoses

Fig. 17 Mammals Human

Lat. Striate perforators
Ant. Choroidal
Posterior Choroidal
Tectal (PCA)
Superior Cerebellar

Inferior Cerebellar (PICA)

Segmental Carotido-
vertebral Anastomoses

telencephalon place increasing demand on the portions of basilar territory, resulting in reversal
carotid system. Notwithstanding, also in some of craniopetal basilar flow to the craniofugal form
mammals such as the sheep, the vertebral system observed in monkeys, apes, and humans (Figs. 16
does not contribute to brain supply, and the basilar and 17).
artery is continuing to demonstrate craniopetal The phylogenetically novel participation of the
flow from the internal cerebral system. However, vertebral system in brain supply is maximized in
in higher mammals the carotid system appears to the human, where the posterior cerebral artery,
reach its functional limit, leading to progressive previously known as the tectal artery in the
annexation of the brainstem and cerebellar terri- “lower” species, is now functionally acquired by
tory by the vertebrobasilar system (Fig. 16). At the posterior circulation, even though develop-
first, this is confined to the PICA region. Progres- mentally it belongs to the carotids (Fig. 17).
sively, however, the vertebral artery annexes Even as such, frequent occurrence of “fetal”
12 M. Shapiro and E. Raz

Pcomm disposition attests to this relatively recent The way the vessels are arranged, i.e., trans-
phylogenetic acquisition (see embryology sec- verse vessels connected by longitudinal ones
tion). Recent territory shifts in this region (Fig. 18), is called metameric, because they ulti-
are borne out by marked variability in PICA- mately relate to the metamers, somites, segments,
AICA dominance, whereas the phylogenetically levels, etc., that are the basic building blocks of
older superior cerebellar arteries are much less our anatomy, similar to the arrangement found in
variable. the lower species.
The early arrangement centers on the six aortic
arches, which relate to the branchial arches
Vascular Neuroembryology (Fig. 19). The aortic arches consist of the dorsal
and ventral aortae, connected by longitudinal
Introduction anastomoses – the arches themselves. There are
a few adult arrangements which are almost
At about 4 weeks of gestation, the neural plate has constant:
just closed, and the brain is a tube with a central
canal of fluid, similar to the spinal cord. A series • The left-sided fourth aortic arch will become
of small vessels wrap around the brain and nourish the adult aortic arch.
it by simple diffusion in a similar way to that of the • The third arch will become the proximal inter-
spinal cord (Fig. 18). nal carotid artery.

Fig. 18
1 Cerebrovascular Development and Evolution 13

Fig. 19

• The dorsal aorta between the third and fourth Stage I (Prechoroidal Stage):
arch regresses, while homologous ventral aorta Approximately 4 Weeks
becomes the common carotid.
• The ventral aorta rostral to the third arch will There is a large vessel arising from the fourth
become the external carotid artery. aortic arch (Fig. 21). Keeping in mind that this is
a simplification, we will call it the common
Other than the dorsal and ventral aortae, there carotid. The place of origin of the internal carotid
is an additional longitudinal system, called the is shown, but again the external carotid is not
longitudinal neural system, related more displayed. The carotid system appears to feed
directly to the developing brain and spinal cord the entire brain, which is in fact true of the early
(Fig. 20), which consists of a series of metameric human embryology; the posterior circulation does
vessels with longitudinal connections. This longi- not exist yet. From this carotid vessel, a series of
tudinal neural system will eventually mature small arteries wrap around the brain and nourish it
into the vertebrobasilar system. We will go by simple diffusion (Fig. 21).
through six stages: the prechoroidal stage At this point, the carotid divides in two rami –
(stage 1), the choroidal stage (stage 2), the telen- the cranial and the caudal. The cranial ramus is
cephalic primitive stage (stage 3), the telence- simply a larger feeding vessel among the many
phalic intermediate stage (stage 4), the smaller vessels that wrap around the brain. It is the
telencephalic final stage (stage 5), and the adult forerunner of the anterior cerebral artery, which is
disposition (stage 6). the phylogenetically oldest telencephalic vessel.
14 M. Shapiro and E. Raz

Fig. 20

The caudal ramus feeds the midbrain and hind- These transverse arteries are in hemodynamic
brain and is the forerunner of the future Pcomm, balance with the caudal ramus of the carotid artery
P1, and part of the basilar. These adult segments in supply of the brainstem. As the longitudinal
therefore all belong embryologically to the ante- neural system matures, the future vertebral artery
rior circulation; this is a quite important concept to begins to meet to demands of the brainstem and
bear in mind in order to understand the vascular cerebellum, and, as a consequence, the anasto-
variation and pathology. Blood flow in the caudal motic transverse arteries typically regress, though
ramus at this point is craniocaudal. never fully disappear. Rarely, one of these vessels
Four transverse vessels are known to anasto- remains dominant and thus continues to supply
mose the caudal ramus with the ICA, the brainstem, for example, in the so-called per-
corresponding to the segmental arrangement men- sistent trigeminal artery. As was mentioned, the
tioned above. These segmental vessels are named posterior circulation does not exist as such in
trigeminal, hypoglossal, and proatlantal type embryological terms. Now look again at the
1 and 2 arteries (Fig. 22). They are very similar small vessels surrounding the brain. From
to the abovementioned transverse metameric among many of these feeders, a dominant vessel
arteries; they receive special attention because of eventually arises to wrap around the brain and
their relationship with branchial arteries and sprout its own small branches. The concept of
nerves. For example, the trigeminal artery sup- selection for a dominant vessel among many is
plies the trigeminal nerve, which is the nerve of very important, because it explains most “vari-
the first arch. ants” seen in the adult. From this arrangement it
1 Cerebrovascular Development and Evolution 15

Fig. 21

is clear that if two of these small vessels continue telencephalon. The choroid plexus of the dien-
to be in hemodynamic balance instead of one cephalon and midbrain is supplied by the artery
vessel being dominant, an apparent duplication called “posterior choroidal” (Fig. 24). The con-
may ensue. nection between telencephalic and diencephalic
central tube – where in the future the lateral and
third ventricles communicate through the foramen
Stage II (Choroidal Stage): of Monro – is important as the area of hemody-
Approximately 5 Weeks namic anterior choroidal/posterior choroidal bal-
ance. The choroid of the medulla (future 4th
The brain continues to be supplied entirely by the ventricle) is being supplied by a dominant feeder
anterior circulation, and the choroid (purple stuff that will become the PICA. Variations of posterior
over the blue tube in Fig. 23) develops as an fossa supply are common, but the 4th ventricle
invagination of the ependymal cells into the cen- choroid plexus is almost always supplied by a
tral canal from the “roof” of the brain and helps PICA-like vessel even when an AICA-PICA var-
supply nutrients to the brain from the inside of the iant is present.
neural tube. The choroid plexus is quite metabol- The early cerebellum is fed by a ramus that
ically active and promotes development of its own belongs to the anterior circulation, the future supe-
vascular supply. At this stage, a vessel is seen rior cerebellar artery. As a rule, the older the
coming off the cranial ramus to feed the plexus vessel, the more constant its presence. And this
of the telencephalon – this is the future anterior is particularly true for the SCA, which is essen-
choroidal artery (Fig. 23). It is a hemodynamic tially always present, while variations involving
balance with the distal cranial ramus (the ACA) the AICA and PICA are quite common (Fig. 25).
and is the earliest feeder of the posterior When a case of agenesis of these early vessels is
16 M. Shapiro and E. Raz

Fig. 22

considered carefully, the vessel is in fact found to the homologue of the future anterior spinal artery.
exist under an incorrect name. For example, ante- The one more laterally related to the vertebral
rior choroidal agenesis in fact represents an bodies will become the future vertebral artery.
enlarged anterior choroidal that supplies the In the anterior circulation, the growing cere-
PCA territory, while the latter is hypoplastic. bellum is going to annex the brainstem forerun-
This is confirmed by the observation that in ners of the AICA and PICA, and the branches of
most mammals the flow in the vertebrobasilar the cranial ramus develop branches of their own,
system is directed caudally and the reversal of some of which will become important adult struc-
flow of the posterior circulation is a very recent tures, such as the Heubner or the MCA, but at this
phenomenon. The important changes taking place point the ACA and the anterior choroidal continue
in the longitudinal neural system (LNS) are to dominate the supply of the telencephalon.
responsible for this and will be illustrated in the
next stage.
Stage IV: Approximately 7 Weeks

Stage III: 6 Weeks The LNS continues to “mature” in an interesting

way: the two longitudinal neural arteries that are
What happens is that the discontinuous LNS the forerunners of the anterior spinal artery and the
plexus coalesces to form actual distinct longitudi- basilar artery fuse into one vessel for most of the
nal vessels (Fig. 20). The one next to the cord is length of the cord and the brainstem. The whole
known as the “longitudinal spinal artery” and is vasculature is starting to resemble the adult
1 Cerebrovascular Development and Evolution 17

Fig. 23

pattern. Still, at this point the flow in the new change in flow direction seems to trigger some
basilar remains caudally directed. The PICA con- of the final events of embryogenesis. The relief of
tinues to expand to cover the cerebellum and the anterior circulation from the responsibility to
acquires its two basic divisions – choroidal and supply the posterior brain and the posterior
parenchymal. The trigeminal, hypoglossal, and fossa allows for continued growth of the PCA
proatlantal vessels continue to regress as the ver- and for development of the true MCA from sev-
tebral axis matures. eral ACA perforators. Indeed, conceptually and
As regards the anterior circulation, the most embryologically, the MCA is simply a very large
important change concerns the development of perforator with cortical territory, similar to the
the future “PCA,” which takes over the territory Heubner. For example, the accessory MCA vari-
previously supplied by the anterior choroidal ant is simply a case of two perforators being in
artery (Fig. 26). The growth of the PCA parallels relative balance and thus developing into
the development of the vertebrobasilar system, by MCA-like vessels.
which it will be annexed. But it is not known
whether it takes place after the flow in the basilar
has been reversed or before. Stage VI: Approximately Adult Pattern

One of the particular curiosities of the late vascu-

Stage V: Approximately 9 Weeks lar embryogenesis is the seemingly sudden
appearance of the Acomm (Fig. 28). Conceptu-
The posterior circulation has assumed a more or ally, it is easy to think of it as a short segment
less adult pattern (Fig. 27). At some point about fusion of the ACA in a fashion homologous to the
this time, the flow becomes craniocaudal. The basilar.
18 M. Shapiro and E. Raz

Fig. 24

To conclude, it is noteworthy to state how this A unilateral fetal posterior cerebral artery is seen
section only illustrates the basic framework of the in 25 % of cases, bilateral in 5 % cases (Figs. 29,
cerebral circulation, but the development of 30, and 31). In Figs. 30 and 32 examples of fetal
smaller vessels, countless branches of branches, Pcomm strokes are demonstrated. The existence
and of the penetrating vasculature continues well of a “duplicated Pcomm” is not possible – as may
after birth, as the brain matures and this level of be confused on occasion with a larger anterior
complexity is quite beyond our scope. choroidal artery.

Practical Aspects of Neurophylogeny Accessory MCA

and Neuroembryology
The MCA (purple in Fig. 33) arises as a dominant
Fetal Posterior Cerebral Artery vessel from a number of lateral striate perforators
of the anterior cerebral artery in the reptile. It may
This configuration is the most common “variant” happen in the case where two perforators remain
of the cerebral circulation. The fetal Pcomm, or dominant, resulting in the “accessory MCA”
fetal origin of the posterior cerebral artery, occurs (Fig. 33). The accessory vessel is, in fact, either
when the Pcomm is larger than the P1 segment a hypertrophied recurrent artery of Heubner – a
and supplies the bulk of the posterior cerebral medial ACA perforator – or another perforator-
artery territory. This reflects the effective persis- like vessel. To qualify as an MCA, the vessel must
tence in carotid supply of the PCA, a reminder of have cortical territory. In the cartoon in Fig. 33,
the phylogeny and embryology of this vessel. the schematic on the left side shows the accessory
1 Cerebrovascular Development and Evolution 19

Fig. 25

MCA configuration where both branches appear cortex remaining under anterior choroidal control
to originate proximal to the A1 complex (which is while PCA contribution is correspondingly
here defined as segment past the more “distal” reduced. This is sometimes erroneously consid-
MCA branch). These are known as “Manelfe ered a “duplicated PCA” (Fig. 36). Mistaking the
type 1” or “type 2” – depending on which branch choroidal artery for the Pcomm can have disas-
is larger. The important feature however is to note trous surgical consequences.
from which vessel the perforators originate and
whether they are medial or lateral. The schematic
on the right side shows the Heubner-type aMCA, Persistent Trigeminal Artery
known as “Manelfe type 3.” Check out the exam-
ples in Figs. 34 and 35. This most common embryonic carotid-
vertebrobasilar anastomosis at the mid-basilar
level represents persistence of segmental
Dominant Anterior Choroidal Artery ICA-basilar connection along the trigeminal
nerve (Fig. 37). The posterior communicating
The anterior choroidal artery, until late in evolu- artery is usually small in these cases.
tion, supplies a substantial portion of cortical ter-
ritory, which in the human is annexed by the PCA.
This annexation is determined by the proximity of Persistent Hypoglossal Artery
the two vessels in the subarachnoid space around
the midbrain. On occasion, the degree of annexa- More caudal than the persistent trigeminal artery,
tion is incomplete, with large portions of the the persistent hypoglossal artery enters the cranial
20 M. Shapiro and E. Raz

Fig. 26

vault through the hypoglossal canal, connecting Variations in Fusion Patterns

the distal vertebral and the proximal internal of the Basilar Artery
carotid arteries. It is, like the trigeminal and
proatlantal vessels, a segmental artery. Examples The basilar artery is formed via coalescence of
in Figs. 38 and 39. multiple channels belonging to the longitudinal
neural system, following development of the ante-
rior circulation. This process sets the stage for
Proatlantal Artery multiple basilar artery variations, which can be
considered from the standpoint of (1) extent of
An embryonic carotido-vertebral anastomosis, fusion and (2) completeness of fusion (i.e.,
the proatlantal artery is, in fact, the fenestrations).
occipital artery. A C1 segmental connection It is conceptually helpful to think of the basilar
between the occipital and vertebral artery is the artery as being “zipped” in the middle, with ver-
proatlantal type I; C2 level connection is tebral and PCA segments being “unzipped.”
proatlantal type II (Fig. 40). Usually, the vertebral Thus, both length and the integrity of the zipper
artery proximal to the proatlantal segment is determine the final configuration of the artery. As
hypoplastic. such, the artery may be:
1 Cerebrovascular Development and Evolution 21

Fig. 27

1. “Short” – corresponding to relative lack of aberration but instead a “deficiency” in basilar

fusion of the caudal segment fusion (Fig. 42).
2. Unfused at the top – corresponding to P1 origin of the superior cerebellar artery. The
“unzipping” at the basilar tip, less common SCA (blue) on the right arises from the P1 seg-
but much more confusing ment. Further “zipping” up would elongate the
3. Fenestrated – broken zipper in the middle basilar trunk, foreshorten both P1 segments, and
shift the SCA onto the basilar (Fig. 43). When
The first one, short basilar, is in fact a contin- especially prominent, basilar tip “unzipping” can
uum in terms of fusion extent (Fig. 41). There is lead to some confusion, as in this patient reported
no “number” to guide what short or long is, but to have a basilar tip aneurysm on CTA. Catheter
some basilar arteries are clearly quite angiogram shows an unfused basilar tip with tor-
foreshortened (the tortuous basilar of the aged tuous P1 vessels simulating aneurysm, further
hypertensive is a different matter). The second complicated by the presence of a fetal Pcomm
variation, involving lack of basilar tip fusion, on the right.
can generate a lot of confusion (Fig. 42). Effec- The third, basilar fenestration (Fig. 44), is quite
tively, the top of the basilar is split in two, so that common and usually of little clinical significance,
one or both superior cerebellar arteries originate except when it is so short as to mimic a dissection
from the P1 segment. This variant is not, there- (Fig. 45). Occasional species, including some
fore, a primary superior cerebellar artery monkeys, do not have a basilar artery at all. The
22 M. Shapiro and E. Raz

Fig. 28

Fig. 29 Bilateral fetal Pcomm (blue arrows) which reflect effective persistence in carotid supply of the PCA. The P1
segment of the PCA (brown) is hypoplastic
1 Cerebrovascular Development and Evolution 23

Fig. 30 CTA reconstruction demonstrating a right fetal related to a clot in both the fetal Pcomm and the MCA with
Pcomm (black arrows). This disposition was particularly a very large infarct. The AChoA territory was preserved
unfavorable in this patient when he/she developed a stroke

Fig. 31 Fetal Pcomm on a cerebral angiogram marked in Notice other unusual arrangements in this case, including a
red on left ICA injection lateral projection. AP vertebral somewhat low origin of the left superior cerebellar artery
artery injection shows a small left P1 segment (orange), as (yellow) and a basilar fenestration (brown)
compared with normal course of the right PCA (green).
24 M. Shapiro and E. Raz

Fig. 32 PCA territory infarction in a case of “fetal the right ICA. Typical diminutive appearance of the basilar
Pcomm.” Time to Peak and rCBV maps (top) demonstrate artery in a patient with bilateral fetal Pcomms is seen on
a completed infarction in right PCA territory. The embolus bottom right T2 image
(right lateral ICA injection on bottom left) originated from

caudal rami are unfused throughout, mimicking identical one, since these branches actually
the anterior cerebral disposition. supply separate territories. Understanding that
arteries in essence are dominant collectors arising
from a number of potential homologous candi-
Duplications dates clarifies this occurrence. The most com-
monly duplicated vessel is the superior
As in the case of accessory MCA, essentially any cerebellar artery. Duplicated AICA is quite rare
vessel on occasion is gifted with a twin, not an (Fig. 46).
1 Cerebrovascular Development and Evolution 25

Anterior Cerebral Artery Variations

The number of potential ACA variations may

seem mind-boggling. Phylogeny and embryology,
however, help make sense of frustration and con-
fusion. The appearance of the ACA system can be
conceptualized in terms of two factors – extent of
fusion and pattern of branch takeoff.
On the fusion side, you can expect anything
from no fusion – that is, no Acomm, various short
segment patterns of fusion (duplicated Acomm),
or elongated segments of fusion (unpaired or azy-
gos arrangements). The unpaired type is one
where a long segment of fusion finally breaks up
into two branches, each of which gives rise to
Fig. 33
individual cortical branches to its respective hemi-
sphere. The azygos type is where the ACAs are
fused for the entire length, with individual
branches to both hemispheres arising from a com-
mon trunk. Thinking of unpaired/azygos arrange-
ments as a kind of “basilar” is helpful and may not
be conceptually incorrect. Many species, such as
dogs, have a “normal” azygos arrangement with
“variants” of unpaired or “human-type” disposi-
tion. In the human, unpaired cases have a much
higher incidence of aneurysm formation at the
eventual branch point, – not a favorable evolu-
tionary development.
On the takeoff pattern side, you can have any
takeoff pattern imaginable, but of particular
importance is the arrangement at the Acomm
region. The frontopolar branch may arise from
the Acomm complex. More importantly, the
callosomarginal-pericallosal split may occur any-
where along the length of ACA from Acomm to
the callosum genu. Particularly when this split
occurs at the Acomm region, a triplicated appear-
ance may result with pericallosal and
callosomarginal trunks arising from one side and
regular A2 from the other, for a total of three
Fig. 34 Catheter angiogram of a patient with type I/II arteries (Fig. 47). Ultimately, ACA analysis in
accessory MCA. Notice how well MCA perforators can terms of some combination of fusion extent and
be seen, originating from the smaller and more distal of the branch takeoff can clarify any ACA arrangement.
two MCA vessels
26 M. Shapiro and E. Raz

Fig. 35 Type III (Heubner-

like) aMCA originating
from the Acomm complex

Fig. 36 The anterior choroidal artery (black), until late in large portions of the cortex remaining under anterior cho-
evolution, supplies a substantial portion of cortical terri- roidal control (blue), while PCA contribution is corre-
tory, which in the human is annexed by the PCA (brown). spondingly reduced (green). On MRA images, the
On occasion, the degree of annexation is incomplete, with Pcomm is labeled in red and choroidal in yellow

anatomy, the physiology, and the pathology of

Conclusion the central nervous system. Countless are the
cases in which this knowledge comes in handy
The deep understanding of the phylogenesis and in the real life of a neuro doctor, radiologist,
embryogenesis of the brain and spine vasculature clinician, or surgeon.
is paramount to develop a good perception of
1 Cerebrovascular Development and Evolution 27

Fig. 37 This diagram

demonstrates the
persistence of the trigeminal
artery. The angiographic
images illustrate a persistent
trigeminal, an accessory
MCA, and an associated
aneurysm treated with coils.
Notice the classic tau shape
of the trigeminal artery in
the lateral view

Fig. 38 More caudal than

the persistent trigeminal
artery (green arrow on the
diagram), the persistent
hypoglossal artery enters
the cranial vault through the
hypoglossal canal,
connecting the distal
vertebral and the proximal
internal carotid arteries. It
is, like the trigeminal and
the proatlantal vessels, a
segmental artery
28 M. Shapiro and E. Raz

Fig. 39 Different patient. Source and reconstructed MRA vertebrals and Pcomms are hypoplastic. Notice the flow
demonstrating a large vessel traversing through the hypo- void on the axial T2-weighted image
glossal canal and becoming the basilar artery. Both

Fig. 40 Stereo AP and lateral views of left ECA injection, (red). This is the proatlantal type I; the proximal occipital artery
opacifying the vertebrobasilar system (yellow) via the C1 (red) is the proatlantal. Occipital artery distal to the vertebral
segmental artery (purple) connection to the occipital artery anastomosis is blue, and internal maxillary artery is black
1 Cerebrovascular Development and Evolution 29

Fig. 41

Fig. 42

Fig. 43 P1 origin of the

superior cerebellar artery.
The SCA (blue) on the right
arises from the P1 segment.
Further “zipping” up would
elongate the basilar trunk,
foreshorten both P1
segments, and shift the SCA
onto the basilar
30 M. Shapiro and E. Raz

Fig. 44

Fig. 45

Fig. 46 Rare case of

duplicated AICA (brown) is
shown on diagram and
frontal view of left vertebral
artery injection
1 Cerebrovascular Development and Evolution 31

Fig. 47 Lateral projection angiogram with early misleading term since not three but two ACA arteries are
callosomarginal takeoff (purple) and pericallosal in red. present. It is the A2 segment on the right that is absent
Corresponding MRA image with left A2 shown in yellow because of early bifurcation into callosomarginal and
and Acomm in dark blue. This is the appearance of a pericallosal
“triplicated ACA” which is discussed above; it is a

8. Lasjaunias P, Manelfe C (1979) Arterial supply for the

References upper cervical nerves and the cervicocarotid anasto-
motic channels: systematization of radiological anat-
1. Abanou A, Lasjaunias P, Manelfe C, Lopez-Ibor L omy. Neuroradiology 18(3):125–131
(1984) The accessory middle cerebral artery 9. Lasjaunias P, Moret J, Manelfe C, Théron J, Hasso T,
(AMCA). Diagnostic and therapeutic consequences. Seeger J (1977) Arterial anomalies at the base of the
Anat Clin 6(4):305–309 skull. Neuroradiology 13(5):267–272
2. Butler AB, Hodos W (2005) Comparative vertebrate neu- 10. Lasjaunias P, Santoyo-Vazquez A (1984) Segmental
roanatomy: evolution and adaptation. Wiley, Hoboken agenesis of the internal carotid artery: angiographic
3. Campos C, Churojana A, Rodesch G, Alvarez H, aspects with embryological discussion. Anat Clin
Lasjaunias P (1998) Basilar tip aneurysms and basilar 6(2):133–141
tip anatomy. Interv Neuroradiol 4(2):121–125 11. Lasjaunias P, ter Brugge KG, Berenstein A (2007)
4. Dimmick SJ, Faulder KC (2009) Normal variants of Surgical neuroangiography: vol. 3: clinical and inter-
the cerebral circulation at multidetector CT angiogra- ventional aspects in children, vol 3. Springer Science &
phy. Radiographics 29(4):1027–1043 Business Media, Berlin
5. Gore AV, Monzo K, Cha YR, Pan W, Weinstein BM 12. Meckel S, Spittau B, McAuliffe W (2013) The persis-
(2012) Vascular development in the zebrafish. Cold tent trigeminal artery: development, imaging anatomy,
Spring Harb Perspect Med 2(5):a006684 variants, and associated vascular pathologies. Neuro-
6. Krings T, Baccin CE, Alvarez H, Ozanne A, Stracke P, radiology 55(1):5–16
Lasjaunias PL (2007) Segmental unfused basilar artery 13. Siqueira M, Piske R, Ono M, Marino Júnior R (1993)
with kissing aneurysms: report of three cases and liter- Cerebellar arteries originating from the internal
ature review. Acta Neurochir (Wien) 149(6):567–574; carotid artery. AJNR Am J Neuroradiol
discussion 574 14(5):1229–1235
7. Lasjaunias P, Braun JP, Hasso AN, Moret J, Manelfe C 14. Vignaud J, Hasso AN, Lasjaunias P, Clay C (1974)
(1980) True and false fenestration of the vertebral Orbital vascular anatomy and embryology. Radiology
artery. J Neuroradiol 7(3):157–166 111(3):617–626
Atherosclerosis: Biological and
Pathological Basis 2
Michele Porcu, Eytan Raz, and Luca Saba

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Atherosclerosis is a extremely complicated and
still well not perfectly understood pathologic
Arteries: Classification and Anatomical
Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
process that involves the arteries of the human
body. It is a multi-step process that starts with
Normal Physiological Aspects . . . . . . . . . . . . . . . . . . . . . . . 34 the endothelial damage and can be influenced
Pathogenesis of Atherosclerosis . . . . . . . . . . . . . . . . . . . . . 34 by many factors, modifiable or not. The defi-
Endothelial Injury and Dysfunction . . . . . . . . . . . . . . . . . . 35 nition of the grade of stenosis determined by an
Plaque Growth and Evolution . . . . . . . . . . . . . . . . . . . . . . . . 35
Types of Plaque and Possible Complications . . . . . . . . . 36
atherosclerotic plaque and the identification of
vulnerable plaques are the main goals for diag-
Imaging Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
nosis, staging and therapies.
Therapeutic Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Keywords
Pathogenesis • Risk factors • Nitric Oxide
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
(NO) • Vulnerable plaque


Atherosclerosis can be defined as a multifactorial

chronic inflammatory disease of the large (elastic)
and medium (muscular) arteries, characterized by
endothelial dysfunction, vascular inflammation,
and deposition of fibro-fatty particles in the con-
text of the vessel wall [1, 2]. This pathological
M. Porcu (*) • L. Saba
Department of Radiology, AOU of Cagliari, Monserrato, process is on the bases of a huge amount of com-
Cagliari, Italy mon clinical conditions, including myocardial
e-mail:; lucasabamd@gmail. ischemia and infarction and cerebrovascular and
peripheral vascular disease.
E. Raz In this chapter, we will briefly analyze the
Neurointerventional Radiology Section, Department of
biological and pathological basis of atherosclero-
Radiology, NYU Langone Medical Center, New York,
NY, USA sis, focalizing our attention in particular on the
e-mail: formation process of the atherosclerotic plaques.
# Springer Science+Business Media New York 2016 33
L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_48
34 M. Porcu et al.

Arteries: Classification and Anatomical Table 1 Principal mediator produced by the endothelium
Structure Molecule Function
Nitric oxide (NO) Vasodilator
The arteries are those tubular shape structures of Prostacyclin (PGI2) Vasodilator
the human in which the blood flows from the heart Endothelin-1 (ET-1) Vasoconstrictor
to the periphery. Endothelium-derived contracting Vasoconstrictor
According to the caliber of the vessel, we can factors (EDCF)
recognize large arteries (caliber between 30 and
7 mm) and medium and small (caliber between are present: these small vessels are fundamen-
7 and 0.2 mm) arteries [3]. Because of the differ- tal for the nutrition of the tunica media and
ent characteristics of the tunica media (see below), could be an important site for
the large arteries are called even elastic arteries, neoangiogenesis and inflammatory
whereas those of medium and small caliber are processes [1].
called muscular arteries.
The vessel wall can be divided into three dif-
ferent layers, which are from the inner to the outer
(1, 2, 3, 4): Normal Physiological Aspects

(a) Tunica intima: it is the innermost, and it An artery is not a simple “conduct” for the blood;
consists of a single layer of endothelial cells by the opposite, it has to be considered a proper
(endothelium) that lie on their basement organ, with proper physiological functions
membrane. The endothelial cells have an (in particular the endothelium).
important role in the production and secretion The endothelial cells can produce and secrete
of several molecules that play a key role in the different molecules after mechanical or chemical
regulation of the blood flux, of the hemostasis, stimulation or the interaction with molecules
and of the inflammatory process. transported in the blood, such as the mediators
(b) Tunica media: this layer is the thickest of the of inflammation [1, 2, 4]: some of the most impor-
three, and it is composed of vascular smooth tant molecules produced by the endothelium are
muscle cells (VSMC), elastin fibers, and reported in Table 1.
extracellular matrix. The elastin fibers are The activity of the VSMC can be regulated
more represented in the tunica media of large directly by the sympathetic and parasympathetic
caliber arteries, whereas the tunica media of system, by the circulating catecholamines (adren-
medium and small arteries consists mainly of aline and noradrenaline) and other molecules such
VSMC. The VSMC are arranged on a circular as angiotensin II, antidiuretic hormone (ADH),
or spiral way. and corticosteroid [4].
The function of this layer in the elastic
arteries is to accumulate the kinetic energy
generated during the systolic phase of the
cardiac cycle, releasing it during the diastolic Pathogenesis of Atherosclerosis
phase in order to facilitate the blood flow to
the periphery; in the muscular arteries, it rules The pathogenesis of atherosclerosis is a compli-
the tone of the arteries and so the amount of cated and still not well-known multistep process.
blood which reaches the organs. Most of these processes are chained one to each
(c) Tunica adventitia: it is the outermost layer other, and a huge number of factors are involved.
and consists of connective tissue in continuity In this paragraph, we will try to explain the
with the perivascular connective tissues. In main elements and mechanisms which lead to the
the context of this layer, the vasa vasorum formation of the atherosclerotic plaque.
2 Atherosclerosis: Biological and Pathological Basis 35

Endothelial Injury and Dysfunction Table 2 Principal modifiable risk factors of

In paragraph two, we have seen how the endo- Principal modifiable risk factors
thelium has to be considered not just an internal Hypertension
smooth cover of the artery but an active tissue Dyslipidemia and hypercholesterolemia
important for many functions. On the surface of Obesity
the endothelium, a great number of receptors are Diabetes
present, for example, insulin, serotonin, hista- Smoking
mine, serum lipoproteins, and adhesion mole-
cules more expressed during the inflammatory
status. endothelial damage induced by the blood flow
The endothelium is also able to produce many (together with the other risk factors we have seen
chemical substances, including nitric oxide before).
(NO) via the endothelial isoform of the nitric
oxide synthases (eNOS) [1, 2, 5]. This chemical
mediator has a very simple chemical structure; it is Plaque Growth and Evolution
unstable, with a very short life (very low seconds),
and once it is produced, it acts on the VSMC Once the endothelial damage has started, the pro-
inducing vasodilation and slowing their prolifera- cess continues with the formation of the athero-
tion and reduces the expression of adhesion mol- sclerotic plaque.
ecules for macrophages on the endothelial surface The first lesions evident in the arteries are the
(see below) [6, 7]. so-called fatty streaks, which are linear lesions
The first step of the pathogenesis of the athero- well evident in particular in younger patients.
sclerosis is the endothelium damage [1, 2, 7]. These kinds of lesions tend to modify during the
This process occurs more frequently in those arte- years, according to various factors:
rial regions in which the blood flow is non-laminar
or turbulent (e.g., the bifurcations or the origin of • Non-modifiable factors (genetic, familiarity)
collateral vessels) [1], and it is favored by “risk • Modifiable factors
factors” such as hypertension, smoking, and
hypercholesterolemia through mechanisms like The modifiable factors include the “risk fac-
free radical oxidation and the activation of multi- tors” mentioned before, i.e., hypertension,
ple genetic pathways [2, 7]. dyslipidemia and hypercholesterolemia, obesity,
The endothelial damage is characterized by the diabetes, and smoke (Table 2): the presence of
expression of pro-inflammatory receptors on their these factors accelerates the process and forma-
surface, the reduction of the production of the NO, tion of the atherosclerotic plaques, increasing the
and the rearrangement of the junction and cyto- probabilities of minor and major cardiovascular
skeletal proteins [2, 8, 9]. The final results are an and cerebrovascular events and pathologies of the
increased endothelial cell turnover, an increased peripheral vascular system [10].
permeability of the endothelium with the recruit- At a microscopic level during this phase, the
ment of inflammatory cells (mainly scavenger macrophages, the T cells, and the VSMC are
macrophages), and the accumulation of lipopro- the most important cellular elements involved;
teins (in particular LDL; see below) beneath the they operate inside the growing lesion in an
tunica intima [2]. This process tends to have a inflammatory context, interacting in particular
positive feedback: the more the permeability, the with the low-density lipoproteins (LDLs).
more the inflammation, the more the lipid accu- LDLs are lipoproteins with a very high rich
mulation, the more the reduction of the caliber of content in cholesterol. These lipoproteins circu-
the lumen of the vessel, and the more the late free in the blood. When the endothelium is
36 M. Porcu et al.

damaged, the LDLs can infiltrate beneath the stiffening of the arterial walls, the physiological
tunica intima and accumulate into the growing mechanisms of narrowing and dilation of the arte-
plaque. At this level, they are susceptible to the rial walls in response to vasoactive molecules and
oxidative stress of various enzymes such as neurovegetative regulation are damaged. Even
myeloperoxidases and lipoxygenases (produced this condition further promotes the atherosclerotic
and released normally during an inflammatory plaque growing process.
process by the immunity cells). The oxidized The continuous growth of the plaque can
LDLs (oxLDLs) thanks to their cytotoxic action extend inside the lumen of the vessel (negative
promote the recruitment of monocytes from the remodeling) or on the external side (positive
blood circulation upregulating the expression of remodeling).
adhesion molecules on the surface of the endothe- The negative remodeling is important because
lial cells [2, 11]. Once they have penetrated inside it is responsible for the reduction of the blood flow
the plaque, they differentiate into resident macro- distally to the plaque, especially during stress
phages [1]. OxLDLs phagocytized by the macro- conditions. For example, a plaque that obstructs
phages are more resistant to degradation of their the lumen of the coronary artery limits the flux of
lysosome, and the macrophages continue to accu- blood to the territories supplied by that vessel: in
mulate cholesterol, becoming foam cells [2]. The the early stages, when the obstruction is not crit-
death of the foam cells led to the formation of the ical (up to 70 % of reduction of the vessel caliber),
necrotic lipidic core inside the plaque [1]. this condition could be compensated during rest,
The macrophages continue to feed this inflam- but during stress condition such as intensive phys-
matory reaction producing chemokines such as ical activity, the presence of the plaque does not
monocyte chemoattractant protein-1 (MCP-1), allow an adequate dilation of the coronary artery
interferon-γ (INF-γ), and macrophage colony- and an adequate blood flow for the heart segments
stimulating factor (MCFS), which increase supplied by the vessel, inducing myocardial
recruitment of other inflammatory cells [2]. ischemia which manifests with pain and
Some of these chemokines attract in the grow- neurovegetative symptoms.
ing plaque and stimulate the proliferation of the On the other hand, the positive remodeling is
VSMC [1]. These cells produce extracellular pro- responsible for the dilation of the vessel involved
teins such as collagen and elastin, producing a in the atherosclerotic process and can lead to the
fibrous cap over the plaque and contributing to formation of aneurysm. The evolution of the
the stability of itself; other molecules produced, aneurismatic lesions is highly variable, even if
the proteoglycans and fibronectin, create a reticu- the most terrible is the rupture of the vessel and
lar web that contribute to trap LDLs inside the the consequent hemorrhage, often fatal.
lesion [12]. Other molecules produced are the
matrix metalloproteinases (MMPs), whose role
in the degradation of the extracellular matrix is Types of Plaque and Possible
very important in the macrophage migration and Complications
destabilization of the plaque [13, 14].
At a microscopic level, it has been seen that as The natural evolution of the atherosclerotic plaque
the plaques increase in complexity, mineralization is highly variable, according to many factors,
processes can develop inside them, leading to the modifiable or not (see before). In particular, we
formation of calcifications [2]. In particular medi- can recognize two types of plaques, non-vulner-
ators of the bone metabolism have been evidenced able and vulnerable.
inside the plaque, close to the calcification The non-vulnerable plaques are characterized
centers [15]. by the presence of a small necrotic lipidic core and
Of course because of the involvement of the a thick fibrous cap and are less prone to rupture.
smooth musculature of the media layer in the The vulnerable plaques are those more prone to
atherosclerotic process and the subsequent rupture and thrombosis [16]. From a histological
2 Atherosclerosis: Biological and Pathological Basis 37

point of view, the vulnerable plaques are usually stenosis of the vessels but even to evaluate the
characterized by the presence of a large fatty composition of the plaques [19].
necrotic core, covered by a very thin fibrous cap The new techniques introduced in magnetic
[17]. The cap consists of molecules (in particular resonance imaging (MRI) are opening the way
collagen, elastin, and proteoglycans) produced for the characterization of the vulnerable
largely by the VSMC and protects the core of plaque [20].
the plaque from the direct contact with the blood Among the invasive methods of imaging, ves-
[16]. The necrotic lipid core is rich in sel angiography is still the gold standard method
prothrombotic molecules; after the fissuring of to evaluate the grade of occlusion of the vessel in
the plaque, the necrotic lipid core is exposed to some fields (e.g., the study of coronary arteries)
the blood flow, unleashing the coagulation cas- and allows even to operate directly at the level of
cade and the formation of thrombus; the thrombus the vessel-operating angioplasty, using stent in
can completely occlude the vessel determining order to make again patent the vessel or injecting
ischemia and then infarction of the tissues sup- thrombolytic drugs intra-arteriously in case of
plied by the vessels and/or can shatter and recent thrombosis. The intravascular ultrasound
embolize peripherally. A vulnerable atheroscle- (IVUS) is another promising technique that
rotic lesion could even fissure after an acute allows assessing the composition of the plaque
intra-plaque hemorrhage because of the rupture with a very small US probe placed at the tip of
of the thin-walled vasa vasorum [18]. an angiographic catheter [21].

Therapeutic Options
Imaging Evaluation
The most important concept to remember is that
Nowadays, we have many diagnostic imaging atherosclerosis is not a local disease, but it must be
methods to evaluate the gravity of the atheroscle- considered a systemic disease of the body; a local-
rotic disease (in particular trying to identify the ization of disease in the carotid arteries could be
vulnerable plaques, more prone to rupture) associated often with the presence of plaques at
suspected on the clinical and laboratory data. In the level of the coronary artery and vice versa.
this paragraph, we will briefly expose the main According to the localization, the characteris-
characteristics of these examinations. tics, and the dimensions of the atherosclerotic
Ultrasound (US) usually is the first imaging plaques, we have different therapeutic options,
technique used; it is safe, repeatable, and very pre- surgical (open surgery or endovascular treat-
cise in particular for the evaluation of the superficial ments) and nonsurgical.
vessels, such as supra-aortic trunks located in the Among the nonsurgical options, the most
neck, and the arteries of the upper and lower limbs. important are the removal and the treatment of
Multidetector computed tomography (MDCT) the modifiable risk factors; in particular, in case
is more accurate for the evaluation of atheroscle- of high levels of LDL, the use of statins could be
rotic plaque; it uses ionizing radiation, and an indicated. These drugs block the cholesterol syn-
intravenous iodinate contrast medium administra- thesis acting on the 3-hydroxy-3-methylglutaryl
tion is required in order to enhance the vessels and coenzyme A reductase at the hepatic level, but it
to evaluate the composition of the plaque and the has also been evidenced that they act directly at
grade of stenosis of the vessel. For example, the the level of the plaque inhibiting some inflamma-
evaluation of coronary arteries with a coronary tory processes, diminishing the recruitment of
computed tomography angiography (CTA) is not inflammatory cells inside the lesion, and increas-
an invasive method for the study of patients with ing the production of NO [22].
low to intermediate risk of coronary artery disease The surgical options are reserved for those
and allows not only to evaluate the grade of cases in which the other therapies have been
38 M. Porcu et al.

resulted insufficient; it is not the goal of this 9. Sima AV, Stancu CS, Simionescu M (2009) Vascular
paragraph to expose in detail the huge number of endothelium in atherosclerosis. Cell Tissue Res
surgical procedures at our disposal, and we invite 10. Whayne TF Jr (2011) Atherosclerosis: current status of
the gentle readers to deepen in the next chapters of prevention and treatment. Int J Angiol 20(4):213–222.
this book and in other specialist texts. doi:10.1055/s-0031-1295520
11. Tsimikas S, Miller YI (2011) Oxidative modification of
lipoproteins: mechanisms, role in inflammation and
potential clinical applications in cardiovascular
Summary disease. Curr Pharm Des 17:27e37
12. Doran AC, Meller N, McNamara CA (2008) Role of
In these pages we have seen how much compli- smooth muscle cells in the initiation and early progres-
sion of atherosclerosis. Arterioscler Thromb Vasc Biol
cated and variable is the pathogenesis of athero- 28:812e9
sclerosis, in which a central role is played by the 13. Heissig B, Hattori K, Friedrich M, Rafii S, Werb Z
endothelium damage and its subsequent dysfunc- (2003) Angiogenesis: vascular remodeling of the extra-
tion. Many factors, modifiable or not, influence its cellular matrix involves metalloproteinases. Curr Opin
Hematol 10:136e41
evolution. The imaging techniques are fundamen- 14. Visse R, Nagase H (2003) Matrix metalloproteinases
tal for the evaluation of the plaque and of the and tissue inhibitors of metalloproteinases:
vessel stenosis, leading the clinicians and the sur- structure, function, and biochemistry. Circ Res
geons to the correct therapeutic approach for the 92:827e39
15. Dhore CR, Cleutjens JP, Lutgens E et al (2001) Differ-
patient. ential expression of bone matrix regulatory proteins in
human atherosclerotic plaques. Arterioscler Thromb
Vasc Biol 21:1998e2003
References 16. Mishra TK et al (2013) An approach to the classifica-
tion, diagnosis and management of vulnerable plaque.
1. Douglas G et al (2010) The pathogenesis of atheroscle- J Indian College Cardiol 3:57–66
rosis. Medicine 42:9; 38(8):397–402 17. Virmani R, Burke AP, Kolodgie FD, Farb A (2002)
2. Tao W et al (2012) Atherosclerosis: pathogenesis and Vulnerable plaque: the pathology of unstable coronary
pathology. Diagn Histopathol 18(11):461–492 lesions. J Interv Cardiol 15:439e446
3. Balboni GC et al (2004) Anatomia Umana (terza edizione, 18. Depre C, Havaux X, Wijns W (1996) Neovascu-
tomo I, chapter 3). Edi.Ermes s.r.l., Milano larization in human coronary atherosclerotic lesions.
4. Rindi G et al (2005) Fisiologia Umana (IX edizione). Cathet Cardiovasc Diagn 39:215e220
UTET S.p.A. Divisione scienze mediche, Corso 19. Sun Z et al (2012) Coronary CT angiography: current
Raffaello, Torino status and continuing challenges. Br J Radiol
5. Hopkins PN (2013) Molecular biology of atheroscle- 85:495–510
rosis. Physiol Rev 93:1317–1542. doi:10.1152/ 20. Yang J et al (2013) MR and targeted molecular MRI of
physrev.00004.2012 vulnerable plaques. Interv Neurol 1(3–4):124–131.
6. Ignarro LJ et al (1987) Endothelium-derived relaxing doi:10.1159/000346767
factor produced and released from artery and vein is nitric 21. Kovárník T et al (2014) Current status of intravascular
oxide. Proc Natl Acad Sci U S A 84(24):9265–9269 ultrasound in interventional cardiology. Vnitr Lek
7. Vanhoutte PM (2009) Endothelial dysfunction; the first 60(12):1062–1067
step toward coronary arteriosclerosis. Circ J 73:595e601 22. Biasucci LM, Biasillo G, Stefanelli A (2010) Inflam-
8. Deanfield JE, Halcox JP, Rabelink TJ (2007) Endothe- matory markers, cholesterol and statins: pathophysio-
lial function and dysfunction: testing and clinical rele- logical role and clinical importance. Clin Chem Lab
vance. Circulation 115:1285e95 Med 48:1685e91
Nonatherosclerotic Vascular Disease:
Biological and Pathological Basis 3
Pierleone Lucatelli, Beatrice Sacconi, and Carlo Catalano

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Atherosclerosis is the main cause of vascular
disease in most cases; in about 10 % of cases,
Takayasu Arteritis (TA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
carotid artery disease is related to
Giant Cell Arteritis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 nonatherosclerotic causes, including several
Fibromuscular Dysplasia (FMD) . . . . . . . . . . . . . . . . . . . 42 unfrequent pathologies such as Takayasu arter-
itis, giant cell arteritis, fibromuscular disease,
Moyamoya Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
moyamoya syndrome, arterial dissection and
Extracranial Internal Carotid Artery extracranial carotid aneurysm. These entities
Aneurysms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
are discussed in the present chapter, with a
Craniocervical Arterial Dissection . . . . . . . . . . . . . . . . . 44 special focus on pathogenesis. Indeed, the
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 aetiology of these diseases is in most cases
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
not completely known, since related to several
factors (genetic, immune and infectious). Early
diagnosis, usually leading to a good patient’s
outcome, is usually achieved after clinical
examination and imaging tests.

Arterial dissection • Carotid artery disease •
Extracranial carotid aneurysm • Fibromuscular
dysplasia • Giant cell arteritis • Moyamoya
syndrome • Nonatherosclerotic disease •
Takayasu arteritis


P. Lucatelli (*) • B. Sacconi • C. Catalano Atherosclerosis is identified as the main cause of

Department of Radiological, Oncological and Anatomo- vascular disease in around 90 % of cases; in the
Pathological Sciences, Sapienza University of Rome,
remaining 10 % of cases, nonatherosclerotic
Rome, Italy
e-mail:; beatrice. causes include several less common entities,; such as Takayasu arteritis, giant cell arteritis,
# Springer Science+Business Media New York 2016 39
L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_49
40 P. Lucatelli et al.

fibromuscular disease, moyamoya syndrome, mainly involves the media and adventitia, usually
arterial dissection, and extracranial carotid aneu- characterized by three stages: [5, 6].
rysm. The pathogenesis of these diseases is in
most cases unclear or even related to several fac- 1. A systemic stage, characterized by signs and
tors, such as genetic, immune, and infectious, symptoms of an acute inflammatory condition,
often associated with triggering events. The cor- such as fever, arthralgia, anemia, and increased
rect diagnosis is currently achieved after clinical erythrocyte sedimentation rate
examination and imaging tests; although US, CT, 2. A vascular inflammatory stage, when vascular
and MRA are useful, catheter angiography repre- stenosis and less frequently aneurysms occur,
sents the gold standard in diagnosing most of with corresponding signs and symptoms
these diseases. The majority of patients have a (stroke, transitory ischemic attacks, hypoten-
good outcome if the specific disease is diagnosed sive ischemic retinopathy with visual symp-
early. toms, vertebrobasilar ischemia, hypertensive
3. A burned-out stage, when fibrosis sets, usually
Takayasu Arteritis (TA) associated with remission According to the
American College of Rheumatology (ACR)
Takayasu arteritis (TA) is a granulomatous arteri- [7], the criteria for assessing the diagnosis are:
tis affecting the aorta and its branches [1]. The first
case was reported by Takayasu in 1905 [2]; lately, • Age at disease onset <40 years
the disease was more comprehensively described • Claudication of extremities
as “pulseless disease” by Shimizu and Sano in • Decreased brachial artery pulse
1951 [3]. It generally affects women in the first • BP difference >10 mmHg
fourth decades of life (nine females/one male), • Bruit over subclavian arteries or aorta
with a general incidence of 2.6 cases per million • Arteriogram abnormality, represented by arte-
per year in the USA and major prevalence in riographic narrowing or occlusion of the entire
patients of Asiatic origin [4]. aorta, its primary branches, or large arteries in
The etiology of TA is still unclear; the under- the proximal upper or lower extremities, not
lying pathologic process is inflammatory, with due to arteriosclerosis, fibromuscular dyspla-
several etiologic factors, either infective or auto- sia, or similar causes (changes usually focal or
immune, having been proposed. The most likely segmental)
hypothesis is that an unknown stimulus triggers
the expression of the 65 kDa heat-shock protein in A patient shall be said to have TA if at least
the aortic tissue which induces the major histo- three of these six criteria are present.
compatibility class I chain-related A (MICA) on According to the classification proposed by
vascular cells. The T cells and NK cells recognize Hata et al. [8], TA can be divided into six types
MICA on vascular smooth muscle cells and based on angiographic involvement:
release perforin, resulting in acute vascular
inflammation; pro-inflammatory cytokines are • Type I – branches of the aortic arch
also released and increase the recruitment of • Type IIa – ascending aorta, aortic arch, and its
mononuclear cells within the vascular wall. Th1 branches
lymphocytes drive the formation of giant cells • Type IIb – Type IIa region plus thoracic
through the production of interferon-γ and acti- descending aorta
vate macrophages with release of VEGF resulting • Type III – thoracic descending aorta, abdomi-
in increased neovascularization and PDGF, nal aorta, renal arteries, or a combination
resulting in smooth muscle migration and intimal • Type IV – abdominal aorta, renal arteries, or both
proliferation. The inflammatory cellular infiltrate • Type V – entire aorta and its branches
3 Nonatherosclerotic Vascular Disease: Biological and Pathological Basis 41

Hence, only types I, IIa, IIb, and V usually leading to vascular complications. In the majority
involve the carotid arteries, whereas the other of patients, a systemic inflammatory syndrome is
types more commonly involve the thoracic also present. The systemic and the vascular com-
descending and abdominal aorta and the renal ponents of giant cell arteritis seem to have differ-
arteries. ent underlying pathogenetic mechanisms: the
Angiography is the gold standard for diagno- vascular inflammation results from abnormal
sis, even though color Doppler imaging can be adaptive immune responses, whereas the systemic
useful in the first part of the diagnostic process, inflammation likely depends on an excessively
showing the mural thickening of the common activated innate immune system [13]. Despite
carotid arteries (hypoechoic in the early stages increased understanding of the inflammatory cas-
and then hyperechoic after the development of cade responsible for the disease process, the initial
fibrosis). CT, and especially MRA due to the event triggering the cascade remains uncertain;
young age of these patients, is also extremely the adventitia is more likely considered as the
helpful, even more during the follow-up [9]. If site of initial immunologic injury [14].
possible, whole-body MRI technique should be According to the American College of Rheu-
employed in order to exclude other localizations matology [15], the criteria for assessing the diag-
of disease [10]. nosis are:
TA is a chronic relapsing and remitting disor-
der. The overall 10-year survival rate is approxi- • Age at disease onset >=50 years
mately 90 %, this rate being reduced in case of • New headache
major complications, such as stroke, intracranial • Temporal artery abnormality (temporal artery
hemorrhage, and graft stenosis and/or occlusion. tenderness to palpation or decreased pulsation,
For this reason, strict management of traditional unrelated to arteriosclerosis of cervical
cardiovascular risk factors is mandatory in order arteries)
to minimize secondary cardiovascular complica- • Elevated erythrocyte sedimentation rate
tions. Approximately 20 % of patients have a • Abnormal artery biopsy (biopsy specimen with
monophasic and self-limited disease; in the artery showing vasculitis characterized by a
remaining 80 % of patients, TA requires immuno- predominance of mononuclear cell infiltration
suppressive treatment, resulting in remission in or granulomatous inflammation, usually with
around 60 % of cases [11]. multinucleated giant cells)

A patient shall be said to have GCA if at least

Giant Cell Arteritis three of these five criteria are present.
The disease usually has an acute or subacute
Giant cell arteritis (GCA), also known as temporal start, characterized by symptoms related to the
arteritis, is the most common form of vasculitis acute inflammatory status (fever, night sweats)
occurring in adults, with a prevalence of persons and headache, jaw pain, and blurred or double
with active or remitted GCA of 200 cases per vision; if the disease is not early diagnosed, com-
100,000 population aged 50 years or older. The plications may occur, such as blindness and less
female-to-male ratio is 3.7:1 [12]. It is a granulo- frequently stroke [9].
matous arteritis affecting large- or medium-sized The diagnostic process starts with clinical
vessels, usually the terminal branches of carotid examination and imaging studies. US findings
arteries (more frequently the temporal and oph- can show swelling of the vessel’s wall as a
thalmic arteries). The clinical manifestations of hypoechoic dark halo around the color-coded
giant cell arteritis result from two different pro- flow in the artery; the disease is segmental; there-
cesses. Inflammatory cells infiltrate the arterial fore, its visualization is suitable for localization of
wall and cause structural damage, eventually the biopsy [16]. CT and MR imaging can show an
42 P. Lucatelli et al.

abnormal wall enhancement of the affected arter- epidemiological evidence for the role of female
ies after contrast media injection in the acute hormones beyond the sex and age distribution of
phase, whereas vascular stenosis and aneurysm FMD has not been described yet. Although no
can be observed in the subacute-chronic phase. etiologic genes for FMD have been identified,
In the chronic phase, the differential diagnosis evidence supports a genetic basis for susceptibil-
between GCA and atherosclerotic disease may ity to FMD; a few authors reported the potential
not be easy, but the different localization of the inheritance pattern to be autosomal dominant
abnormalities (temporal and ophthalmic arteries with variable penetrance. However, several fac-
for GCA, ubiquitarious for atherosclerosis) can tors have limited the identification and character-
lead to the correct diagnosis. Standard test for ization of genes contributing to FMD, such as
definitive diagnosis is biopsy of the temporal disease rarity, variable phenotype, and gene-
artery, being more samples needed because the environment interactions [19]. FMD is usually
inflammation usually does not involve all parts described in terms of the affected arterial layer
of the artery [10]. and the composition of the lesions; depending
The prognosis for patients with untreated on the type of FMD, the narrowing (stenosis) of
GCA is extremely poor; these patients may suffer the artery is caused by an excess of either
blindness or death from myocardial infarction, the fibrous or muscular components of the arte-
stroke, or dissecting aortic aneurysm. On the rial wall [18].
contrary, with prompt and adequate therapy, full Many people with FMD do not have any symp-
recovery is the rule. Symptoms from temporal toms; symptoms can occur if the stenosis is severe
arteritis usually improve within days of treatment enough to restrict blood flow through the affected
with corticosteroids, except for those symptoms artery or if dissection occurs. Symptoms of FMD
related to an effective vision damage that in the carotid artery include headaches, ringing or
occurred before initiation of therapy, which are “swishing” noise in the ears, or light-headedness;
often irreversible. The mean duration of treat- advanced cases of FMD can cause stroke or a
ment is 2 years [17]. transient ischemic attack [9].
FMD is often accidentally diagnosed when the
beaded appearance in the arteries is observed dur-
Fibromuscular Dysplasia (FMD) ing examinations performed for other reasons.
Noninvasive imaging studies such as duplex ultra-
Fibromuscular dysplasia (FMD) is an angiopathy sound, MRA, and CTA can be used to confirm the
affecting medium-sized arteries, more frequently diagnosis of FMD and determine the extent of the
in young women of childbearing age. Among lesions. In general, angiographic studies are
patients with identified FMD, renal involvement performed only when the diagnosis is not clear
occurs in 60–75 % and cerebrovascular involve- or if the patient requires a therapeutic procedure
ment in 25–30 %; involvement of visceral arteries such as a balloon angioplasty [10].
and arteries of the limbs is less commonly The most common type of FMD is the medial
observed (about 9 % and 5 %, respectively); in fibrodysplasia (75–80 % of cases), affecting the
the case of cerebrovascular localization, the inter- tunica media; it is characterized by areas of vessel
nal carotid artery is more frequently affected stenosis alternating with areas of ectasia, resulting
(C2 segment) [9, 18]. in a classic “beads on a string” appearance on
The etiology of FMD is not known, even angiograms. Intimal and perimedial fibroplasias
though several factors have been considered as are less common (10 % of cases, respectively),
involved in its pathogenesis. More in detail, hor- caused by collagen deposits in the intima and in
monal factors such as estrogen have been pro- the outer portion of the tunica media; in the inti-
posed; however, although in the US Registry mal type, a concentric, smooth, narrowing (with-
91 % of registrants were female, clear supporting out beads) appearance of arteries can be observed,
3 Nonatherosclerotic Vascular Disease: Biological and Pathological Basis 43

whereas the perimedial type has a “beads on a According to the Classification of the Japanese
string” appearance, but with “beads showing a Health Ministry, there are four clinical forms of
smaller diameter in comparison to the normal the moyamoya disease: ischemic, hemorrhagic,
vessel” [20]. epileptic, and “other.” Clinical presentation in
When a patient is diagnosed with FMD in the children is usually ischemic, related to occlusion
carotid arteries, additional imaging studies may be of internal carotid artery or one of the branches of
obtained to evaluate the other blood vessels, espe- the Willis circle, resulting in paresis, sensory
cially vertebral arteries, Willis circle, and renal impairment, involuntary movements, headaches,
arteries. dizziness, or seizures; mental retardation is often
When FMD is present without any symptoms, present. In adults, the hemorrhagic form, espe-
it usually does not require intervention; risk fac- cially with subarachnoid hemorrhage, is more fre-
tors for vascular disease, such as high blood pres- quently observed as a result of hemorrhage of
sure, diabetes, and high cholesterol, should be fragile vessels [9, 26].
evaluated and treated, and imaging studies should In the affected cerebral vessels, pathological
be performed at regular intervals to evaluate the examinations do not show atherosclerotic or
disease progression. Angioplasty is recommended inflammatory lesions. Currently, the major pro-
for patients with FMD of the internal carotid teins believed to be involved in the pathogenesis
artery who experience TIAs or stroke related to are vascular endothelial growth factor (VEGF),
severe arterial narrowing. Stenting is rarely nec- basic fibroblast growth factor (bFGF), hepatocyte
essary only, in the case of carotid or vertebral growth factor (HGF), transforming growth factor-
artery dissection or carotid aneurysm. Surgery β (TGF-β), and granulocyte colony-stimulating
depends upon the location and the extent of dis- factor (G-CSF) [27]. The cause of stenosis is the
ease and consists in removing or bypassing the overgrowth of the smooth muscle layer, with
affected portion of the artery to restore normal thrombotic changes. The disease leads to different
blood flow; reconstructive surgery may be degrees of stenosis and occlusions of arteries of
recommended for patients with an aneurysm of the anterior part of the Willis circle and to the
the internal carotid or vertebral arteries [19, 21]. development of the collateral vasculature. The
vessels of the collateral circulation are formed as
a result of the widening of the existing vessels
Moyamoya Syndrome or development of new perforating arteries. There
are three main pathways of collateral circulation –
Moyamoya is a disease causing intimal thickening parenchymal, meningeal, and transdural. In the
of the walls and stenosis of the terminal branches vessels of the collateral circulation, there may
of the internal carotid arteries bilaterally; the term appear thrombotic changes, which are the cause
“moyamoya” (Japanese for “puff of smoke”) of ischemic symptoms. An increased blood flow
refers to the angiographic appearance of abnormal through thin collateral walls during stress, as well
vascular collateral networks that develop adjacent as the presence of microaneurysms, is the proba-
to the stenotic vessels [22, 23]. Moyamoya dis- ble cause of intracranial hemorrhages [28].
ease occurs primarily in Asians; the female-to- Before CTA or MRA were introduced to a wide
male ratio of moyamoya disease is 1.8:1. The clinical practice, the final diagnosis of the vascular
disease has two peaks of incidence in the first changes was based on cerebral angiography.
and fourth decades of life [8, 24]. The cause of Although CT examination is sufficient to diagnose
moyamoya disease is not clear. The disease is ischemic or hemorrhagic stroke in the course of the
believed to be hereditary; a few authors suggested disease, MRA can be considered as the first imaging
that the transmission may be autosomal dominant technique for definitive diagnosis, considering the
with incomplete penetrance based on age and young age of the patients [29]. To support the diag-
genomic imprinting factors [25]. nosis, the following findings should be observed:
44 P. Lucatelli et al.

• Stenosis or occlusion at the terminal portion of involved with the aneurysm and performing a
the internal carotid artery or the proximal por- bypass from the normal artery below the aneu-
tion of the anterior or middle cerebral arteries rysm to the normal artery above the aneurysm.
• Abnormal vascular networks in the vicinity of Stroke and cranial nerve injuries (particularly XII
the occlusive or stenotic areas pair) are potential risks for surgical treatment of
• Bilaterality of the findings (although some aneurysms with a diameter of more than 3 cm
patients may initially present with unilateral and/or extending to the skull base; endovascular
involvement) treatment may be preferred in appropriate cases as
an alternative to surgical therapy; since there is no
Mortality rates from moyamoya disease are risk of cranial nerve injury, it allows to treat
around 10 % in adults and 4 % in children. lesions that are hard to reach surgically, and
About 50–60 % of affected patients experience a there is no need for general anesthesia [31].
gradual deterioration of cognitive function, likely
due to recurrent strokes. The outcome of the dis-
ease depends on the severity of the hemorrhage, Craniocervical Arterial Dissection
the prognosis on recurrent attacks. Cases of mild
clinical course are normally treated conserva- Craniocervical artery dissection (CCAD) is one of
tively. In severe cases, surgery is indicated, the major causes of ischemic symptoms in young
including direct anastomoses, indirect procedures, adults. It has a prevalence of 2.6 cases per
and combined therapies [30]. 100, 000 persons per year; vertebral artery dissec-
tions are less common than carotid artery dissec-
tions. Genetic factors such as constitutional
Extracranial Internal Carotid Artery weakness of the arterial wall (such as in FMD or
Aneurysms in monogenic connective tissue disease, mainly
Ehlers-Danlos syndrome or Marfan’s syndrome)
Aneurysms of the extracranial internal carotid artery might have a role in the pathophysiology of
are extremely rare, being more commonly observed CCAD; environmental factors such as minor
in males (prevalence ranging from 0.1 % to 3.7 %); trauma act as a trigger [32]. Arterial dissections
most of them are of atherosclerotic origin. In case of begin with a tear in the intima or media resulting
nonatherosclerotic etiology, they can be congenital in bleeding in the arterial wall; expansion of the
(due to collagen disorders, FMD), infectious (TBC, wall by intramural blood causes compression and
HIV, mycotic), and posttraumatic. True aneurysms narrowing of the lumen, contributing to the for-
must be differentiated from pseudoaneurysms, mation of an intraluminal thrombus. The intramu-
which are usually iatrogenic (due to carotid punc- ral hematoma can create a false lumen that might
ture or endarterectomy). For a definitive diagnosis reconnect with the true lumen and forms parallel
of aneurysm, the diameter of the ectatic artery must flow; alternatively, wall rupture through the
increase of at least 50 %, if compared to the normal adventitia causes extraluminal bleeding. CCAD
vessel diameter [9, 10]. can be asymptomatic and discovered during rou-
Most of these aneurysms are asymptomatic, tine examinations. The most common symptom is
whereas TIA or stroke is the most common pain in the head and neck and in the region of
cause of hospital admission, related to emboliza- dissections, usually following a minor trauma. If
tion from the aneurysmal contents. Spontaneous the dissection compromises the arterial lumen or
rupture or bleeding is very rare, but fatal compli- causes thrombus formation, clinical symptoms are
cations are seen particularly in cases with mycotic related to ischemia; the closer the dissection to the
aneurysms [31]. brain is, the higher the possibility of brain infarc-
The mainstay of treatment of extracranial tion is present. On the contrary, if the dissection is
carotid artery aneurysms is surgical repair; it con- more extracranial, the probability of local symp-
sists in resecting the portion of the carotid artery toms from space-occupying lesions is higher;
3 Nonatherosclerotic Vascular Disease: Biological and Pathological Basis 45

bleeding in the subadventitial wall results in com- extracranial carotid aneurysm must be taken into
pression of the adjacent structures, such as lower account. These less common diseases have some
cranial nerves. Patients with subadventitial intra- similarities, such as an unclear pathogenesis and a
cranial dissections often present with subarach- good outcome if the specific disease is diagnosed
noid hemorrhage [32, 33]. early.
CCAD can be noninvasively diagnosed by
performing MRA and CTA [34]; color Doppler
flow imaging showed good results in visualization
of dissections, even though the main limitation is
represented by intracranial dissection (which is 1. Jennette JC (1994) Nomenclature of systemic vasculi-
the most common site of localization). The tradi- tis. Arthritis Rheum 37:187–192
tional method for CCAD diagnosis is catheter 2. Numano F et al (1996) Takayasu arteritis-five doctor in
angiography that may show irregular luminal the history of Takayasu arteritis. Int J Cardiol
narrowing occlusion, pseudoaneurysm, intimal 3. Shimizu K, Sano K (1951) Pulseless disease.
flap, double lumen, or distal branch J Neuropathol Clin Neurol 1.7–1.47
occlusion [35]. 4. Hall S et al (1985) Takayasu arteritis. A study of
There is no general consensus regarding opti- 32 North American patients. Medicine 64:89–99
5. Arnaud L, Haroche J, Mathian A (2011) Pathogenesis
mal management of internal carotid artery dissec- of Takayasu’s arteritis: a 2011 update. Autoimmun Rev
tion, but the choice among medical, endovascular, 11(1):61–67
and surgical options may depend on the type of 6. Matsunaga N, Hayashi K, Sakamoto I et al (1997)
and anatomic location of dissection; while surgery Takayasu arteritis: protean radiologic manifestations
and diagnosis. Radiographics 17:579–594
has a limited role, candidates for angioplasty and 7. Arend WP et al (1990) The American College of Rheu-
stent placement include patients with persistent matology 1990 criteria for the classification of
ischemic symptoms despite adequate Takayasu arteritis. Arthritis Rheum 33:1129–1134
anticoagulation, patients with a contraindication 8. Hata A, Noda M, Moriwaky R et al (1996) Angio-
graphic findings of Takayasu arteritis: new classifica-
to anticoagulant therapy, and patients with signif- tion. Int J Cardiol 54(suppl):S155–S163
icantly compromised cerebral blood flow [36, 37]. 9. Bartels E, Bartels S, Poppert H (eds) (2012) New trends
Overall, the prognosis for spontaneous internal in neurosonology and cerebral hemodynamics-an
carotid artery dissection is favorable, with about update. Perspect Med 1:244–249
10. Catalano C, Anzidei M, Napoli A (editors): (2013)
75 % of patients making a good recovery; the Cardiovascular CT and MR imaging. Springer-Verlag
reported mortality is less than 5 %. The functional Mailand Italia Srl; ISBN 978-88-470-2868-5.
outcome is also generally good, and recurrence of 11. Phillip R, Luqmani R (2008) Mortality in systemic
cerebral ischemia and carotid artery dissection is vasculitis: a systematic review. Clin Exp Rheumatol
26(5 Suppl 51):S94–S104
rare; the risk of recurrence is highest in the first 12. Salvarani C, Gabriel SE, O’Fallon WM, Hunder GG
month and then remains around 1 % per year (1995) The incidence of giant cell arteritis in Olmsted
for about a decade. Headache may persist for a County, Minnesota: apparent fluctuations in a cyclic
long time, in some cases for years, after the pattern. Ann Intern Med 123(3):192–194
13. Weyand CM, Goronzy JJ (2003) Giant-cell arteritis and
dissection [38]. polymyalgia rheumatica. Ann Intern Med
14. Weyand CM, Goronzy JJ (2000) Pathogenic principles
Summary in giant cell arteritis. Int J Cardiol 75(Suppl 1):S9–S15,
discussion S17–S9
15. Hunder GG, Bloch DA, Michel BA, Stevens MB,
Atherosclerosis is the main cause of vascular dis- Arend WP, Calabrese LH et al (1990) The American
ease, being responsible in around 90 % of cases of College of Rheumatology 1990 criteria for the classi-
carotid disease; in the remaining cases, fication of giant cell arteritis. Arthritis Rheum
nonatherosclerotic causes such as Takayasu arter- 16. Schmidt WA, Kraft HE, Vorpahl K (1997) Color
itis, giant cell arteritis, fibromuscular disease, duplex ultrasonography in the diagnosis of temporal
moyamoya syndrome, arterial dissection and arteritis. N Engl J Med 337:1336–1342
46 P. Lucatelli et al.

17. Borchers AT, Gershwin ME (2012) Giant cell arteritis: 28. Lim M, Cheshier S, Steinberg G (2006) New vessel
a review of classification, pathophysiology, formation in the central nervous system during tumor
geoepidemiology and treatment. Autoimmun Rev growth, vascular malformations, and Moyamoya. Curr
11(6–7):A544–A554 Neurovasc Res 3(3):237–245
18. Luscher TF, Lie JT, Stanson AW et al (1987) Arterial 29. Tarasów E, Kułakowska A, Łukasiewicz A et al (2011)
fibromuscular dysplasia. Mayo Clin Proc Moyamoya disease: diagnostic imaging. Pol J Radiol
62(10):931–952 76(1):73–79
19. Olin JW, Gornik HL, Bacharach JM et al (2014) 30. Burke GM, Burke AM, Sherma AK et al (2009)
Fibromuscular dysplasia: state of the science and crit- Moyamoya disease: a summary. Neurosurg Focus
ical unanswered questions: a scientific statement from 26(4):E11
the American Heart Association. Circulation 31. Yetkin U, Y€ urekli, G€
uz A (2006) Extracranial inter-
129(9):1048–1078 nal carotid artery aneurysms. Internet J Thorac
20. Begelman SM, Olin JW (2000) Fibromuscular dyspla- Cardiovasc Surg 10(2)
sia. Curr Opin Rheumatol 12(1):41–47 32. Debette S, Leys D (2009) Cervical artery dissections:
21. Olin JW, Pierce M (2008) Contemporary management predisposing factors, diagnosis and outcome. Lancet
of fibromuscular dysplasia. Curr Opin Cardiol Neurol 8:668–678
23(6):527–536 33. Caplan LR (2008) Dissections of brain-supplying
22. Kuroda S, Houkin K (2008) Moyamoya disease: cur- arteries. Nat Clin Pract Neurol 4:34–42
rent concepts and future perspectives. Lancet Neurol 34. Debette S, Provenzale JM (2009) MRI and MRA for
7(11):1056–1066 evaluation of dissection of craniocerebral arteries: les-
23. Scott R, Smith ER (2009) Moyamoya disease and sons from the medical literature. Emerg Radiol
moyamoya syndrome. N Engl J Med 360:1226–1237 16:185–193
24. Uchino K, Johnson A, Claiborne S, Tirschwell DL 35. Houser OW, Mokri B, Sundt TM et al (1984) Sponta-
(2005) Moyamoya disease in Washington State and neous cervical cephalic arterial dissection and its resid-
California. Neurology 65:956–958 uum: angiographic spectrum. Am J Neuroradiol
25. Mineharu Y, Takenaka K, Yamakawa H et al (2006) 5:27–34
Inheritance pattern of familial moyamoya disease: 36. Baumgartner RW (2010) Management of spontaneous
autosomal dominant mode and genomic imprinting. dissection of the cervical carotid artery. Acta Neurochir
J Neurol Neurosurg Psychiatry 77(9):1025–1029 Suppl 107:57–61
26. Shamim S, Kumar J, Jamalvi SW et al (2008) Moya 37. Xianjun H, Zhiming Z (2013) A systematic review of
Moya disease in a child. J Coll Physicians Surg Pak endovascular management of internal carotid artery
18(4):252–253 dissections. Interv Neurol 1(3–4):164–170
27. Weinberg DG, Arnaout OM, Rahme RJ et al (2011) 38. Schievink WI (2001) Spontaneous dissection of the
Moyamoya disease: a review of histopathology, bio- carotid and vertebral arteries. N Engl J Med
chemistry, and genetics. Neurosurg Focus 30(6):E20 344(12):898–906
Essentials of Transcranial Doppler
Ultrasound 4
Jonathan D. Kirsch

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Transcranial Doppler (TCD) ultrasound imag-
ing is a noninvasive, portable imaging tech-
Anatomy and Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
nique that can be used to evaluate the
Physics of Doppler Ultrasound . . . . . . . . . . . . . . . . . . . . . 49 intracerebral arteries. TCD ultrasound has
Applications of Transcranial Doppler become an important modality in the monitor-
Ultrasound . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 ing and evaluation of vasospasm of the intra-
Vasospasm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 cerebral arteries in patients with subarachnoid
Sickle Cell Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Brain Death . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
hemorrhage due to aneurysmal rupture.
Internal Carotid Artery Occlusion . . . . . . . . . . . . . . . . . . . . 58 Advances in transcranial ultrasound imaging
Other Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 utilizing gray-scale, color Doppler flow, and
Ultrasound Biosafety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 spectral Doppler allow for direct visualization
and flow velocity measurements within the
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
vessels in real time. This is a significant
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 improvement over the older “blind” technique
which utilized a range-gated Doppler probe
that inferred the vessels being interrogated by
a complex set of parameters including vessel
depth, flow direction, and probe orientation.
TCD has also been found to be useful in the
evaluation of patients with sickle cell disease
who may be at the risk of stroke and determi-
nation of brain death, internal carotid artery
(ICA) occlusion and collateral pathways, ste-
nosis, arteriovenous malformations, and intra-
cardiac right-to-left shunts.

J.D. Kirsch (*)
Transcranial Doppler (TCD) ultrasound has
Department of Diagnostic Radiology, Yale University
School of Medicine, New Haven, CT, USA emerged as an important technique for imaging
e-mail: of the intracranial vasculature. Originally
# Springer Science+Business Media New York 2016 47
L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_36
48 J.D. Kirsch

developed as an indirect “blind technique” involv- internal carotid arteries (ICA) which bifurcate
ing range-gated spectral Doppler, the technique into the middle and anterior cerebral arteries
has advanced in recent years with the ability to near the skull base (MCA, ACA). The anterior
image and visualize the vessels directly utilizing cerebral arteries communicate via the anterior
gray-scale, color Doppler, and spectral Doppler communicating artery (AComm). Posteriorly,
imaging. TCD has found its greatest application the vertebral arteries join to form the basilar
and utility in the neurosurgical intensive care unit artery which then divides into the posterior
in monitoring for vasospasm in patients with cerebral arteries (PCA). The posterior communi-
intracranial bleeds due to ruptured aneurysms. cating arteries complete the circle joining the
However, the technique has also found promising internal carotid artery system anteriorly with
and potential uses in the evaluation of stenosis, the posterior circulation.
internal carotid artery occlusion, brain death, Transcranial Doppler imaging can be utilized
intracranial arteriovenous malformations, intra- to visualize the circle of Willis, the ophthalmic
cardiac right-to-left shunts, and sickle cell artery, and the cephalad portions of the vertebral
disease [1]. and basilar arteries at the skull base.
Although the intracranial vessels can be Imaging of the circle of Willis can be achieved
imaged by other techniques such as conventional by transcranial Doppler imaging. Before the
angiography, CT angiography (CTA), or MR advent of high-resolution gray-scale and color
angiography (MRA), these techniques require Doppler flow imaging, Aaslid et al. demonstrated
transport of the patient to the diagnostic radiology that cerebral arterial flow velocities could be mea-
department. Conventional angiography is inva- sured using pulsed Doppler US through the tem-
sive in nature with the potential risk of vessel poral bone or the foramen magnum [2]. The
damage or stroke. All three techniques involve technique was “blind” in that the vessels them-
the use of contrast agents which carry the risk of selves were not directly visualized. The identity of
anaphylactoid reaction for MR and CT contrast the vessel being interrogated was determined by
agents, nephrotoxicity for CT contrast, and the depth of the vessel from the skull, its flow
nephrogenic systemic fibrosis with MR contrast. direction, and the transducer orientation in rela-
CTA and conventional angiography, in addition, tion to the skull [3]. Using this complex approach,
require exposure to ionizing radiation. Aaslid was able to measure and determine normal
Transcranial Doppler ultrasound, on the other mean flow velocity (MFV) ranges of the intracra-
hand, is a noninvasive imaging modality that can nial arteries (Table 1). This technique had obvious
be performed portably, if necessary, in the ICU limitations related to lack of visualization of the
setting. TCD does not require the use of contrast vessels and the inability to angle correct the mea-
or ionizing radiation and serial follow-up exami- surements of flow velocity (see section “Physics
nations can be easily performed. This chapter will of Doppler Ultrasound”).
review the relevant anatomy, physics, and tech- With advances in ultrasound equipment and
niques of transcranial Doppler ultrasound and the the utilization of gray-scale, color Doppler flow,
interpretation of the findings. and spectral Doppler imaging, direct visualization
of the vessel and more accurate flow velocity
measurements are now possible. Imaging of the
Anatomy and Technique circle of Willis is performed via a transtemporal
approach utilizing a 5–1 MHz sector-array trans-
The circle of Willis is an anastomotic ring of ducer with most of the imaging being performed
arteries that sits at the base of the brain encircling at the lower frequencies (1–2 MHz) to allow pen-
the sella turcica and pituitary gland. It provides etration of the temporal bone. The transducer is
communication between the internal carotid placed on the temporal bone either above the
arteries and the vertebrobasilar system. The ante- zygomatic arch and anterior to the external audi-
rior portion of the circle is comprised of the tory canal or slightly more cephalad and posterior
4 Essentials of Transcranial Doppler Ultrasound 49

Table 1 Parameters for evaluation of the intracranial vessels with range-gated spectral Doppler. Before the advent
of transcranial Doppler utilizing direct visualization of the vessels with gray-scale and color Doppler flow imaging, the
cerebral vessel being interrogated by spectral Doppler had to be inferred from multiple parameters as noted in the above
table. The vessel being measured was inferred from its depth from the skull, its flow direction with respect to the
transducer, and the transducer orientation to the vessel and skull. Mean flow velocity (MFV) ranges shown are those
determined by Aaslid et al. [2]
Flow Transducer
Artery Depth (mm) Window direction orientation MFV (cm/s)
Ophthalmic artery 40–50 Orbital Toward Slightly medial 16–26
Middle cerebral artery 35–60 Temporal Toward En face 46–86
Anterior cerebral artery 60–75 Temporal Away Anterior 41–76
Posterior cerebral artery 60–75 Temporal Toward Posterior 33–64
Vertebral artery 45–75 Transforaminal Away Superior and 27–55
Basilar artery 70–120 Transforaminal Away Superior 30–57

over the earlobe. Imaging is performed from the intracerebral vessels. High-frequency (5–10
left and right temporal bones with color Doppler MHz) linear, curved, or sector transducers can be
flow imaging utilized to visualize the vessels. used depending on the size of the child. Utilizing
Flow velocity measurements (both MFV and the anterior fontanelle, the ICA, MCA, ACA, and
peak systolic velocity) are obtained. With the anterior communicating artery (AComm) can be
transtemporal approach, visualization of the visualized in the coronal plane (Fig. 3). Midline
MCA, ACA, PCA, and terminal ICA is possible. sagittal imaging allows visualization and flow
At our institution, spectral Doppler waveforms velocity measurements from the ACA,
and flow velocities are obtained from the proxi- pericallosal, callosomarginal, and frontopolar
mal, mid, and distal MCA and single measure- arteries in addition to the vein of Galen (Fig. 4).
ments are obtained in the visualized portion of the Patency of the superior sagittal sinus can be dem-
ACA, PCA, and terminal ICA. onstrated through the anterior fontanelle or an
The basilar and distal vertebral arteries can be open sagittal suture in the coronal or sagittal
visualized via a transforaminal approach. The ver- planes (Fig. 5). The posterior circulation can be
tebral arteries enter the skull base through the imaged through a transforaminal approach, as in
foramen magnum and then join to form the basilar the adult, or through an open posterolateral fonta-
artery at the base of the medulla oblongata. The nelle just posterior to the mastoid process.
basilar artery courses cephalad, anterior to the
medulla and pons, in the prepontine cistern. To
visualize these vessels utilizing the transforaminal Physics of Doppler Ultrasound
window, the transducer is placed in the midline
below the occiput and angled cephalad (Fig. 1). An understanding of the basic underlying physical
The ophthalmic artery and carotid siphon of principles related to spectral Doppler ultrasound is
the ICA can be visualized via the transorbital essential for the interpretation of transcranial
window. The probe is lightly placed on the closed Doppler measurements and results. All vascular
eyelid with a small amount of gel applied to the studies utilizing spectral Doppler make use of two
closed eyelid to aid in acoustic transmission well-known physical principles when measuring
(Fig. 2). flow velocities and evaluating for stenosis and/or
The open fontanelles in infants provide an vasospasm: the Doppler effect and Bernoulli’s
excellent acoustic window for visualizing the principle.
50 J.D. Kirsch

Fig. 1 Transtemporal and transforaminal acoustic win- Doppler flow imaging, the circle of Willis can be visualized
dows (a). The transducer is placed superior to the zygo- (b). The transforaminal approach allows visualization of
matic arch and anterior to the external auditory canal or the basilar and vertebral arteries (c). ACA anterior cerebral
more cephalad above the earlobe for the transtemporal artery, ICA terminal internal carotid artery, MCA middle
approach. For the transforaminal approach, the transducer cerebral artery, PCA posterior cerebral artery, PComm pos-
is placed in the midline below the occiput and angled terior communicating artery
cephalad. Utilizing the transtemporal approach and color

The Doppler effect, first described in 1842 by transducer at a certain frequency, fo, and is then
the physicist Christian Doppler, relates the change reflected back to the transducer from the moving
in frequency of a sound wave observed (or other red blood cells at a different frequency, fr. This
periodic event) if the source of the wave is moving frequency shift ( fd = fr  fo) and its relation to
relative to the observer. This effect is commonly velocity can be expressed by the following
seen in everyday life when an automobile with a equation:
siren passes by. The sound waves increase in
frequency as the car approaches the observer and c  fd

decreases in frequency as the car passes by 2 cos θ  fo
resulting in a frequency shift in the sound of the
siren heard by the observer. For flowing blood, a where v is the velocity of the flowing blood, c is
similar shift in frequency is noted when interro- the speed of the sound wave, and θ is the angle
gated by the sound beam from an ultrasound between the incident beam of the sound wave
transducer. The sound wave is emitted from the arising from the transducer and the direction of
4 Essentials of Transcranial Doppler Ultrasound 51

Fig. 2 Transorbital acoustic window: The transducer is carotid siphon can be visualized by color Doppler flow
placed lightly on the closed eyelid (a) and utilizing the imaging (b). OA ophthalmic artery (Image used with per-
globe as an acoustic window, the ophthalmic artery and mission from Kirsch et al. [1])

Fig. 3 Coronal color Doppler flow imaging through the anterior cerebral artery (ACA), M1 segments of the middle
circle of Willis (a, b), utilizing the anterior fontanelle as an cerebral arteries (MCA), and the intracranial internal
acoustic window, demonstrates the A1 segments of the carotid artery (ICA)
52 J.D. Kirsch

blood flow within the vessel (Fig. 6). From the its direction. A shift to a higher frequency is a
equation, it becomes apparent that the most accu- positive Doppler shift and indicates flow toward
rate flow velocities would be obtained with the the transducer. A shift to a lower frequency is a
transducer parallel to the direction of blood flow negative Doppler shift and indicates flow away
as θ = 0 and cos θ = 1. The greatest error in from the transducer. This frequency shift and
measurement and loss of signal would occur as directionality can be color-coded to indicate flow
the angle approached 90 and cos θ approaches direction. Conventionally with color Doppler
zero. It is recommended that velocity measure- flow imaging, red is used to indicate flow toward
ments be obtained with θ < 60 to obtain the the transducer and blue for away; however, this
most accurate measurements. color designation is arbitrary and can be switched
The Doppler equation allows one to not only so attention must be paid to the color bar on the
calculate the velocity of the flowing blood but also image. The color on the top of the bar is toward
the transducer and the color below is away
(Fig. 7).
Bernoulli’s principle, as applied to flowing liq-
uids, can be utilized for the diagnosis of areas of
stenosis and vasospasm. According to the princi-
ple, as fluid flows through conduits of differing
diameters, the velocity and pressure associated
with the fluid will change. As fluid flows from a
conduit (or blood vessel) of greater diameter to
one of smaller diameter, the flow velocity must
increase to allow the same volume of blood to
flow through the narrower area (this is analogous
to one putting one’s thumb over the opening of a
garden hose as water is flowing out resulting in a
high-velocity jet of water being formed).
Fig. 4 Sagittal midline color Doppler flow imaging Bernoulli’s principle and the law of conservation
through the anterior fontanelle demonstrates the of energy also state that the pressure in the region
pericallosal artery, the callosomarginal artery (CM), the
anterior cerebral artery (ACA), the vein of Galen, and the of increased velocity will decrease (Fig. 8). The
corpus callosum (CC) increase in velocity noted in areas of narrowing is

Fig. 5 Coronal gray-scale (a) and sagittal color Doppler flow imaging (b) through the anterior fontanelle demonstrate the
superior sagittal sinus (arrows)
4 Essentials of Transcranial Doppler Ultrasound 53

a mainstay for diagnosing stenosis or vasospasm

in blood vessels. As described above, these flow
velocities can be easily calculated by the
Doppler equation. As elucidated by Spencer and
Reid [4], there is a fairly linear relationship
between flow velocity and the degree of stenosis
(up to a critical stenosis at which point flow veloc-
ities decline due to extreme narrowing of the
vessel). In general, the higher the velocity of the
blood flow, the greater the degree of stenosis or
Spectral Doppler allows measurements of flow
velocity in a specified area of a vessel (the sample
volume) over time and displays these flow veloc-
ities as a time/velocity curve throughout the car-
Fig. 6 Doppler effect. The velocity of blood flow can be diac cycle. From this waveform, peak systolic
calculated utilizing the Doppler effect. The ultrasound trans-
ducer emits a sound wave pulse at a certain frequency, fo, velocity (PSV), the velocity of the flowing blood
which is reflected off the moving red blood cells back to the at the peak of systole on the curve, and
transducer at a different frequency, fr. This resulting fre- end-diastolic velocity (EDV), the velocity just
quency shift ( fd = fr  fo) can be used in the Doppler before the next systolic upstroke, can be mea-
equation to calculate the velocity of the flowing blood. The
angle, θ, is the angle between the incident sound wave beam sured. The spectral waveform also indicates direc-
and the direction of the flowing blood. The most accurate tionality of flow. Conventionally, a waveform
flow velocity measurements are obtained with θ < 60 above the baseline (positive) indicates flow
(Image used with permission from Kirsch et al. [1]) toward the transducer and a waveform below the
baseline (negative) indicates flow away. Again,
close attention should be paid to the scale as this
can be inverted on the screen. The spectral line
and the area underneath it (the spectral envelope)
can indicate the type of flow present. A narrow
spectral line with a clear envelope beneath it indi-
cates fairly uniform (laminar) flow with a narrow
range of velocities. Filling in of the spectral win-
dow indicates a broader range of flow velocities
within the sample volume consistent with turbu-
lent flow seen with stenosis or narrowing (Fig. 9).
The amount of flow during the different portions
of the cardiac cycle provides information as to
whether the vessel is feeding a high-resistance
system such as a muscle bed or extremity or a
low-resistance system such as an end organ. High-
Fig. 7 Color Doppler flow imaging allows one to visual- resistance waveforms will have little to no flow
ize vessels more easily and also encode flow direction by
during diastole, whereas low-resistance wave-
the color map. The color bar in the upper right-hand corner
(thick arrow) denotes flow direction. Although the color forms will have more abundant forward flow
designation is arbitrary and can be reversed, the colors on during diastole.
the top half of the bar codes are for flow toward the The resistive index (RI) is a measure of
transducer (red) and the colors on the bottom, away
the resistance to the flow of blood into an
(blue). In this example, the right MCA is flowing toward
the transducer and the left MCA and PCAs are organ and is calculated utilizing the following
flowing away formula:
54 J.D. Kirsch

Fig. 8 Bernoulli’s principle. As fluid (blood) flows the area of narrowing and its pressure will decrease. This
through conduits of differing diameters, the velocity and principle, along with the Doppler effect, allows for ultra-
pressure of the fluid will change. As blood flows from a sonographic detection of areas of vasospasm or stenosis
vessel of greater to narrowed diameter (either due to vaso- (Image used with permission from Kirsch et al. [1])
spasm or stenosis), the velocity of the blood will increase in

Mean flow velocity (MFV) can be calculated

from the spectral Doppler waveform and is essen-
tially the average flow velocity over the time/
velocity waveform. Thresholds for PSV and
MFV have been determined that correlate with
varying degrees of vasospasm (see section

Applications of Transcranial Doppler

Fig. 9 Spectral Doppler waveform. The spectral Doppler
waveform is a time/velocity curve that demonstrates the Vasospasm
blood flow velocities over time within a defined sample
window in the vessel being evaluated. The area below the Transcranial Doppler ultrasound has found its
curve or spectral line is called the spectral envelope or
greatest application in the neurosurgical intensive
window. With turbulent blood flow, this window will “fill
in” representing the myriad flow velocities associated with care unit screening patients for vasospasm sec-
turbulent flow (as seen in this example). The spectral ondary to subarachnoid hemorrhage (SAH) from
Doppler also can be used to measure peak systolic velocity aneurysmal rupture. Vasospasm is used to
(thick arrow) and end-diastolic velocity (thin arrow) from
describe both the clinical picture of delayed-
which the mean flow velocity and the resistive index can be
calculated (table to left of CDF image). A waveform above onset neurological deficits following SAH and
the baseline (positive) indicates flow toward the transducer the imaging findings of vessel narrowing. Vaso-
and a waveform below indicates flow away from the spasm from SAH has been associated with a
15–20 % risk of stroke or death [5]. About half
of those affected will demonstrate clinical signs
ðPSV  EDVÞ and symptoms. It has been reported that vaso-
RI ¼ spasm doubles the risk for mortality in SAH
In low-resistance waveforms, EDV will be fairly Vasospasm following SAH is generally absent
high as there is forward flow of blood into the in the first 72 h post bleed. Vasospasm generally
organ during diastole and the ratio will, therefore, occurs by day 3, peaking between days 6 and
be lower. In organs or systems that have a high 12, and then resolving 15–20 days after the
resistance to blood flow, EDV will fall toward bleed [10]. The clinical findings associated with
zero and the ratio will approach 1. vasospasm usually manifest themselves 4–5 days
4 Essentials of Transcranial Doppler Ultrasound 55

post hemorrhage. Insidious onset of confusion Table 2 MCA evaluation for vasospasm: Mean flow
and decreasing levels of consciousness are the velocity, peak systolic velocity, and Lindegaard ratio cut-
offs used at our institution for evaluation of varying
initial presenting signs which may progress to degrees of vasospasm in the middle cerebral arteries
focal neurologic defects, infarction, coma, and
Severity of MFV PSV Lindegaard
death. The severity and extent of vessel narrowing vasospasm (cm/s) (cm/s) ratio
has been seen to correlate with the clinical Mild 120–150 200–250 3.0–4.5
findings. vasospasm
The pathogenesis of cerebral vasospasm is Moderate 150–200 250–300 4.5–6.0
poorly understood but results primarily from vasospasm
prolonged smooth muscle contraction. Data sug- Severe >200 >300 >6.0
gests that calcium-dependent and calcium-
independent vasoconstriction is taking place.
Breakdown products from the blood in the sub-
arachnoid space may also be playing a direct or
indirect role. Other factors such as an imbalance Table 3 ACA and PCA evaluation for vasospasm:
between vasoconstrictors and vasodilators includ- Mean flow velocity and peak systolic velocity parameters
ing nitric oxide, prostaglandins, prostacyclin, free used at our institution for evaluation of the presence of
vasospasm in the anterior and posterior cerebral arteries.
radicals, and endothelin may also be involved in The Sloan ratio is calculated by dividing the MFV of the
the pathogenesis of vasospasm following ACA by the MFV of the distal extracranial ICA (NA not
SAH [11]. applicable)
Screening of patients at risk for vasospasm is MFV PSV Sloan
essential as patients cannot be placed prophylac- Vessel (cm/s) (cm/s) ratio
tically on treatment for potential vasospasm. Med- Anterior cerebral >80 >120 >4.0
ical treatment for vasospasm consists of “triple-H
Posterior cerebral >85 >120 NA
therapy”: induced hypertension, hemodilution,
and hypervolemia which are all aimed at improv-
ing cerebral perfusion. Calcium channel blockers,
such as nimodipine, have also been recommended the velocity, the greater is the degree of
but should be used cautiously to avoid the delete- vasospasm.
rious effects of hypotension [7]. Unfortunately, For intracranial vessels, cutoff values for PSV
these treatments carry their own inherent risks and MFV have been established that correlate
and potential complications including renewed with various degrees of vasospasm in the MCA
bleeding from the aneurysm site or into areas of and for the presence of vasospasm in the ACA or
infarction, increased cerebral edema, myocardial PCA (Tables 2 and 3). For the MCA, MFV in the
infarction, and congestive heart failure. If medical 120–150 cm/s range correlates with mild vaso-
therapy is ineffective, catheterization of the cere- spasm, whereas MFV in the 150–200 cm/s and
bral vessels can be performed with a vasodilating 200–250 cm/s ranges correlate with moderate and
agent injected. TCDI allows for quick and nonin- severe vasospasm, respectively [2, 4, 12].
vasive screening and monitoring of the patients in Torbey et al. have shown that the age of the
the ICU setting for vasospasm to determine if patient does play a role in the incidence and diag-
medical or interventional treatment is necessary. nosis of cerebral vasospasm. Older patients (>68
Vasospasm results in a narrowing of the vessel years) had a lower incidence of symptomatic cere-
lumen, and as noted above, the velocity of the bral vasospasm and a shorter window in which
blood flow within a vessel is inversely propor- vasospasm may occur (up to 10 days post bleed
tional to its diameter until a critical narrowing for older patients as compared to younger patients
(usually greater than 95 %) is reached at which that demonstrated symptomatic vasospasm as late
point blood flow through the extremely narrowed as 16 days following SAH). With age, cerebral
vessel actually decreases. In general, the higher blood flow velocities decrease and older patients
56 J.D. Kirsch

were seen to develop symptomatic vasospasm at more indicative of an etiology other than vaso-
lower MFV than younger patients [13]. spasm as the cause of the flow velocity elevation
Physiologic processes such as autoregulation [14, 15]. An accurate measurement of the extracra-
and hyperemia and medically induced conditions nial distal ICA MFV is important as slight varia-
including hypertension and hypervolemia may also tions in this measurement can result in over- or
result in elevated flow velocities within the intra- underestimating the presence of vasospasm. The
cranial vessels. To help distinguish between these velocity should be obtained as close to the skull
different etiologies for elevated flow velocity in the base as possible (Fig. 10).
MCA, Lindegaard et al. developed a ratio utilizing The Sloan ratio (MFV of the ACA divided by
the MFV of the MCA and the MFV of the distal the MFV of the extracranial ipsilateral ICA) can
extracranial ipsilateral ICA (VMCA/VICA). aid in the detection of vasospasm in the ACA. A
Lindegaard ratios in the range of 3.0–6.0 are indic- Sloan ratio greater than 4.0 is considered indica-
ative of mild to moderate vasospasm and ratios tive of vasospasm.
>6.0 are indicative of severe stenosis. Elevated Baseline and serial examinations are important
flow velocities with a Lindegaard ratio <3.0 is to obtain as elevations of flow velocities within a

Fig. 10 (a) Noncontrast CT demonstrates an aneurysmal respectively). The Lindegaard ratio, however, was mark-
SAH. Left frontal craniectomy has been performed to help edly elevated (11.5) consistent with severe vasospasm. (b)
relieve intracranial pressure (*). Transcranial Doppler 3D reconstruction from a CT angiography performed after
imaging (TCDI) demonstrates mild to moderately elevated the TCD revealed severe vasospasm of the left MCA
peak systolic velocity and mean flow velocities in the left (arrow, c) (Images used with permission from Kirsch
middle cerebral artery (280 cm/s and 150 cm/s, et al. [1])
4 Essentials of Transcranial Doppler Ultrasound 57

Fig. 11 (a) Initial transcranial Doppler of the right ante- showed a normal caliber vessel (arrow). TCDI obtained
rior cerebral artery (ACA) in a patient with SAH demon- 10 days later demonstrates the PSV and MFV to have more
strates peak systolic and mean flow velocities in the normal than doubled in the same vessel. (c) Repeat CT angiogram
range (67 cm/s and 32 cm/s, respectively). (b) 3D recon- now demonstrates vasospasm of the right ACA (d, arrow)
struction form a CT angiogram performed at this time (Images used with permission from Kirsch et al. [1])

vessel between subsequent exams may be indica- vertebral arteries and dividing this value into the
tive of vasospasm. Mean flow velocity increases highest MFV obtained in the basilar artery. A
of greater than 50 % or absolute increases in BA/VA MFV ratio of 2.0–3.0 plus a basilar mean
velocity of greater than 50 cm/s over a 24 h period flow velocity greater than 85 cm/s was found to be
may indicate vasospasm in the vessel (Fig. 11). indicative of mild to moderate vasospasm. A ratio
Through the transforaminal window, the distal >3.0 was found to be highly accurate in diagnosing
vertebral arteries (VA) and basilar artery (BA) can severe vasospasm [16].
be visualized and evaluated for vasospasm involv-
ing the posterior circulation. Sviri et al. demon-
strated good sensitivity and specificity utilizing a Sickle Cell Disease
ratio of the MFV of the extracranial vertebral artery
and the basilar artery in conjunction with MFV of Sickle cell disease is the result of a coding muta-
the basilar artery. The ratio was calculated using the tion in the beta chain gene for hemoglobin. The
average of the mean flow velocities of both mutation results in substituting thymine for
58 J.D. Kirsch

adenine, GAG to GTG, in the sixth codon. The subarachnoid hemorrhage are the chief causes of
resulting abnormal hemoglobin (HbS) under deoxy brain death in adults, whereas in children, abuse,
conditions forms polymers and demonstrates motor vehicle accidents, and asphyxia are more
increased viscosity and decreased solubility. This common etiologies [22, 23]. Clinically, the deter-
polymer formation in red blood cells (RBCs) mination of brain death includes documentation
which are homozygous for HbS results in the clas- of coma, absence of brain stem reflexes, and
sic sickled shape of the RBCs. Factors other than apnea. Confirmatory tests include cerebral angi-
local tissue hypoxia that can precipitate the sickling ography which should demonstrate no filling of
process include dehydration, infection, acidosis, the intracerebral vessels from their point of entry
and cold weather. Recurrent episodes of sickling into the skull, electroencephalography which
result in membrane damage and loss of the ability should demonstrate absence of electrical activity
of the RBCs to reattain their normal shape. These and lack of reactivity to intense somatosensory or
properties of HbS and the resulting physiological audiovisual stimuli, and cerebral scintigraphy
changes in RBCs result in the clinical and physical which demonstrate absence of radioisotope within
findings of the disease including hemolytic anemia, the brain [12]. TCDI has been found to have a
vaso-occlusive crisis, and multiple organ damage high sensitivity and specificity for the evaluation
due to microinfarcts [17]. of brain death and can be utilized to evaluate the
In children, the most common cerebrovascular intracranial vessels utilizing the transtemporal
manifestation of sickle cell disease is cerebral approach [24].
infarction due to occlusive vasculopathy. An Due to increased intracranial pressures,
approximately 11 % incidence of stroke is seen transcranial Doppler waveforms will demon-
in patients with sickle cell disease by the age of strate initially decreased forward diastolic flow.
20. Incidence of stroke is highest in the first As pressure continues to increase, reversed dia-
decade [18]. Fibrous proliferation of the intima, stolic flow will be noted resulting in a reverber-
usually within the intracranial ICA, proximal ating or oscillating flow pattern. Reversed
MCA, and ACA, leads to vessel narrowing, occlu- diastolic flow, however, is not pathognomonic
sion, stroke, and infarction. In stroke-free chil- for brain death as other entities such as mass-
dren, those with high cerebral blood flow occupying lesions, large strokes, edema, or her-
velocities in the distal ICA or proximal MCA niation may also result in increased intracranial
had a significantly increased risk of stroke [19]. pressure with resulting loss of diastolic flow.
Children with sickle cell disease and mean Small, low-velocity systolic peaks or spikes
flow velocities below 170 cm/s were considered may be noted (Fig. 12). Care must be taken in
at low risk for stroke in the 60 months post TCD. the interpretation of apparent lack or cessation of
Patients with MFV in the 171–199 cm/s range flow in the vessels as technical factors such as
were recommended to have more frequent TCD poor skull penetration by the ultrasound beam
testing. Adams et al. have shown that children may result in false-positive exams.
with mean flow velocities greater than 200 cm/s
have a significantly increased risk of stroke and
showed significant decrease in the risk of stroke Internal Carotid Artery Occlusion
when treated with blood transfusions that
decreased the HbS concentration to less than Atherosclerotic plaques consisting of necrotic
30 % of the total hemoglobin [20, 21]. cells, lipids, and cholesterol crystals can result in
stenosis of the internal carotid artery which, in
conjunction with superimposed thrombosis, can
Brain Death result in complete occlusion of the vessel. Emboli,
originating either from a cardiac source or ather-
Transcranial Doppler imaging can aid in the diag- oma from the aortic arch, are the other most fre-
nosis of brain death. Traumatic brain injury and quent cause of occlusion. The proximal 2 cm of
4 Essentials of Transcranial Doppler Ultrasound 59

Fig. 12 Brain death: (a–c) Transcranial Doppler demon- lack of intracranial arterial flow with an abrupt cutoff of
strates reversed diastolic flow (thin arrows) in the right and activity at the skull base because increased intracranial
left middle cerebral arteries and low-velocity, systolic pressure exceeds cerebral perfusion pressure. Note also
spikes in the anterior cerebral artery (thick arrow), wave- the “hot nose” sign which is related to the decreased flow
forms typical for brain death. (d) Noncontrast CT demon- within the internal carotid arteries compared to the
strates diffuse cerebral edema with loss of gray-white increased flow in the external carotid circulation (e)
matter differentiation. Nuclear scintigraphy demonstrates
60 J.D. Kirsch

the origin of the artery and the carotid siphon are stenoses greater than 50 % were elevations in flow
the two most common sites for occlusion. The velocities greater than 150 % at the site of stenosis
occlusion may be clinically silent due to a collat- compared to proximal or distal flow velocities in
eral pathway that forms from the external carotid the vessel or flow velocity ratios >2.5. Detection
artery to the ophthalmic artery which reverses its of turbulent blood flow, aliasing, visualization of
flow to feed the intracranial ICA/MCA. Duplex narrowing on color Doppler flow imaging, and
and transcranial Doppler imaging can demon- distal tardus-parvus waveforms have also aided
strate this collateral pathway. Ultrasonogra- in detection of stenosis [25, 26].
phically, the occluded internal carotid artery will Although not approved by the Food and Drug
be filled with echogenic material and will demon- Administration for clinical use in the United
strate absence of flow by color Doppler flow States, ultrasound contrast-enhancing agents
imaging and spectral Doppler. The common such as Levovist (Schering AG) can help aid in
carotid artery waveform, which has normally the detection of vessels and areas of stenoses
low resistance with forward diastolic flow, will [17]. Intracranial vessels may be difficult to see
frequently convert to a high-resistance waveform due to an insufficient temporal bone window, poor
due to the occlusion. If the collateral pathway insonation angle, or low flow velocities or vol-
develops, the common carotid artery may main- umes. Imaging after intravenous injection of
tain its normal low-resistance pattern. The ipsilat- Levovist, galactose-based microbubbles (<5 μm
eral external carotid artery, which normally has a in diameter) coupled with palmitic acid, has been
high-resistance waveform (little forward diastolic found to increase significantly the ability to detect
flow), will become “internalized” with its wave- vessels and obtain spectral Doppler waveforms
form changing from a high-resistance pattern to a compared to pre-contrast examinations. Detection
low-resistance pattern. The ipsilateral ophthalmic of collateral flow through the anterior and poste-
artery can be visualized via the transorbital rior communicating arteries in patients with high-
approach and will demonstrate flow reversal grade ICA stenosis or occlusion was also signifi-
away from the transducer and globe (Fig. 13). cantly improved with the use of an echo-contrast
Other collateral pathways are possible including agent [27, 28].
feeding of the anterior circulation from the poste- Transcranial power Doppler imaging has been
rior vertebrobasilar system via the posterior com- useful in the detection of intracranial aneurysm.
municating arteries or filling of the ipsilateral The detection of aneurysm by TCDI is dependent
MCA via the contralateral MCA and anterior com- on the size of the aneurysm, its location, and its
municating artery. morphology. Detection rates of 47–73 % have
been reported in the literature [29, 30]. Aneurysms
in the anterior communicating and basilar arteries
Other Applications had the highest detection rates (>50 %) as did
those greater than 10 mm in diameter. Aneurysms
TCDI has found a limited role in detection of in the posterior communicating and pericallosal
stenosis, aneurysms, arteriovenous malformations, arteries had the lowest detection rate. MCA and
and intracardiac right-to- left shunts. ICA aneurysms had detection rates of about 50 %.
For stenosis of intracranial vessels, TCDI has Morphologically, spherical and multi-loculated
been found to be useful but not definitive for the aneurysms were more easily detected than elon-
diagnosis. Confirmatory studies such as CTA or gated ones. The administration of IV echo-
conventional angiography are often necessary. An contrast agent significantly increased the detec-
elevation in flow velocity of 50–150 % (ratio of tion rate for aneurysms [31]. TCDI has also
1.5–2.5) at the site of stenosis as compared to the shown promise in the long-term surveillance of
flow velocity just proximal or distal to the stenosis aneurysms treated by detachable coils. High sen-
has been used by Roubec et al. as criteria for the sitivity and specificity was seen in the ability of
detection of stenoses less than 50 %. Criteria for TCDI to detect and distinguish aneurysms that
4 Essentials of Transcranial Doppler Ultrasound 61

Fig. 13 (continued)
62 J.D. Kirsch

Fig. 13 (continued)
4 Essentials of Transcranial Doppler Ultrasound 63

Fig. 13 Internal carotid artery (ICA) occlusion: Gray- The ipsilateral external carotid artery normally has a high-
scale ultrasound demonstrates echogenic material filling resistance waveform with little diastolic flow but has
the lumen of the right internal carotid artery (a, arrows). become “internalized” and demonstrates increased forward
Power and spectral Doppler demonstrates no flow within diastolic flow (arrow, e). The right ophthalmic demon-
the vessel consistent with occlusion (b, c). The common strates flow reversal (f) away from the transducer (wave-
carotid artery (CCA) has maintained its normal form below baseline), away from the globe, and toward the
low-resistance (d) pattern as a collateral pathway has brain. (g) Normal waveform and flow direction in the left
formed to the brain via the ipsilateral ophthalmic artery. ophthalmic artery for comparison (OA ophthalmic artery)

had been successfully treated demonstrating com-

plete occlusion from those with significant resid- Ultrasound Biosafety
ual flow [32].
In patients with patent foramen ovale (PFO) Although generally considered a safe imaging
and right-to-left intracardiac shunting, TCDI has modality, diagnostic ultrasound does carry poten-
shown high sensitivity and specificity for tial biosafety concerns. As such, diagnostic ultra-
establishing the diagnosis. Patent foramen ovale, sound imaging should be performed only by
a remnant of the fetal circulation, is a flap-like personnel who have an understanding of the
opening in the atrial septum primum and potential biohazards related to ultrasound imaging
secundum that persists into adulthood. It is esti- and who are fully trained in its safe and
mated to be present in 20–25 % of the general proper use.
population and can cause paradoxical embolic Ultrasound waves are mechanical compression
events resulting in stroke, migraine or migraine- waves through a given medium. The waves prop-
like symptoms, systemic embolism, and neuro- agate in a longitudinal fashion with alternating
logic decompression sickness [33, 34]. The pres- areas of compression and expansion. The term
ence of a PFO and right-to-left shunt is usually “ultrasonic” refers to sound waves above the audi-
diagnosed by transesophageal echo (TEE) ble range and generally is considered to be sound
[35]. TCDI offers a less-invasive option for diag- waves above a frequency of 20,000 Hz (0.02
nosing PFO by intravenously injecting 1 cm3 of MHz). Medical ultrasound utilizes frequencies of
microbubbles and then evaluating for the presence sound generally in the range of 1–15 MHz (1–15
of bubbles in the MCA by TCDI via a  106 Hz).
transtemporal window indicating a right-to-left Ultrasonic sound waves and the resultant
shunt is present. The technique is showing prom- imaging results in the deposition of energy into
ise as being as accurate as TEE for the the body that can potentially cause damage to the
diagnosis [36]. tissues secondary to thermal and mechanical
64 J.D. Kirsch

(nonthermal) mechanisms. Attenuation of the heat from the bone to the soft tissues. Exposure to
ultrasonic sound wave as it passes through the the eye by ultrasound should also be kept to a
soft tissues or bone results in loss of energy due minimum due to low perfusion in the eye, espe-
to absorption and scatter. Absorption of the sound cially the lens, which results in decreased capabil-
wave results in the wave’s energy being converted ity for heat dissipation [41].
to heat. Spectral Doppler has a greater heating Two indices have been devised, the thermal
potential than other modes of ultrasound (gray- index (TI) and the mechanical index (MI) which,
scale and color Doppler flow imaging). Since through an on-screen display, give an indication to
sound waves are mechanical energy being com- the sonographer of the potential for ultrasound-
posed of alternating areas of compression and induced bio-effects.
expansion, nonthermal injury can also result Three types of thermal indices have been
from direct interaction of the sound wave with devised. The thermal index for the soft tissue
the soft tissue. Rarefaction from the passing ultra- (TIS) is utilized when the ultrasound beam only
sound wave can result in cavitation of the tissue. insonates the soft tissue; the TIB (thermal index
The presence of contrast agents containing for the bone) is used when the ultrasound beam
microbubbles can increase the probability of cav- insonates the bone at or near its focal zone; and the
itation and nonthermal effects [37, 38]. TIC (thermal index for cranial bone) is used for
To limit the potential for biosafety effects, the transcranial Doppler imaging when the probes are
key principle in the safe use of ultrasound includes adjacent to the bone (the skull) during scanning.
the ALARA (as low as reasonably achievable) The TI is an indicator of the relative potential
principle. The principle, as applied to medical for tissue temperature rise and biological effect. It
diagnostic ultrasound exams, consists of using is defined as the ratio of the emitted acoustic
the lowest output power and the shortest scan power to the power required to raise the tempera-
time reasonably possible to obtain the necessary ture of the tissue by 1  C [41]. The TI is intended
diagnostic information to formulate an accurate to give an approximate guide to the likely maxi-
diagnosis. It is acknowledged that it may be nec- mum temperature rise that might be incurred after
essary to use longer scan times or higher outputs if a long exposure to ultrasound. A greater TI repre-
that is what is required to obtain accurate diagnos- sents a higher heating potential and a higher risk
tic information [39]. for tissue injury. The MI is a rough guide as to the
As noted above, insonation of the soft tissue likelihood of cavitation occurring. In general, the
with ultrasound can result in a rise in temperature risk of cavitation increases with an increasing
of the tissue due to energy deposition causing MI. Various recommendations are present in the
potential thermal injury. The extent of damage literature for suggested thresholds for TI and MI
depends on the degree of temperature rise, the depending on the age of the patient, the body part
duration of exposure to the elevated temperature, being scanned, and the duration of exposure to the
the type of tissue exposed, its cellular proliferation ultrasound beam [39–41].
rate, and its potential for regeneration. Tempera-
tures of 39–43  C have been shown to cause
detrimental effects in vitro. Temperatures above Summary
44  C can result in denaturation of proteins
[40]. Self-heating of the ultrasound transducers Transcranial Doppler ultrasound is an extremely
(probes) can occur but is more an issue with useful imaging modality for evaluating the intra-
endo-cavitary transducers than for transducers cerebral arterial system. It has found its greatest
placed on the skin where heat dissipation is utility in evaluating and monitoring patients in the
greater. The presence of the bone (e.g., skull) in neurosurgical intensive care unit for vasospasm
the ultrasound beam increases the likelihood of following subarachnoid hemorrhage due to aneu-
temperature rise due to direct absorption of the rysmal rupture. The technology has also found
beam by the bone and subsequent conduction of clinical utility in evaluating and screening
4 Essentials of Transcranial Doppler Ultrasound 65

pediatric patients with sickle cell disease who may Council, American Heart Association. Stroke
be at risk of stroke. In conjunction with clinical 40(3):994–1025
10. Rasulo FA, De Peri E, Lavinio A (2008) Transcranial
findings, electroencephalography, and nuclear Doppler ultrasound in intensive care. Eur J
medicine imaging, TCDI has also been shown to Anaesthesiol Suppl 42:167–173
be a useful adjunct in determining brain death. 11. Kolias AG, Sen J, Belli A (2009) Pathogenesis of
Other possible uses for the imaging modality cerebral vasospasm following aneurysmal subarach-
noid hemorrhage: putative mechanisms and novel
include detection of stenosis, arteriovenous approaches. J Neurosci Res 87:1–11
malformations, determination of collateral path- 12. Wozniak MA, Sloan MA, Rothman MI et al (1996)
ways with occlusion of the extracranial internal Detection of vasospasm by transcranial
carotid artery, and right-to-left intracardiac shunts Doppler sonography: the challenges of the anterior
and posterior cerebral arteries. J Neuroimaging
due to patent foramen ovale. 6(2):87–93
Biosafety concerns are of paramount impor- 13. Torbey MT, Till-Karsten H, Bhardwaj A et al (2001)
tance. Although ultrasound is considered a very Effect of age on cerebral blood flow velocity and inci-
safe imaging modality, the practicing sonographer dence of vasospasm after aneurysmal subarachnoid
hemorrhage. Stroke 32:2005–2011
should be aware and cognizant of the potential 14. Lindegaard KF, Nornes H, Bakke SJ et al (1989) Cere-
injuries that could result to soft tissue and bone bral vasospasm diagnosis by means of angiography
from potential thermal and mechanical injury. and blood velocity measurements. Acta Neurochir
15. White H, Venkatesh B (2006) Intensive applications of
transcranial Doppler in the ICU: a review. Intensive
References Care Med 32(7):981–994
16. Sviri GE, Britz GW, Douville CM et al (2006)
1. Kirsch JD, Mathur M, Johnson MH et al (2013) Transcranial Doppler grading criteria for basilar artery
Advances in transcranial Doppler US: imaging ahead. vasospasm. Neurosurgery 59(2):360–366
Radiographics 33:E1–E14 17. Maakaon JE, Taher AT, Besa EC (2014) Sickle cell
2. Aaslid R, Markwalder TM, Nornes H (1982) Noninva- anemia. Medscape reference: emedicine.medscape.
sive transcranial Doppler ultrasound recording of flow com/article/205926-overview
velocity in basal cerebral arteries. J Neurosurg 18. Frempong KO (1991) Stroke in sickle cell disease:
57(6):769–774 demographic, clinical and therapeutic considerations.
3. Lupetin AR, Davis DA, Beckman I et al (1995) Semin Hematol 28:213–219
Transcranial Doppler sonography. Part 1. Principles, 19. Adams RJ, MCvie V, Nichols FT et al (1992) The use
technique, and normal appearances. Radiographics of transcranial ultrasonography to predict stroke in
15:179–191 sickle cell disease. NEJM 326:605–610
4. Spencer MP, Reid JM (1979) Quantitation of carotid 20. Adams RJ, McKie VC, Carl EM et al (1997) Long
stenosis with continuous-wave (C-W) Doppler ultra- term stroke risk in children with sickle cell disease
sound. Stroke 10(3):326–330 screened with transcranial Doppler. Ann Neurol
5. Marshall SA, Nyquist P, Ziai WC (2010) The role of 42:699–704
transcranial Doppler ultrasonography in the diagnosis 21. Adams RJ, McKie VC, Hsu L et al (1998) Prevention
and management of vasospasm after aneurysmal sub- of a first stroke by transfusions in children with sickle
arachnoid hemorrhage. Neurosurg Clin N Am cell anemia and abnormal results on transcranial Dopp-
21(2):291–303 ler ultrasonography. N Engl J Med 339:5–11
6. Keyrouz SG, Diringer MN (2007) Clinical review: 22. Wijdicks EFM (2001) The diagnosis of brain death.
prevention and therapy of vasospasm in subarachnoid N Engl J Med 344(16):1215–1221
hemorrhage. Crit Care 11(4):220 23. Ropper AH, Kehne SM, Wechsler L (1987)
7. Condette-Auliac S, Bracard S, Anxionnat E et al (2001) Transcranial Doppler in brain death. Neurology
Vasospasm after subarachnoid hemorrhage: interest in 37:1733–1735
diffusion-weighted MR imaging. Stroke 24. Petty GW, Mohr JP, Pedley TA et al (1990) The role of
32:1818–1824 transcranial Doppler in confirming brain death: sensi-
8. Awasthi D (1997) Cerebral Vasospasm: current think- tivity, specificity, and suggestions for performance and
ing and future trends. interpretation. Neurology 40:300–303
rosurgery/nervecenter/spasm.html 25. Roubec M, Kuliha M, Jonszta T et al (2011) Detection
9. Bederson JB, Connolly ES Jr, Batjer HH et al (2009) of intracranial arterial stenosis using transcranial color-
Guidelines for the management of aneurysmal sub- coded duplex sonography, computed tomographic
arachnoid hemorrhage: a statement for healthcare pro- angiography, and digital subtraction angiography.
fessionals from a special writing group of the Stroke J Ultrasound Med 30:1069–1075
66 J.D. Kirsch

26. Feldman E, Wilterdink JL, Kosinski A et al (2007) The 34. Schwerzman M, Nedeltchev K et al (2005) Prevalence
stroke outcomes and neuroimaging on intracranial in size of directly detected patent foramen ovale in
atherosclerosis trial. Neurology 68:2099–2116 migraine with aura. Neurology 65:1415–1418
27. Hansberg T, Wong KS, Droste DW et al (2002) Effects 35. Schneider B, Zienkiewicz T, Jansen V et al (1996)
of the ultrasound contrast-enhancing agent Levovist on Diagnosis of patent foramen ovale by transesophageal
the detection of intracranial arteries and stenoses in echocardiography and correlation with autopsy find-
Chinese by transcranial Doppler ultrasound. ings. Am J Cardiol 77:1202–1209
Cerebrovasc Dis 14(2):105–108 36. Jauss M, Kaps M, Keberle M et al (1994) A compari-
28. Droste DW, Jurgens R, Weber S et al (2000) Benefit of son of transesophageal echocardiography and
echocontrast-enhanced transcranial color-coded transcranial Doppler sonography with contrast medium
duplex ultrasound in the assessment of intracranial for detection of patent foramen ovale. Stroke
collateral pathways. Stroke 31(4):920–923 25:1265–1267
29. Becker G, Greiner K, Kaune B et al (1991) Diagnosis 37. Miller DL, Averkiou MA, Brayman AA et al (2008)
and monitoring of subarachnoid hemorrhage by Bioeffects considerations for diagnostic ultrasound
transcranial color-coded real-time sonography. Neuro- contrast agents. J Ultrasound Med 27(special
surgery 28:814–820 issue):611–632
30. Klotzsch Z, Nahser HC, Fischer B et al (1996) Visual- 38. Bioeffects Committee of the American Institute of
ization of intracranial aneurysms by transcranial Ultrasound in Medicine (2008) American Institute of
duplex sonography. Neuroradiology 38:555–559 Ultrasound in Medicine consensus report on potential
31. Turner CL, Kirkpatrick PJ (2000) Detection of intra- bioeffects of diagnostic ultrasound. J Ultrasound Med
cranial aneurysms with unenhanced and echo contrast 27:503–515
enhanced transcranial power Doppler. J Neurol 39. The British Medical Ultrasound Society (2009) Guide-
Neurosurg Psychiatry 68:489–495 lines for the safe use of diagnostic medical ultrasound
32. Turner CL, Higgins JN, Kirkpatrick PJ (2003) Assess- equipment
ment of transcranial color-coded duplex sonography 40. O’Brien WD, Deng CX, Harris GR et al (2008) The
for the surveillance of intracranial aneurysms treated risk of exposure to diagnostic ultrasound in postnatal
with Guglielmi detachable coils. Neurosurgery subjects: thermal effects. J Ultrasound Med
53(4):866–871 27:517–535
33. Hagen PT, Scholz DG, Edwards WD (1984) Incidence 41. Nelson TR, Fowlkes B, Abramowicz JS et al (2009)
and size of patent foramen ovale during the first Ultrasound biosafety considerations for the practicing
10 decades of life: an autopsy study of 965 normal sonographer and sonologist. J Ultrasound Med
hearts. Mayo Clin Proc 59:17–20 28:139–150
DSA Imaging Protocol and Technique
Eytan Raz

Contents Abstract
Preprocedural Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . 68 Currently, digital subtraction cervico-cerebral
angiography (DSA) is seldom the first-line
Patient Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
study to evaluate the vasculature of the neck
Tray . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 or of the head and is usually performed after
Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 duplex ultrasound, CT angiography, or MR
angiography, in cases where additional details
Vascular Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
or assessment is needed. Many times, DSA is
Catheter and Wire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 obtained in the same setting just before the
Cervico-Cerebral Angiography . . . . . . . . . . . . . . . . . . . . . 70 interventional endovascular treatment.
Projections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Special Situations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Cerebral angiogram • DSA • catheter • wire •
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 vascular access

Obviously, this approach is justified when the

same information can be obtained with a nonin-
vasive approach, but it is the belief and the expe-
rience of the author that cervico-cerebral
angiography may add many details and informa-
tion compared to other techniques and has to be
kept in the armamentarium to use in all cases in
which vascular questions remains open after non-
invasive studies are performed. Multiple such
examples will be presented in the current chapter.
Another introductory point that needs to be
pointed out is that DSA is an operator-dependent
technique, and as such, a wide spectrum of quality
of exam and spectrum of complications incidence
E. Raz (*) will result: in good hands, a DSA can have as low
Neurointerventional Radiology Section, Department of
as 0.03 % incidence of stroke.
Radiology, NYU Langone Medical Center, New York,

# Springer Science+Business Media New York 2016 67

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_50
68 E. Raz

Preprocedural Assessment order to take the routine morning medications.

Once the patient arrives in the department, a
Visiting the patient before the angiogram is man- 20–22 G IV is placed by the nursing team,
datory. This evaluation should focus on the clini- connected to an intravenous line that will serve
cal history, previous surgical history, history of as a means to administer fluids and medications.
medications, and allergies. Many important This peripheral IV line is also a lifeline, in the
changes in the procedure are usually tailored event of an emergency. After the tray is set up, the
based on all the findings of the preprocedural patient is brought to the angiography suite, appro-
assessment. priately positioned on the table, and connected to
For elective angiograms, a basic thorough continuous EKG, oxygen saturation monitoring,
evaluation, other than medical clearance, includes and intermittent blood pressure monitoring. Pubic
blood work (CBC with differential, PT, PTT, hairs close to the groin are removed, after which
BMP, and, for female patients, hCG), urinalysis, the groins can be prepped and the patient draped.
chest x-ray, and EKG. The reason to obtain these The patient’s head is held with a tape and aligned
studies is to evaluate the general medical risk, to carefully along the long axis of the detector.
evaluate for renal insufficiency or renal failure
(worsened by use of iodine contrast), and to eval-
uate for ongoing infections and for bleeding diath-
esis. If the patient is pregnant – and the procedure Tray
can be postponed – that is usually the best option.
The authors perform angiograms on The basic tray should have the following items:
anticoagulated patients knowing that a micro-
puncture kit and a groin closure device will prob- • Femoral drape
ably be used. • Skin prep tray
During the preprocedural assessment, the • Needle box
neurointerventionalist will educate the patient • Sterile labels
about the exam. Focusing on the sensations that • Telfa
the patient will experience during the procedure • 18 G 22 G–25 G needles
will ultimately improve the collaboration and, • Disposable scalpel no. 11
consequently, improve the quality of the study • Five 12 cc syringes
and decrease the rate of complications. The most • Light handle cover
important points to discuss with the patient are: • 3-way and 1-way stopcocks
the sensations in the groin when the local anes- • Torque device
thetic is injected, the warming sensation when the • Small J wire 0.038
contrast is injected, and the possible dizziness and • Terumo 0.038/150 cm
appearance of flashes behind the eyes. Likewise, • Heparin 4,000 units/1 L bag
also exemplifying the usual commands that are • Lidocaine 1 %
given prior to the injection is needed. A written • 4 F or 5 F micropuncture kit
informed consent needs to be signed by the patient • 4 F or 5 F vertebral catheter
before the procedure.

Patient Preparation Procedure

In elective angiograms, the autors request the Before starting the procedure, a time-out should
patient to be NPO since the midnight before. be conducted before the incision. The time-out
The only exception is usually a sip of water in should involve the entire operative team, use
5 DSA Imaging Protocol and Technique 69

active communication, be briefly documented, continuous flushing technique is preferable; both

and use a checklist which should include: (1) cor- pros and cons can be found in other excellent
rect patient identity using at least name and date of reviews. Double flushing consists of aspiration
birth, (2) allergies, pregnancy, or any other patient of the contents of the catheter with one 10 ml
specific issue, and (3) agreement on the procedure syringe followed by partial aspiration and irriga-
to be done. tion with a second syringe, while continuous
flushing technique consists in connecting the cath-
eter hub with a Y-connector plus three-way stop-
cock connected to a heparinized pressurized
Vascular Access saline bag. On average, a catheter should be
flushed every 90–120 s and after each wire place-
A successful angiography begins with the groin ment. The tip of the syringe should be oriented
puncture. Before puncturing, localization of the inferiorly so that the air will rise upward toward
femoral head with x-ray is suggested, especially the plunger and away from the catheter. As a
for obese patients, in which the fat folds may syringe is connected to the catheter or stopcock,
masquerade the anatomic landmarks. The right it is best to allow a meniscus of blood to form by
common femoral artery, which is in the vast allowing slight backflow into the hub of the cath-
majority of cases the aimed vessel for the access, eter: this meniscus is then brought together
is in fact located in the medial upper quadrant of against the syringe meniscus in a meniscus-to-
the femoral head. Local anesthesia is performed meniscus technique which reduces the likelihood
using 1 % lidocaine using two different needles: of air being trapped between the tip of the syringe
first superficial injection of 2 cc using a 25 G and the hub. After the syringe has been removed
needle, wait 30 s and then injection of approxi- from the end of the catheter, the hub should be
mately 8 cc deep to the artery with a 22 G needle, flushed and the best way to do it is using a
after careful evaluation of no blood return to avoid so-called Jabbie in angio suite jargon, which
injection in the bloodstream. Because of its pH, as essentially is a syringe with a dilator attached to
the lidocaine is being injected, it can be painful. the tip. The Jabbie fits easily into the hub and
Prior to puncturing, a small skin incision (few allows a jet of flush to be injected for cleaning
millimeters long) is made using a no. 11 scalpel the hub.
parallel to the skin fold and dissected using a
hemostat in order to facilitate the following pas-
sage of needle and catheter and to prevent the
subcutaneous forming of hematoma, creating a Catheter and Wire
tract with the surface. For discussion of the vari-
ous techniques to obtain access in the femoral The kind of catheter chosen is a personal decision.
artery, please report to excellent discussions else- In the authors’ experience, they start each angio-
where such as in Harrigan and Deveikis book. gram with a vertebral catheter leaving the option
Once the access is obtained, two main options to exchange for another shape in case of difficult
exist: pushing the catheter over the wire without a access. The vertebral catheter is a simple curve
sheath in place, or placing a sheath in order to catheter, meaning that it has only a primary curve
facilitate the possible catheter exchange. Both and does not need to be formed. The catheter is
pros and cons can be detailed for either technique. advanced over the wire in the ascending aorta, and
The main advantage of not placing the sheath is the wire is then pulled out. In a standard angio-
that a smaller hole is present in the artery. gram, the first vessel to be catheterized is the right
The catheter is then advanced under fluoro- common carotid artery. In order to select this
scopic guidance in the descending aorta. Another vessel, once it is in the ascending aorta, the cath-
option to face at this point is if a double flushing or eter needs to be pulled and rotated
70 E. Raz

counterclockwise. Once in the ostium of the ves- Table 1 Flow rate for different vessels
sel, additional selective catheterization can be Injector
obtained getting a new roadmap, reinserting the settings
wire and readvancing the catheter over the wire. Vessel injected ml/s ml
Every time the wire is removed and the catheter is Aortic arch 20 25
in position for injection, a flow check injecting Common carotid artery (CCA) 8 12
1–2 cc of contrast under fluoroscopy should be Subclavian artery 6 15
performed in order to confirm the appropriate Internal carotid artery (ICA) 6 8
External carotid artery (ECA) 3 6
catheter positioning and exclude iatrogenic dis-
Vertebral artery 6 8
section during the catheter advancement.
3D angiogram in internal carotid artery 3 20
A bifurcation run in both frontal and lateral
projections may be obtained if necessary; other-
wise, a roadmap to access the ICA may be
enough. Advancing the catheter in the ICA with If the automatic injection is used, the following
attention to have the vertebral catheter curve par- are the volumes and rates used – keep in mind
allel to the bloodstream and not against the vessel that, in jargon, the flow rate followed by the total
wall is important. This is usually obtained by volume is indicated in the angio suite (e.g., 5 by
rotating the catheter slightly counterclockwise in 7 means 7 cc of contrast to be injected with a rate
the right ICA and clockwise in the left ICA. of 5 cc/s) (Table 1).
If difficulties are encountered to access a cer- Another feature to keep in mind is the rise: a
vical vessels, other techniques can be considered, “rate rise” in cerebral angiography refers to a
such as the breath-holding technique or exchang- progressive acceleration of contrast over the first
ing the catheter for a double curved one such as second of the injection. The linear rate rise feature
the Simmons – please consult other excellent is especially useful when a catheter is near the
resources to good use of these catheters and how origin of a vessel and a sudden injection of con-
to form and unform. Keep in mind that the more trast may result in recoil of the catheter proximal
complex the vasculature and the more catheters outside of the desired vessel. The PSI setting of
are used, the higher the incidence of the injector refers to the maximum pressure
complications. (pounds per square inch) that the injector will
In some cases, an aortogram is needed, for generate while injecting the contrast.
example, in case an arch study was not obtained
with MRA or CTA, in case of possible vasculitis
or in patients with trauma to exclude an aortic tear.
If the aortogram is needed, a groin sheath has to be Cervico-Cerebral Angiography
placed. The catheter used for aortogram is a pigtail
catheter which is placed in the proximal ascending Biplane angiography is nowadays the standard of
aorta, proximal to the innominate artery. care for cervico-cerebral angiograms and allows
The contrast is diluted in a 2:1 solution with for two planes to be acquired simultaneously with
heparinized saline. This ratio allows excellent a single contrast injection, limiting the total
opacification of the bloodstream with a relatively amount of time and contrast needed to complete
limited contrast load. The injection of contrast the study.
(iopromide, which is a nonionic agent with
300 mg of iodine/ml) is different to obtain
roadmap or to obtain angiographic images. For Projections
roadmaps smooth continuous injection is
performed, while for angiography, fast, strong Standard projections are illustrated in the figures;
injection with appropriate handling is needed. the x-ray detector has a limited space, and as such,
5 DSA Imaging Protocol and Technique 71

the space needs to be optimized. Standard Most cerebral angiographic runs are obtained
arch projection is a left anterior oblique that with a rate of two frames per second (FPS). A
opens the view of the aortic arch. To better variable frame rate can also be obtained limiting
visualize the origin of the vertebral artery, a the rate for venous phase. Usually, an imaging
Towne’s view is necessary, given the origin of sequence lasts approximately 10–12 s in order to
the artery in the superoposterior aspect of the visualize the arterial, capillary, and venous
subclavian artery. The Towne’s view is obtained phases.
angulating the detector 30–40 cranially aligning Modern angiography makes use of three-
the petrous ridges below the superior rim of the dimensional images in various occasions. In
orbits. order to obtain the source 3D images, only a
Standard neck arteries views are the straight single plane is needed, which is rotated around
frontal and lateral projections (Fig. 1). If a single the head of the patient before and during the
plane is used, an ipsilateral oblique frontal view is injection of contrast in a CT-like manner
preferable. (Fig. 6). The volumetric images thus obtained
Before obtaining the cranial images, careful can be postprocessed in different ways and
evaluation of the alignment of the head along the reconstructed in the preferred views (Fig. 7). An
long axis of the detector should be checked. In the obvious advantage of the 3D images is that they
standard frontal (PA) view of the internal carotid can further guide the best 2D projections useful
injection, the petrous ridges are aligned with the that can be selected based on the 3D reconstruc-
lower third of the orbits (Figs. 2a–e and 3a–f). In tion (Fig. 8).
the lateral projection, careful evaluation of the After the angiogram, careful evaluation of the
alignment of the orbital roof and of the external images obtained is done before pulling the cathe-
auditory canals is needed (Figs. 2f–j and 3g–j). ter out of the body. This suggestion has two
While for a CCA injection, the outer skull has to purposes: first, to do a thorough evaluation in
be included on the field of view, for an ICA case additional images are needed and, second,
injection the positioning of the detector has to be carefully evaluate that no evidence of complica-
adjusted just to cover the inner skull. For vertebral tions such as dissection or thromboembolism is
injection intracranial views, both a frontal and present. Once these things are ascertained,
Towne’s views are needed, while for the lateral the management of arteriotomy can start,
projection, the C1 vertebra coverage needs to be preferably accomplished with manual compres-
checked (Figs. 4 and 5). Special views are sion, with pressure applied approximately
obtained in particular situations: for example, to 1–2 cm above the skin incision for 15–20 min,
better see the ICA cavernous segment, the com- or longer if oozing is still noted after that time.
municating segment, or the middle cerebral artery In certain situations, a closure device can be
branches, a Haughton view may be obtained; in used for a mechanical closure and multiple
order to position for this view, the x-ray tube devices for this purpose are available in the
needs to be tilted caudally towards the shoulder market.
of the side of interest. To better visualize the ICA After the angiogram, they request that the
bifurcation or the anterior communicating artery patient lies flat on bed for 6 h with the accessed
complex, an ipsilateral transorbital oblique is leg extended and with head of bed at <30 . Vital
recommended. To memorize all the useful pro- signs need to be checked q1h. Nursing
jections for the different vessels can be cumber- team should check the groin puncture as
some; the best way to obtain the best views is to follows: q15min for 1 h, q30min for 2 h, and
create a 3D virtual map of the cerebral vasculature then q1h for 3 h. The patient may restart the
in thier mind and use this virtual map to guide the usual diet after the angiogram with encourage
best projection based on the orientation of the water intake. The IV can be discontinued before
vessel of interest. discharge.
72 E. Raz

Fig. 1 Standard angiographic views for the CCA bifurcation: lateral, subtracted and unsubtracted (a, b) and straight
frontal subtracted and unsubtracted (c, d)
5 DSA Imaging Protocol and Technique 73

Fig. 2 (continued)
74 E. Raz

Fig. 2 Right ICA injection. In the standard frontal view of late venous (e) views. (f–j) Lateral projection in early
the internal carotid injection, petrous ridges are aligned arterial – subtracted (f) and unsubtracted (g) –, capillary
with the lower third of the orbits. (a–e) Frontal projection (h), late capillary/early venous (i), and early venous (j)
of right ICA in early arterial – subtracted (a) and views
unsubtracted (b), – capillary (c), early venous (d), and

Special Situations For evaluation of intracranial aneurysms, a

complete four-vessel angiogram should be
For evaluation of atherosclerotic disease, an performed: in 15 % of patients with SAH, multi-
aortogram is suggested as well as routine neck ple aneurysms are present. If surgery is an option,
views of both common carotid and vertebral an external carotid injection is also performed in
arteries. order to have a map of the vessels for possible
5 DSA Imaging Protocol and Technique 75

Fig. 3 (continued)
76 E. Raz

Fig. 3 Left ICA injection. In the standard frontal view of early venous (e), and late venous (f) views. In the lateral
the internal carotid injection, petrous ridges are aligned projection (g–j) the alignment follows the orbital roof and
with the lower third of the orbits. (a–e) Frontal projection the external auditory canals: early arterial – subtracted (g)
of left ICA in early arterial – subtracted (a) and and unsubtracted (h) – capillary (i), and late venous (j)
unsubtracted (b) – early capillary (c), late capillary (d), views

intracranial arterial bypass (superficial temporal obtain the best view to show the neck aneurysm.
artery, internal maxillary artery, etc.). When an Sometimes, cross-compression may be necessary
aneurysm is identified, the mere diagnosis is not to visualize the anterior communicating artery by
enough: for thorough planning, the anatomy of the compressing the contralateral carotid and
aneurysm has to be defined, especially trying to allowing crossflow through the Acomm. In a
5 DSA Imaging Protocol and Technique 77

Fig. 4 (continued)
78 E. Raz

Fig. 4 Left vertebral artery injection. In the standard capillary (d), and late venous (e) views. In the lateral
frontal view of the internal carotid injection, petrous ridges projection (f–k), the alignment follows the orbital roof
are aligned with the lower third of the orbits. (a–e) Townes and the external auditory canals: early arterial – subtracted
projection of left vertebral artery in early arterial – (f) and unsubtracted (g) – late arterial (h), capillary (i),
subtracted (a) and unsubtracted (b) – late arterial (c), early venous (j), and late venous (k) views
5 DSA Imaging Protocol and Technique 79

Fig. 5 Additional vertebral artery injection views. Additional injection with frontal projection (a, b) and double oblique
(c) views

similar fashion, the carotid artery can be com- understand the subjacent anatomy. Angiograms of
pressed during vertebral artery injection to better other branches are often the most important runs
visualize the Pcomm. in cases of a high-flow AVM because partial
In cases of arteriovenous malformations, high- opacification is visualized with higher details
rate runs are obtained (3–5 FPS), which help to (Fig. 9). Another important point when imaging a
80 E. Raz

Fig. 6 Source images from 3D injection. These are the are similar to tomographic images obtained from a CTA,
source images acquired during injection (in this case, left and can be reconstructed in a similar fashion
ICA injection) that are used to create the 3D views. They
5 DSA Imaging Protocol and Technique 81

Fig. 7 (a–c) The volumetric images obtained can be postprocessed in different ways and reconstructed in the preferred
views using different color algorithms

patient with AVM is that, in order to see the normal also the ascending and deep cervical arteries may
brain venous drainage, a longer imaging sequence is need to be separately imaged according to the
necessary. In AVM ECA injection should be location.
obtained in order to ascertain ECA feeders. What to look for on an angiogram:
For evaluation of dural arteriovenous fistulas, The neuroradiologist should be familiar with
the ECA, some of its branches, more commonly the evaluation of angiographic images. Multiple
the ascending pharyngeal and the occipital, but notions should be kept in mind when looking at an
82 E. Raz

angiogram. Differently from a CTA, the contrast

is injected inside a single vessel and hemody-
namic consequences may be noted such as the
inflow of nonopacified blood from other
vessels (Fig. 10). The angiography is a dynamic
technique and all the phases have to be
evaluated, early arterial, late arterial, capillary,
early venous, and late venous. In order to
better understand the anatomy, or in order to iden-
tify some branches (especially the ECA
branches), the unsubtracted images are very help-
ful (Fig. 11).
In the arterial phase, the vessel contour and size
are evaluated. The arterial anatomy is visualized
with identification of anatomical variants or
nonvisualization of expected vessels; careful eval-
uation of early draining veins is accomplished
Fig. 8 After analysis of 3D images, the best view to
during this phase. On the capillary phase, the
evaluate the neck of the aneurysm is selected and the 2D
acquisition is then performed. Compare with Fig. 7a blush of physiologic or pathologic structures is

Fig. 9 (a, b) In cases of arteriovenous malformations, opacification is visualized with higher details. In this
angiograms of other branches are often the most important example of a right frontal AVM, better details of the
runs in cases of high-flow AVM because partial nidus are obtained during vertebral artery injection
5 DSA Imaging Protocol and Technique 83

identified: in order to better see it, hard Patency of the veins is visualized. Abnormal per-
windowing may be helpful. On the venous sistent arterial opacification in this phase should
phase, dominance of different veins is identified, prompt to think of possible thromboembolism.
which can help to better tailor an intervention.


Even though noninvasive methods to depict the

vessels of the neck and head are nowadays avail-
able, catheter-based angiography is still the gold
standard to visualize the vasculature. This proce-
dure needs to be performed following strict
criteria in order to decrease as much as possible
the incidence rate and increase as much as possi-
ble the diagnostic yield.

Fig. 10 Different from a CTA, the contrast is injected

inside a single vessel and hemodynamic consequences
may be noted such as the inflow of nonopacified blood
from other vessels

Fig. 11 (a, b) External carotid artery injection. In order to better understand the anatomy, or in order to identify some
branches (especially the ECA branches), the unsubtracted images are very helpful
84 E. Raz

References 4. Björkesten G, Halonen V (1965) Incidence of intracra-

nial vascular lesions in patients with subarachnoid
1. Fifi JT, Meyers PM, Lavine SD, Cox V, Silverberg L, hemorrhage investigated by four-vessel angiography.
Mangla S, Pile-Spellman J (2009) Complications of J Neurosurg 23(1):29–32
modern diagnostic cerebral angiography in an academic 5. Lin JP, Kricheff II (1972) Angiographic investigation
medical center. J Vasc Interv Radiol 20(4):442–447 of cerebral aneurysms. Technical aspects. Radiology
2. Thiex R, Norbash AM, Frerichs KU (2010) The safety 105(1):69–76
of dedicated-team catheter-based diagnostic cerebral 6. Harrigan MR, Deveikis JP (2012) Handbook of cerebro-
angiography in the era of advanced noninvasive imag- vascular disease and neurointerventional technique.
ing. AJNR Am J Neuroradiol 31(2):230–234 Springer Science & Business Media. New York, NY,
3. Willinsky RA, Taylor SM, TerBrugge K, Farb RI, USA, 850 p
Tomlinson G, Montanera W (2003) Neurologic compli- 7. Osborn AG (1999) Diagnostic cerebral angiography.
cations of cerebral angiography: prospective analysis of Lippincott Williams & Wilkins, Philadelphia, 462 p
2,899 procedures and review of the literature. Radiology 8. Morris P (2013) Practical neuroangiography. Lippincott
227(2):522–528 Williams & Wilkins, Philadelphia, 528 p
Part II
Carotid and Vertebral Artery: Anatomy,
Diseases, and Treatment
Anatomy of the Neck Arteries
Nasim Sheikh-Bahaei, Tomasz Matys, and Jonathan H. Gillard

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 The anatomy of neck arteries, normal varia-
tions, and anastomoses between different arter-
Aortic Arch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
ies is discussed in this chapter. All major
Common Carotid Artery . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 arteries of the neck originate from the aortic
Cervical Internal Carotid Artery . . . . . . . . . . . . . . . . . . . 89 arch via three main vessels: the
brachiocephalic trunk, left common carotid
External Carotid Artery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
(CCA), and left subclavian arteries. The CCA
External Carotid Artery Anastomoses . . . . . . . . . . . . . 92 courses superiorly in the neck, anteromedial to
Vertebral Artery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 the jugular vein and alongside the vagus nerve.
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 CCA typically divides at the level of C3 or C4
vertebral body into internal and external
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
carotid arteries. The cervical segment of the
internal carotid artery (ICA) usually does not
have branches in the neck, unless there is rem-
nant of carotid-basilar anastomoses such as the
hypoglossal artery or proatlantal artery type I.
The external carotid artery (ECA) is the
smaller branch of the CCA and runs
anteromedial to ICA. It has six branches in
the neck before entering the parotid gland and
divides into two terminal branches: superficial
temporal artery and internal maxillary artery.
The ECA supplies most of the neck structures,
scalp, and meninges. The vertebral artery is
usually the first branch of the subclavian artery,
running superiorly in the transverse foramen of
C6 to C1 with no cervical branches. It then
N. Sheikh-Bahaei (*) • T. Matys • J.H. Gillard courses between the C1 and foramen magnum
Department of Radiology, School of Clinical Medicine, and enters intracranial space. There are exten-
University of Cambridge, Cambridge Biomedical Campus,
sive anastomoses between carotid and verte-
Cambridge, UK
e-mail:;; bral arteries, the knowledge of which is;

# Springer Science+Business Media New York 2016 87

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_2
88 N. Sheikh-Bahaei et al.

essential to prevent disastrous complications scope of this chapter; in particular, developmental

during interventions. anomalies leading to the formation of right or
double aortic arches will be omitted here. The
Keywords most common anatomical variants, knowledge
Aortic arch • Common carotid artery • Internal of which is of paramount importance for
carotid artery • External carotid artery • Verte- performing safe catheter angiography, deserve a
bral artery • External-internal carotid artery brief mention however.
anastomoses • Embryonic carotid-basilar sys- The most frequent anatomical variants are the
tem anastomoses • Superior thyroid artery • common origin of the brachiocephalic trunk and
Ascending pharyngeal artery • Lingual artery the left common carotid artery seen on average in
• Facial artery • Occipital artery • Posterior 13 % of population and the origin of the left
auricular artery • Superficial temporal artery • common carotid artery from the brachiocephalic
Internal maxillary artery trunk encountered in 9 % of patients [3, 4]. Both
of these variants are more common in blacks
(25 % and 10 %, respectively). Instead of arising
Introduction separately, the left common carotid artery shares
the aortic arch origin with the brachiocephalic
In this chapter, the anatomy of the aortic arch and trunk (Fig. 1b) or arises from the brachiocephalic
neck arteries is described. The aortic arch config- trunk slightly more cranially (1–2.5 cm distal to its
uration and its common variations are reviewed as ostium) (Fig. 1c). As a result, only two major
well as the anatomy of CCA, ECA, cervical ICA, vessels ascend from the aortic arch. The common
and vertebral artery. The normal variations and trunk variant, or sometimes both of the above
anastomoses between different arteries, the variants, has been referred to as “bovine” arch
knowledge of which is crucial for performing implying high frequency of such configuration in
safe catheter angiography, are also discussed. cattle. In fact, the aortic arch in cows more often
The aim of this chapter is to illustrate the applica- gives rise to a single common vessel, and these
tion of anatomical knowledge in clinical practice. branching patterns are not specific on any species.
Thus, “bovine-type” aortic arch should be
avoided as a misnomer and replaced with an ana-
Aortic Arch tomical terminology [3, 4].
The arch origin of the left vertebral artery as a
All major arteries of the neck originate directly or fourth separate arch vessel is seen in 0.5 % of
indirectly from the aortic arch. In the standard patients; in such cases the left vertebral artery
configuration, the aortic arch gives rise to three most commonly arises between the left common
great vessels: the brachiocephalic trunk (also carotid and the left subclavian artery (Fig. 1d) and
termed brachiocephalic or innominate artery), sometimes distal to the left subclavian artery. Rare
the left common carotid artery, and the left sub- variants include the bi-innominate artery (sym-
clavian artery. The brachiocephalic trunk then metrical right and left innominate arteries each
bifurcates at the level of the right sternoclavicular bifurcating into the common carotid and subcla-
joint into the right common carotid and the right vian vessels, Fig. 1e) and bicarotid trunk (bifur-
subclavian arteries (Fig. 1a). This standard cating into the right and left common carotid
branching pattern occurs in approximately 70 % vessels, with separate origins of subclavian arter-
of patients [1, 2]. ies, Fig. 1f). Different combinations in which the
There are many possible anatomical configu- neck vessels arise separately from the aortic arch
rations of the aortic arch and its branches. are also encountered, in extreme case giving rise
Description of all these variations is beyond the to six aortic arch branches (separate right and left
6 Anatomy of the Neck Arteries 89

Fig. 1 Branching patterns of the aortic arch. (a) Standard (e) Bi-innominate artery. (f) Bicarotid trunk. (g) Separate
configuration. (b) Common origin of the brachiocephalic origins of subclavian, common carotid, and vertebral arter-
trunk and the left common carotid artery. (c) Origin of ies. (h) Aberrant right subclavian artery. R/L-SA right/left
the left common carotid artery from the brachiocephalic subclavian artery, R/L-VA right/left vertebral artery, R/L-
trunk. (d) Arch origin of the left vertebral artery. CCA right/left common carotid artery

subclavian, common carotid, and vertebral arter- the vagus nerve. The common carotid artery typ-
ies, Fig. 1g) [1, 2]. ically divides at the level of C3 or C4 vertebral
The most frequent anomaly of the aortic arch body into the external carotid artery (ECA) and
seen in 0.4–2 % of population is the aberrant right the internal carotid artery (ICA); this corresponds
subclavian artery (Fig. 1h) arising distal to the left approximately to the superior border of the thy-
subclavian artery and passing to the right side of roid cartilage. Bifurcation can occur as low as T2
the mediastinum behind the esophagus causing its and as high as C2. Normally CCA does not have
typical posterior indentation and dysphagia any named cervical branches, but occasionally the
lusoria [2]. vertebral artery or proximal ECA branches can
arise from it (Fig. 2) [1].

Common Carotid Artery

Cervical Internal Carotid Artery
The right and left common carotid arteries (CCA)
course superiorly on both sides of the neck lying The cervical segment of the ICA (C1) extends
within the respective carotid space, anteromedial from the carotid bifurcation to the skull base
to the internal jugular veins and accompanied by where the artery enters the carotid canal
90 N. Sheikh-Bahaei et al.

Fig. 2 (a) Volume rendering demonstrating the cervical The vertebral artery (VA) enters the transverse canal at C6
arteries. Common carotid artery (CCA) bifurcates into the level. Note a dominant left vertebral artery with hypoplas-
internal (ICA) and external (ECA) carotid artery at the tic right vertebral artery. (b) Posterolateral view. Distal to
level of C4 vertebral body. Note the variant origin of the the CCA bifurcation, the ICA lies posterior and lateral to
left common carotid artery from the brachiocephalic trunk. the ECA

transitioning into the petrous (C2) segment. At the basilar artery) and persistent proatlantal artery
bifurcation, the ICA most frequently arises type I (proatlantal intersegmental artery, joining
posterolaterally to ECA; its medial origin from the vertebrobasilar system after passing through
CCA is seen in 10–15 % of the population. The the foramen magnum) [1].
most proximal part of the ICA shows a focal
dilatation termed the carotid bulb or carotid
sinus. Reversal of flow and turbulent flow pattern External Carotid Artery
in this region predispose it to atherosclerotic
plaque formation. Ascending cranially, the ICA The ECA is the second branch of the CCA, which
passes medial to the ECA and lies medial to the unlike the ICA has multiple branches in the neck
internal jugular vein and anterior to the transverse and supplies most of the neck structures. The ECA
processes of the upper cervical vertebrae. Usually extends from the carotid bifurcation, usually at the
there are no major named branches of the C1 level of the upper border of the thyroid cartilage, and
segment of the ICA. Important exceptions include curves anteriorly before inclining backward behind
remnants of the carotid-basilar system anastomo- the neck of the mandible. The ECA is anteromedial
ses such as persistent hypoglossal artery (arising to the ICA [1]. It gives six main branches in the neck
from the cervical ICA between C1 and C3 and and, after entering the parotid gland, divides into its
passing through the hypoglossal canal to join the two terminal branches. In comparison to ICA, it has
6 Anatomy of the Neck Arteries 91

smaller caliber and rapidly tapers along its course as 3. The lingual artery is the second anterior branch
it gives multiple branches. of the ECA. Initially it runs obliquely upward
The following structures cross the ECA anteri- to the hyoid bone. Then it curves downward
orly: the hypoglossal nerve (CN XII), posterior and forward, forming a loop, which is crossed
belly of digastric muscle, stylohyoid muscle and by the hypoglossal nerve [1]. It passes beneath
ligament, and facial nerve (CN VII) within the the digastric and stylohyoid muscles and
parotid gland, while the pharyngeal wall, superior ascends and courses forward underneath the
laryngeal branch of the vagus nerve (CN X), and tongue. It supplies the tongue, oral cavity, and
deep lobe of the parotid gland are posterior to submandibular gland and has four branches:
ECA. The pharyngeal branch of the vagus nerve suprahyoid, dorsal lingual, deep lingual, and
(CN X) and glossopharyngeal nerve (CN IX) are sublingual arteries. The lingual artery may
between the ECA and ICA [5]. have common origin with the facial artery
(10–20 %) [5].
Branches of ECA 4. The facial artery originates above the lingual
1. The superior thyroid artery is the first branch artery. It is sheltered by the ramus of the man-
of ECA, which arises anteriorly and descends dible and courses above the submandibular
to the apex of the thyroid. It supplies the larynx gland, then curves upward over the body of
and superior thyroid and anastomoses with the the mandible, and passes anterosuperiorly
inferior thyroid artery which is a branch of the across the cheek to the angle of the mouth
thyrocervical trunk. The superior thyroid artery and nasolabial fold. It then runs toward the
may arise from CCA bifurcation (20 %) or inner canthus of the eye and terminates as the
have common origin with the lingual artery as angular artery. The facial artery is a very tortu-
thyrolingual trunk (2.5 %) or with both lingual ous vessel and supplies the face, lip, palate, and
and facial arteries as the thyrolinguofacial superficial neck [1, 5]. Its cervical branches are
trunk (2.5 %). Its main branches are the hyoid the ascending palatine artery, tonsillar,
artery, sternocleidomastoid branches, superior submental, and glandular branches. Branches
laryngeal artery, and cricothyroid branch [1, 5]. in the face include inferior and superior labial
2. The ascending pharyngeal artery is the arteries, inferior alar artery, and lateral nasal
smallest branch of the ECA which arises pos- branch [8]. The facial artery has anastomoses
teriorly near the carotid bifurcation. It is a long with the ophthalmic artery, branch of the ICA,
vessel that ascends superiorly between the ICA and other ECA branches [7, 8].
and ECA deeply in the neck [5]. After a short 5. The occipital artery originates from the poste-
common trunk, it divides into two major rior aspect of the ECA opposite to the facial
branches: anteriorly, the pharyngeal trunk, artery, at the level of the posterior belly of the
which is extracranial, and, posteriorly, the digastric muscle [1, 5]. It ascends posterosu-
neuromeningeal trunk, which is intracranial. periorly between the occiput and C1 in the
The other branches are inferior tympanic and occipital groove toward the superior nuchal
musculospinal arteries. The pharyngeal trunk line and then becomes subgaleal and extends
supplies the nasopharynx, oropharynx, and up to the vertex. It supplies the scalp, upper
eustachian tube. The neuromeningeal branches cervical musculature, and posterior fossa
supply the dura, CN IX–XII, and first, second, meninges. Branches of the occipital artery are
and third cervical roots. Inferior tympanic and sternocleidomastoid, auricular, mastoid,
musculospinal branches supply the middle ear descending, and occipital [9]. The occipital
and prevertebral muscles, respectively [6]. The artery is a remnant of the embryologic type I
ascending pharyngeal artery has several impor- and type II proatlantal arteries. It corresponds
tant and potentially dangerous anastomoses to C1 and C2 segmental arteries and, therefore,
with internal carotid, internal maxillary, verte- retains extensive connection to the vertebral
bral, and occipital arteries [6, 7]. artery in adulthood [10]. It is critical to
92 N. Sheikh-Bahaei et al.

understand the type and pattern of anastomoses anterior to the external pterygoid muscle and
between occipital and vertebral arteries to enters the pterygopalatine fossa. It terminates
avoid any disastrous complications during by sending off six branches to deep facial
embolization. structures: the posterior superior alveolar
6. The posterior auricular artery arises from the artery, infraorbital artery, artery of pterygoid
posterior ECA above the occipital artery and canal, pharyngeal artery, greater palatine
opposite to the apex of the styloid process artery, and sphenopalatine artery [12].
[1]. It ascends posteriorly beneath the parotid
and along the styloid process toward the ear. It There are several potential anastomoses
supplies the pinna, scalp, external auditory between the internal maxillary artery and ICA: in
canal, and chorda tympani. Common origin the orbital region, potential collateral route from
of occipital and posterior auricular arteries is both proximal and distal internal maxillary arteries
seen in 12.5 % [5]. to the ophthalmic branch of ICA. In the cavernous-
7. The superficial temporal artery is the smaller petrous region, the major anastomotic route is
of the two terminal branches of the ECA. It between the middle and accessory meningeal arter-
arises in the parotid gland behind the neck of ies and inferolateral trunk of the cavernous ICA [7].
the mandible and runs superiorly between the
deep and superficial lobes of the parotid gland
[1]. The transverse facial artery comes off the External Carotid Artery Anastomoses
superficial temporal artery before it leaves the
parotid gland. Then the superficial temporal There are three main regions of anastomoses
artery passes superficially over the zygomatic between the ECA and ICA and vertebral arteries:
process and divides into two branches: frontal
and parietal. It supplies the scalp and some 1. Anterior or orbital region: ECA anastomoses
parts of the face [11]. through internal maxillary and facial arteries
8. The internal maxillary artery arises within the with ophthalmic branch of ICA.
substance of the parotid gland behind the man- 2. Middle or petrocavernous: ECA connects with
dibular neck. It then runs forward between the petrocavernous branches of the ICA via inter-
ramus of the mandible and sphenomandibular nal maxillary and ascending pharyngeal
ligament and courses along the pterygoid mus- branches.
cle to the pterygopalatine fossa [5]. The inter- 3. Posterior or cervical: anastomoses between the
nal maxillary artery is divided into three major ECA and vertebral artery via occipital or
parts based on its relationship to the external ascending pharyngeal arteries [7]. Persistent
(lateral) pterygoid muscle. The first (mandibu- embryonic connection between the ECA and
lar) section is posterior to the external ptery- vertebrobasilar system is called type II
goid muscle and courses along its lower border proatlantal artery [1, 10].
giving rise to five branches: the deep auricular
artery, anterior tympanic artery, middle menin-
geal artery, accessory meningeal artery, and Vertebral Artery
inferior alveolar artery. The second (pterygoid
or muscular) section is within the pterygoid The vertebral artery usually arises as the first
muscle and runs obliquely forward and upward branch of the subclavian artery, proximal to the
as well as medially through the infratemporal thyrocervical trunk. Less frequent anatomical var-
fossa. Branches arising from this section are iations include the vertebral artery originating
entirely muscular including: anterior and pos- together with the thyrocervical trunk, from the
terior deep temporal branches, pterygoid thyrocervical trunk, from the subclavian artery
branches, masseteric artery, and buccinator distal to the thyrocervical trunk, or from the com-
artery. The third section (pterygopalatine) is mon carotid artery. In its typical course, the
6 Anatomy of the Neck Arteries 93

vertebral artery courses superiorly and medially to The level at which the vertebral artery enters
enter the transverse foramen of C6; this part of the the transverse foramen is variable and dependent
vertebral artery is termed an extraosseous segment on which embryonic intersegmental artery per-
(V1). The foraminal (V2) segment of the vertebral sists to form the vertebral artery during develop-
artery ascends through the transverse foramina of ment. Vertebral arteries often demonstrate
C6 to C1 vertebrae. At the level of the axis, the variation in size with unilateral (usually left
artery makes a sharp superolateral turn to pass sided) dominance [1]. Cervical branches of the
through its obliquely positioned transverse fora- vertebral artery include segmental muscular and
men. In the extraspinal (V3) segment between the spinal branches, often with a single branch to
C1 transverse foramen and foramen magnum, the brachial widening of the cervical cord (which
artery courses medially and posteriorly lying can also arise from the thyrocervical or
superior to the arch of C1. In the last intracranial costocervical trunk or superior thyroid artery). In
segment (V4), the artery pierces the dura and the presence of carotid-basilar anastomoses (per-
courses superomedially over the clivus to unite sistent hypoglossal artery or proatlantal artery
with its contralateral counterpart into the basilar type I from the ICA or proatlantal artery type II
artery (Fig. 3) [1]. from the ECA), the vertebral artery proximal to
anastomosis may be hypoplastic or absent.


All major neck arteries originate from the aortic

arch. In the standard configuration the aortic arch
gives rise to three great vessels: the
brachiocephalic trunk, the left common carotid
artery, and the left subclavian artery. The most
frequent anatomical variants are the common ori-
gin of the brachiocephalic trunk and the left com-
mon carotid artery (13 %). The common carotid
arteries course superiorly on both sides of the neck
and divide at the level of C3 or C4 vertebral body
into the ECA and ICA. The cervical segment of
ICA (C1) extends from the carotid bifurcation to
the skull base. At the bifurcation, the ICA most
frequently arises posterolaterally to the ECA and
then passes medially. Usually there are no major
named branches of the C1 segment of the ICA.
Important exceptions are persistent hypoglossal
artery and proatlantal artery type I as the remnants
of the carotid-basilar system anastomoses. The
ECA is the smaller branch of the CCA which has
six branches in the neck before entering the
Fig. 3 Volume rendering demonstrating the typical course
of the vertebral artery. The segment proximal to the trans- parotid gland and divides into two terminal
verse foramen of C6 is termed V1; the segment coursing branches: superficial temporal artery and internal
through the transverse canal from C6 to C1 is termed V2; maxillary artery. The ECA supplies most of the
note a sharp superolateral turn at the C2 transverse fora-
neck structures, scalp, and meninges. The verte-
men. Extraspinal V3 segment stretches from the transverse
foramen of C1 to the site at which the artery pierces the bral artery is usually the first branch of the sub-
dura and enters the cranium clavian artery, running superiorly in the transverse
94 N. Sheikh-Bahaei et al.

foramen of C6 to C1 with no cervical branches. It clarification of a common misnomer. AJNR Am J

then courses between the C1 and foramen mag- Neuroradiol 27(7):1541–1542
5. Standring S (2011) Gray’s anatomy, 39th edn. Living-
num and enters intracranial space. stone, Churchill
There are three main regions of anastomoses 6. Hacein-Bey L, Daniels DL, Ulmer JL, Mark LP et al
between the ECA with ICA and vertebral artery: (2002) The ascending pharyngeal artery: branches,
(1) anterior or orbital region, (2) middle or anastomoses, and clinical significance. Am J
Neuroradiol 23:1246–1256
petrocavernous, and (3) posterior or cervical. Per- 7. Geibprasert S, Pongpech S, Armstrong D, Krings T
sistent embryonic connection between the ECA (2009) Dangerous extracranial–intracranial anastomo-
and vertebrobasilar system is called type II ses and supply to the cranial nerves: vessels the
proatlantal artery. It is very crucial to identify neurointerventionalist needs to know. Am J
Neuroradiol 30:1459–1468
these anastomoses or embryological remnants to 8. Lohn JWG, Penn JW, Norton J, Butler PEM (2011)
prevent disastrous complications during The course and variation of the facial artery and vein,
interventions. implications for facial transplantation and facial sur-
gery. Ann Plast Surg 67:184–188
9. Alvernia JE, Fraser K, Lanzino G (2006) The occipital
artery: a microanatomical study. Neurosurgery
References 58(1 Suppl):114–122
10. Chen PR, Siddiqui AH, Chen PR (2011) Intracranial
1. Harnsberger HR, Macdonald AJ (2006) Diagnostic and arterial collateralization: relevance in neuro-
surgical imaging anatomy, 1st edn, Brain, head and endovascular procedures In: Peres JFP (ed)
neck, spine. Amirsys, Salt Lake City Neuroimaging for clinicians- combining research and
2. Lippert H, Pabst R (1985) Arterial variations in man: practice. InTech Europe, Rijeka, p 175–202
classification and frequency. J F Bergmann, Munchen 11. Pinar YA, Govsa F (2006) Anatomy of superficial
3. Katz JC, Chakravarti S, Ko HH, Lytrivi ID, temporal artery and its branches: its importance for
Srivastava S, Lai WW et al (2006) Common origin of surgery. Surg Radiol Anat 28:248–253
the innominate and carotid arteries: prevalence, 12. Kashiwagi N, Nakanishi K, Kozuka T, Sato Y
nomenclature, and surgical implications. J Am Soc et al (2010) Vascular supply with angio-CT for
Echocardiogr 19(12):1446–1448 superselective intra-arterial chemotherapy in advanced
4. Layton KF, Kallmes DF, Cloft HJ, Lindell EP, Cox VS maxillary sinus cancer. Brit J Radiol 83:171–178
(2006) Bovine aortic arch variant in humans:
Imaging of Carotid Atherosclerosis
Jie Sun, Niranjan Balu, and Chun Yuan

Contents Abstract
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 Growing interest in the in vivo characterization
of plaque pathology has furthered the develop-
Pathology and Implications for Imaging . . . . . . . . . 96
Pathogenesis of Carotid Atherosclerosis . . . . . . . . . . . . 97
ment of vessel wall imaging approaches
Histopathological Characteristics . . . . . . . . . . . . . . . . . . . . 97 beyond traditional techniques of luminal imag-
Plaque Morphology and Clinical Presentation . . . . . . 98 ing. By leveraging histopathological informa-
MR Imaging Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 tion from carotid endarterectomy specimens,
Field Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 magnetic resonance imaging has been proven
Coils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 capable of providing qualitative and quantita-
Multi-slice Black-Blood MRI . . . . . . . . . . . . . . . . . . . . . . . 100 tive information on a number of morphological
3D Black-Blood MRI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Multi-Contrast MRI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 and pathological features of carotid atheroscle-
Lipid-Rich Necrotic Core . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 rosis, including fibrous cap and necrotic core,
Intraplaque Hemorrhage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 intraplaque hemorrhage, plaque neovascu-
Calcification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 larization, and inflammation. These technical
Loose Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Neovasculature and Inflammation . . . . . . . . . . . . . . . . . . . 105 advancements present new opportunities to
Measurement of Plaque Morphology expand the understanding of the pathophysiol-
and Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 ogy of ischemic stroke and devise more effi-
Current Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 cient diagnostic and prognostic tools to address
Imaging-Based Patient Stratification . . . . . . . . . . . . . . . . 106 contemporary clinical problems in the manage-
Carotid Artery Stenting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 ment of carotid atherosclerosis. From a clinical
Cryptogenic Stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
In Vivo Insights on Pathobiological Mechanisms . . . 109
perspective, this article introduces the various
Efficacy Markers for Testing Therapeutic magnetic resonance techniques for carotid ath-
Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 erosclerosis imaging. The background and
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 pathological basis of direct plaque imaging in
carotid arteries are discussed, followed by
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
available solutions and key technical consider-
ations, as well as promising applications that
have arisen from these techniques.

J. Sun • N. Balu • C. Yuan (*) Keywords

Department of Radiology, University of Washington,
Carotid artery • Atherosclerosis • Magnetic
Seattle, WA, USA
e-mail:; ninja@u.washington. resonance imaging • Plaque rupture • Black-

# Springer Science+Business Media New York 2016 95

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_34
96 J. Sun et al.

blood • Multi-contrast • Ischemic stroke • morphology and clinical complications has

Carotid endarterectomy • Carotid artery shifted the focus of imaging of carotid atheroscle-
stenting • Cryptogenic stroke rosis from luminal narrowing to plaque instability,
i.e., the risk of acute plaque rupture and thrombo-
sis. Magnetic resonance imaging (MR, MRI) has
Introduction emerged as a promising imaging modality for
in vivo assessment of plaque instability. Over the
Atherosclerosis is a systemic condition that years, many validation studies have leveraged
affects large- and medium-sized arteries, such as carotid endarterectomy specimens to correlate
the aorta and its branches. Carotid atherosclerosis, in vivo imaging findings with histopathological
especially atherosclerotic plaques at the carotid features. Carotid MRI has been shown capable of
bifurcation, is one of the most prevalent manifes- characterizing a number of pathological features
tations of systemic atherosclerosis. It has been and biological activities of atherosclerotic plaques
estimated from pooled population-based data that are implicated in plaque rupture, including
that the prevalence of moderate carotid stenosis fibrous cap and necrotic core, intraplaque hemor-
(>=50 %) is 4.8 % and 2.2 % for men and women rhage, plaque neovascularization, and inflamma-
less than 70 years old, respectively. The numbers tion [2–5]. As such, direct MR imaging of carotid
increase further still in those aged more than atherosclerosis presents a unique opportunity to
70 years to 12.5 % and 6.9 % [1]. develop more effective diagnostic and prognostic
Although the majority of carotid atheroscle- approaches for clinical research and practice.
rotic plaques have a benign clinical course, a
portion of them result in ischemic stroke, a leading
cause of death and disability worldwide. A pri- Pathology and Implications
mary goal of clinical imaging of carotid athero- for Imaging
sclerosis is therefore to identify lesions that will
lead to ischemic stroke despite control of risk Atherosclerotic plaques are small but complex
factors. Atherosclerotic plaques typically become pathological structures with considerable mor-
flow-limiting as they progress. Unfortunately, the phological variations (Fig. 1). For imaging scien-
simple assessment of luminal narrowing has tists and radiologists to develop effective imaging
proven to be of limited value in predicting the protocols and properly interpret image findings, a
risk of ischemic stroke. Expanded knowledge on basic understanding of the pathology of carotid
the relationship between atherosclerotic plaque atherosclerosis is essential.

Fig. 1 Atherosclerosis in the carotid bulb from three (b) The plaque is mainly fibrotic with small areas of lipid
endarterectomy specimens showing morphological var- accumulation (white arrows) and calcification. (c) Pieces
iations. (a) A mature necrotic core with intraplaque hem- of dense calcification comprise the main portion of the
orrhage (asterisk) is the main component of the plaque. plaque (hollow arrow)
The overlying fibrous cap is an inhomogeneous thickness.
7 Imaging of Carotid Atherosclerosis 97

Pathogenesis of Carotid microvasculature present in large- and medium-

Atherosclerosis sized arteries, with the physiological function of
supplying oxygen and nutrients to the outer por-
Atherogenesis is now appreciated as a multifacto- tion of the medial layer). Initially considered as a
rial process in which the arterial wall interacts compensatory mechanism to alleviate arterial wall
with a number of systemic and local factors. hypoxia as wall thickness increases, accumulating
The disturbed flow conditions at carotid bifurca- evidence from recent studies suggests that plaque
tion make the carotid bulb a predilection site of neovascularization has an active role in
extracranial carotid artery for atherosclerosis. atherothrombosis by contributing to the develop-
Nonetheless, more extensive disease, detectable ment, progression, and destabilization of athero-
by improved imaging approaches, is not uncom- sclerotic plaques. Leaky microvessels may
mon, particularly in the common carotid artery. facilitate the entry of inflammatory cells as well
Individual variations in the geometry of carotid as plasma proteins. Furthermore, it is thought that
bulb and the unknown extent of outward episodic ruptures of plaque microvessels are
remodeling make it difficult to discern the true likely the origin of intraplaque hemorrhage, a
plaque burden by angiography or similar common finding in advanced human lesions.
techniques. Improvements in research approaches, for both
The accumulation of extracellular lipids and histopathology and imaging, have recently
cell debris, the migration and proliferation of enabled a better appreciation of the prominent
smooth muscle cells, and extracellular matrix pro- role of intraplaque hemorrhage in atherogenesis
duction are recognized processes in the formation as well as the pathophysiology of clinical mani-
of plaques with well-developed necrotic core festations. However, mechanistic insights about
(s) and an overlying fibrous cap, or fibroatheroma intraplaque hemorrhage are still limited at this
as described by Virmani et al. in coronary athero- moment, as the phenomenon is rarely seen in
sclerosis [6]. Structural changes are paralleled by prevailing animal models.
the recruitment and activation of immune cells
that respond to local insult and modulate the
inflammatory process. A variety of immune cells Histopathological Characteristics
including monocytes, lymphocytes, and, more
recently, granulocytes have been shown to play a To date, much of the understanding of carotid
role in the pathogenesis and prognosis of athero- atherosclerotic plaque morphology is based on
sclerosis. Furthermore, it has been suggested histopathological examination of carotid endarter-
through dedicated experimental and human ectomy specimens, which do not cover the full
autopsy studies that immune cells can be highly spectrum of lesion severity as compared to human
adaptive in response to microenvironmental sig- autopsy. The peripheral sections of advanced
nals. In accord, functionally indistinct monocytes carotid specimens are presumed to give a glimpse
may differentiate into polarized macrophage sub- of early or intermediate lesions, which frequently
populations in the atherogenic milieu, taking on show proteoglycan accumulation, lipid retention,
heterogeneous biological functions that may sta- or early stage necrotic cores. Based on these
bilize or destabilize the plaque [7]. Therefore, the observations, it appears sensible to describe
biological complexity of plaque inflammation has carotid plaque morphology with nomenclature
placed a higher demand on molecular imaging by proposed for classifying coronary atherosclerosis.
emphasizing the activities rather than the mere Yet caution should be exercised when making an
presence of immune cells. Another morphological analogy for a specific feature with regards to its
change that initiates early in the disease process is role in pathogenesis and prognosis, given the dis-
plaque neovascularization, resulting from the pen- tinct differences in vessel diameter and flow con-
etration of the medial layer (i.e., tunica media) by ditions between coronary circulation and
hyperplastic vasa vasorum (i.e., the nutritional extracranial carotid circulation.
98 J. Sun et al.

Few studies have attempted to compare plaque mural thrombus constituted of platelets or fibrin
morphologies between coronary and carotid ath- on plaque surface, which was associated with
erosclerosis, yet notable differences seem to be either plaque rupture (90.1 %) or erosion (9.9 %)
present. The fibrous cap overlying necrotic core(s) in patients with stroke [9]. Furthermore, fresh
is at risk of rupture in both coronary and carotid thrombus was more frequently observed when
lesions. The high mechanical forces in the carotid carotid endarterectomy was performed closer to
circulation may render the fibrous cap of carotid symptom onset. With carotid plaques collected
atherosclerosis more vulnerable to rupture, as it has from participants of the North American Symp-
been reported that ruptured carotid plaques have tomatic Carotid Endarterectomy Trial (NASCET)
much thicker fibrous cap and fewer macrophage and the Asymptomatic Carotid Atherosclerosis
infiltration than ruptured coronary plaques [8]. Study (ACAS) trials (71 symptomatic, 170 asymp-
Thrombotic activity typically ensues, yet the high tomatic), Fisher et al. noted that thrombus was
flow velocity is likely a limiting factor for the associated with plaque ulceration, both of which
propagation of mural thrombus and the healing of were more prevalent in symptomatic patients than
fibrous cap rupture. In line with the hypothesis, in asymptomatic patients [10]. Another study by
ulceration is reported more frequently in carotid Redgrave et al. examined 570 consecutive symp-
atherosclerosis than in coronary atherosclerosis, tomatic carotid plaques and demonstrated a con-
whereas total thrombotic occlusion of carotid ste- tinued post-event decline of plaque vulnerability
nosis is not as common as in coronary atheroscle- features including cap rupture and plaque inflam-
rosis. Fibrous cap erosion and calcified nodules are mation [11]. These studies have provided impor-
the other two types of surface disruption reported in tant insights onto the pathogenesis of ischemic
the coronary artery, which may be accountable for events associated with carotid atherosclerosis. A
over a third of atherothrombotic events in coronary typical plaque responsible for a recent ischemic
atherosclerosis [6]. However, in the carotid artery, stroke can be described as one with active throm-
plaque rupture is predominant, while plaque ero- bosis on the luminal surface, triggered by plaque
sion does not seem to be a major cause for rupture and exposure of thrombogenic materials
atherothrombosis [9]. Although calcified nodules to the flowing blood. Subsequently, acute plaque
are more frequently seen in carotid atherosclerosis rupture and thrombosis may lead to artery-to-
in comparison to coronary atherosclerosis, its role artery embolism or thrombotic occlusion, which
in atherothrombosis in the carotid artery has yet to is in line with clinical observations.
be determined. Mechanisms of plaque rupture are not yet fully
understood. In the study by Redgrave et al. [11],
two morphological features demonstrated inde-
Plaque Morphology and Clinical pendent association with cap rupture – intraplaque
Presentation hemorrhage and inflammatory cell infiltration of
fibrous cap. Other studies based on data from
Morphological and pathological features of coronary atherosclerosis have suggested that
carotid atherosclerosis have been shown to be fibrous cap thickness and necrotic core size are
associated with downstream neurological ische- good discriminators of plaques at risk for rupture
mic events. Despite the longtime focus of clinical from those that are stable [12]. Of note, the degree
imaging on luminal stenosis, chronic stenosis or of stenosis could fit in here as one of the risk
occlusion from carotid atherosclerosis rarely leads factors for plaque rupture, considering the
to ischemic stroke and is therefore not considered increased mechanical forces associated with
the pathological substrate of ischemic stroke. In a local narrowing. Some of these risk factors may
study of 269 carotid plaques obtained en bloc have a more dynamic natural course, such as cap
from surgical endarterectomy (96 ischemic stroke, inflammation. It is plausible that in vivo charac-
91 transient ischemic attack, 82 asymptomatic), terization of major risk factors of cap rupture may
Spagnoli et al. observed a high prevalence of fresh enable accurate estimation of the short-term or
7 Imaging of Carotid Atherosclerosis 99

long-term risk of neurological ischemic events in core, intraplaque hemorrhage, etc. Indeed, a
patients with carotid atherosclerosis. unique advantage of MR plaque imaging is to
employ more than one contrast weightings to
help differentiate various atherosclerotic
MR Imaging Techniques components.
Thus, carotid vessel wall imaging places spe-
In contrast to luminal imaging using bright blood cial technical constraints on field strength, coils,
MR angiography, vessel wall visualization and and pulse sequences, which are required for accu-
the identification of components of atherosclero- rate identification of pathology. Carotid MRI pro-
sis require suppression of signal from intraluminal tocols need to consider all these aspects while
blood. Thus, MRI sequences commonly referred being tailored for each study. This section outlines
to as black-blood MRI are essential for vessel wall the established carotid MRI sequences and emerg-
imaging. Normal extracranial carotids have thin ing trends in carotid atherosclerosis imaging.
pulsating walls and are often situated in deep
tissue planes in patients with a large neck. Carotid
bifurcation, the most frequent site of carotid ath- Field Strength
erosclerosis, has a complex flow pattern with
recirculating flow. Thus, complete flow suppres- Initial development of carotid vessel wall MRI
sion in carotid bifurcation requires special consid- took place on clinical 1.5-T scanners. Moving to
eration of the flow conditions at carotid higher field strengths can increase the available
bifurcation. SNR. Initial studies comparing 1.5 and 3-T carotid
In addition to blood suppression, consider- MRI found an improvement of wall SNR and
ations of spatial resolution, signal-to-noise ratio lumen/wall CNR by 1.5-fold for T1-WI imaging
(SNR), motion suppression, etc., are also impor- and 1.7–1.8-fold for proton-density-weighted/T2-
tant to obtaining diagnostic quality vessel wall WI imaging [13]. With optimization of sequences
images. Atherosclerotic plaque components are for change in field strength, plaque components
small but complicated structures that require can be identified with similar repeatability at 3-T.
high spatial resolution for identification and quan- However, quantification of some components,
tification. Of note, the requirement of high spatial such as calcification and intraplaque hemorrhage,
resolution must be achieved with good SNR may differ due to higher susceptibility effects at
within a clinically acceptable scan time. Another 3-T. Additionally, the specific absorption rate
reason to avoid long scan times is to reduce (SAR) is higher at 3-T and should be accounted
motion artifacts. Due to the unique location of for in protocol optimization. Carotid MRI at
carotid arteries, patient swallowing has been one higher field strengths is challenging due to inho-
of the major causes for degraded image quality of mogeneous excitation, suboptimal fat suppres-
carotid MRI, whereas it will be less of a problem if sion, and higher susceptibility. Despite these
the scanning time is short. disadvantages, carotid imaging is moving into
Contrast-to-noise ratio (CNR) is equally 7-T due to the advantages of higher SNR.
important in order to characterize vessel wall
pathology. Delineation of the lumen requires
good CNR with adequate blood suppression, Coils
while delineation of the outer wall requires good
fat suppression. This allows plaque boundaries to At 1.5 and 3-T, body coil transmission is used for
be clearly identified and disease burden to be homogenous excitation, which is also useful for
accurately estimated. Furthermore, MRI many blood suppression sequences that use
sequences with special contrast mechanisms are nonselective excitation. However, body coil
required for identification of various plaque com- reception does not provide a high enough SNR
ponents such as calcification, lipid-rich necrotic for carotid vessel wall MRI within clinically
100 J. Sun et al.

acceptable scan times. Thus, phased array surface flows through the imaging slice and produces
coils with a small diameter and applied close to a black-blood effect. This method can be used
the neck provide improved SNR and are often for short TE sequences such as T1-WI con-
used for vessel wall MRI. Phased array coils trast but is prone to flow artifacts depending
with four, six, and eight coil elements have been on the thickness of the saturation slab and its
designed for carotid imaging. Compared to a four- distance from the imaging slice. A traveling
element design, an eight-element design was saturation band is used to ensure that the
shown to provide 1.7-fold higher SNR and larger distance to all slices remains the same and a
coverage along the axis of the artery [14]. Some of short delay time is used so that blood signal
these coil designs are available commercially on does not recover before being imaged.
all major scanner vendor platforms. MRI at 7-T Advantages of inflow saturation include low
also requires custom coil design, but these are SAR, wide availability, and selectivity in
currently available for research purposes only. suppressing arteries and/or veins.
(c) Double inversion-recovery (DIR) and variants:
DIR preparation consists of two inversion
Multi-slice Black-Blood MRI pulses followed by a delay before image acqui-
sition. A nonselective inversion pulse is
Traditionally, the carotid vessel wall is scanned followed by a slice-selective re-inversion of
using 2 mm thick slices with higher in-plane res- the imaging slice. Blood with inverted magne-
olution of 0.5–0.6 mm. A stack of 12–16 slices tization proximal and distal to the imaging slice
centered on the carotid bifurcation covering a flows into the imaging slice while simulta-
3 cm region is used in a multi-slice black-blood neously undergoing T1 relaxation. The inver-
MRI protocol. Blood suppression can be achieved sion delay time (TI) is calculated such that
by black-blood preparation of the magnetization blood magnetization is zero during image
prior to gradient or spin-echo acquisition. The acquisition, thereby causing a black-blood
sequences used for black-blood imaging can be effect. Both arterial and venous blood are there-
briefly described as below: fore nulled by DIR. The long TI time allows
sufficient time for blood inflow and ensures
(a) Spin-echo acquisition: Direct spin-echo MRI good blood suppression. However, artifacts
without any magnetization preparation can may still occur due to slow or recirculating
provide black-blood contrast with proper flow since the non-inverted blood within the
choice of echo time (TE). Only spins imaging slice may not have time for outflow
experiencing both the excitation and from the imaging slice. Increasing imaging
refocusing spins will produce signals. If after slice or slab thickness tends to worsen flow
experiencing the 90 pulse blood flows artifacts with DIR. DIR has better performance
through the slice during TE and does not than inflow suppression and is widely available
experience the refocusing pulse, it will pro- on all vendor scanner platforms. Accurate
duce a black-blood effect in the acquired echo. lumen identification in vessel wall imaging
Long TE spin-echo or fast spin-echo requires at least a DIR preparation. The pri-
sequences such as T2-WI images exhibit this mary disadvantage of DIR for carotid imaging
effect. However, slow flow and recirculating is that it is a single slice method. Multi-slice
flow are not suppressed by this method. DIR (MDIR) preparations have been proposed
Increasing slice thickness will also lead to to reduce scan time associated with DIR
more flow artifacts. [15]. The current most efficient method uses a
(b) Inflow saturation: In this method, the signal slab-selective re-inversion rather than a slice-
from a thick slab proximal to the carotid bifur- selective inversion, thus allowing up to eight
cation is completely dephased every repeti- slices to be acquired with good flow suppres-
tion time (TR). The dephased blood then sion within a single TR [16].
7 Imaging of Carotid Atherosclerosis 101

(d) Motion sensitization: All the above methods 3D Black-Blood MRI

perform best with laminar blood flow with
artifacts arising in cases of slow flow, turbu- MRI studies of atherosclerotic plaque have gen-
lent flow, and recirculating flow. Flow in the erally used 2–3 mm thick axial slices, since black-
normal carotid bifurcation can be blood techniques such as inflow suppression or
recirculating or turbulent, causing plaque- DIR rely on inflow or outflow for efficient blood
mimicking artifacts on black-blood MRI. suppression. Compared to the traditional
Use of motion-sensitized-driven equilibrium two-dimensional (2D) sequences, 3D black-
preparation (MSDE) can effectively suppress blood MRI can improve the accuracy and the
artifacts from recirculating or turbulent flow repeatability of plaque measurements due to
[17]. MSDE with two refocusing pulses is less higher spatial resolution in the z direction and
affected by B1 inhomogeneity and eddy cur- less partial volume averaging. However, this
rent effects than MSDE with one refocusing requires blood suppression in a relatively thick
pulse [18]. MSDE is not dependent on inflow slab, and therefore use of DIR can result in
of blood into the imaging slice. Hence, it can increased flow artifacts. Since MSDE flow sup-
be used for non-axial slice acquisition. Addi- pression does not depend on flow, 3D black-blood
tionally, MSDE is also compatible with acqui- MRI can be achieved with high isotropic spatial
sition of thick slabs and is therefore used for resolution with this technique. A 3D MSDE pre-
three-dimensional (3D) black-blood MRI. An pared rapid gradient echo (3D-MERGE) vessel
extension of motion sensitization uses a train wall imaging technique can provide
of delay alternating with nutation for tailored 0.7  0.7  0.7 mm3 isotropic resolution within
excitation (DANTE) subpulses with inter- a 2-min scan time with good blood suppression
spersed gradient dephasers for suppression (Fig. 2) [20]. Since 3D-MERGE datasets are iso-
of blood flow [19]. DANTE with high ampli- tropic, interactive reformatting of multiple image
tude dephasing gradients can also be used for planes can be used for plaque visualization. Spin-
3D MRI. echo-based sequences with variable flip angle

Fig. 2 3D isotropic high-

(0.7  0.7  0.7 mm3)
carotid artery images
acquired with
3D-MERGE. Two major
advantages offered by 3D
isotropic black-blood vessel
wall imaging are shortened
scan times for clinical utility
and flexible reformation
capability for improved
plaque visualization. Note
that vessel wall boundaries
are clearly visible on all
reformats (Reprinted with
permission from Balu
et al. [20])
102 J. Sun et al.

Table 1 Multi-contrast carotid MRI protocol for 3-T

Pre/post- Proton-
contrast Time-of- density 3D-MERGE
T1-WI flight T2-WI -weighted SNAP sequence sequence
Image mode 2D 3D 2D 2D 3D 3D
Scan plane Axial Axial Axial Axial Coronal Coronal
TR (ms) 800 20 4,800 4,800 13 10
TE (ms) 10 4.6 50 10 4 5
Flip angle ( ) 90 50 90 90 11 6
FOV (cm) 14  14 14  14 14  14 14  14 14  14  3.2
Resolution 0.55  0.55 0.55  0.55 0.55  0.55 0.55  0.55 0.7  0.7  0.7 0.7  0.7  0.7
Slice 2 1 2 2 Not applicable Not applicable
thickness (interpolated)
Blood QIR/MSDE Venous DIR/MSDE DIR/MSDE Not applicable MSDE
suppression inflow
Fat Yes No Yes Yes Yes Yes
3D-MERGE 3-dimensional MSDE prepared rapid gradient echo, DIR double inversion-recovery, FOV field of view,
MSDE motion-sensitized-driven equilibrium preparation, PSIR phase-sensitive inversion-recovery, QIR quadruple
inversion-recovery, SNAP simultaneous non-contrast angiography and intraplaque hemorrhage, TFE turbo field echo,
TSE turbo spin echo

long echo trains have also been suggested for complementary fashion is essential for compre-
imaging plaque burden [21]. Further studies on hensive evaluation of atherosclerotic plaques.
the use of 3D sequences for identification of Table 1 shows a typical carotid MRI protocol
plaque components are required before 3D MRI with traditional 2D multi-contrast MRI in com-
can be used in a similar way to current 2D MRI parison to newer isotropic 3D MRI sequences.
protocols. Single contrast weightings may be used when
only a subset of plaque components is of interest.
Sequences of interest for detecting important
Multi-Contrast MRI plaque components are briefly described. Table 2
shows the appearance of plaque components on
Atherosclerotic plaques may range from simple each of the different contrast weightings.
wall thickening to complicated plaques with mul-
tiple plaque components. By varying pulse
sequence parameters, image contrast may be Lipid-Rich Necrotic Core
manipulated such that specific plaque components
are visible in each contrast weighting. For exam- Lipid accumulation in plaques evolves from lipid
ple, a highly T1-WI fat-suppressed sequence will pools to necrotic cores containing cell debris with
show intraplaque hemorrhage with high signal progression of atherosclerosis. Fat suppression
intensity compared to fibrous tissue or muscle. used for plaque imaging suppresses signal from
Thus, a multi-contrast MRI protocol where multi- triglycerides in subcutaneous fat but does not
ple contrast weightings are used in a suppress signal from necrotic core, where the
7 Imaging of Carotid Atherosclerosis 103

Table 2 Plaque component classification using multi-contrast MRI

Time-of-flight T1-WI Proton-density-weighted T2-WI Post-contrast T1-WI
Lipid-rich necrotic core o o/+ o/+ /o
Intraplaque hemorrhage + + /o /o /+
Loose extracellular matrix o /o + + +
Dense fibrous tissue o o o +
Signal intensities are relative to adjacent muscle
+ hyperintense, o isointense, hypointense

signal mainly originates from free cholesterols Intraplaque Hemorrhage

and cholesteryl esters. Necrotic core is isointense
on T1-WI sequences, but can be seen on T2-WI Bleeding into the plaque may come from either
sequences as a hypointense region due to its short the lumen secondary to surface rupture or from the
T2 components. Thus, a non-contrast MRI proto- vasa vasorum originating from the adventitia.
col requires a T2-WI sequence for detection of Hemoglobin in the red blood cells then passes
necrotic core. through several stages of degradation. One of the
Sensitivity and specificity for detection of intermediate stages is the formation of highly
necrotic core can be improved by administration paramagnetic methemoglobin. Methemoglobin is
of gadolinium contrast agent such as gadopentetic visible as hyperintense regions on T1-WI images
acid (Gd-DTPA) [2]. Due to necrotic contents, such as T1-WI fast spin echo or time-of-flight
gadolinium uptake into necrotic core is negligible images. However, the sensitivity and specificity
compared to surrounding fibrous tissue. Post- for detection of intraplaque hemorrhage can be
contrast difference between necrotic core and improved by use of highly T1-WI sequences [23].
fibrous tissue is maximal 5–10 min after contrast Magnetization-prepared rapid gradient echo
injection. Therefore, necrotic core can be readily (MPRAGE) is often used for detection of
identified by comparing pre-contrast and post- intraplaque hemorrhage [3]. The MPRAGE
contrast black-blood T1-WI images. However, sequence employs an inversion recovery preced-
adequate blood-suppression post-contrast is chal- ing a gradient echo acquisition. Fat suppression is
lenging. The T1 of blood decreases sharply after also employed to reduce signals from subcutane-
contrast administration, but the change in blood ous fat. MPRAGE has been shown to have higher
relaxivity is not predictable since it is dependent sensitivity and specificity for intraplaque hemor-
on distribution volume – an unknown patient- rhage detection than T1-WI fast spin echo or time-
dependent factor. Thus, while black-blood tech- of-flight. However, blood suppression in
niques such as DIR provide good blood- MPRAGE can be challenging since TI has to be
suppression pre-contrast, adjustment of TI of optimized simultaneously for both blood suppres-
DIR post-contrast is empirical and does not pro- sion and visualization of intraplaque hemorrhage.
vide adequate blood suppression in all subjects. A A better approach combines MPRAGE acquisi-
quadruple inversion-recovery (QIR) preparation tion with phase-sensitive reconstruction (PSR).
has been introduced for this purpose [22]. QIR While PSR requires the additional acquisition of
consists of two DIR modules in tandem and con- a reference image, it provides additional benefits
sists of two TIs that are adjusted such that blood over blood suppression. PSR is used in the simul-
suppression is achieved for the set repetition time taneous non-contrast angiography and intraplaque
(TR) of the sequence. QIR blood suppression is hemorrhage (SNAP) MRI sequence where the
T1 insensitive, thus enabling the same sequence nonselective inversion pulse of MPRAGE is
parameters to be used for both pre- and post- replaced with a slab-selective inversion [24]. The
contrast T1-WI MRI. TI of SNAP is adjusted such that blood flowing
104 J. Sun et al.

Fig. 3 SNAP imaging and histopathology of a compli- Axial reformats of the bifurcation are shown in (d) with an
cated plaque with intraplaque hemorrhage and ulcera- ulcer (arrows) shown as a dark depression and intraplaque
tion. Two image portions are simultaneously acquired. The hemorrhage as hyperintense signals. Imaging findings are
portion of non-contrast MR angiography (a) shows a high- confirmed on matched histological cross sections
grade stenosis in the proximal external carotid artery (Masson’s trichrome). Note that the good blood suppres-
(arrowhead) and ulcerated plaque in the proximal internal sion in (d) allows clear distinction of intraplaque hemor-
carotid artery (ICA) (arrow). The intraplaque hemorrhage rhage from lumen (Reprinted with permission from Wang
portion (b) can be fused with the angiogram portion (c). et al. [24])

into the imaging plane is inverted only a single calcification in multi-contrast plaque MRI. How-
time compared to the blood in MPRAGE that may ever, two issues need to be noted. First, it is
be inverted multiple times. During image acquisi- common to witness calcifications that are either
tion at TI, flowing blood which has a long T1 has close to luminal surface or even in direct contact
negative longitudinal magnetization, while short T1 with flowing blood in carotid atherosclerosis. In
components such as intraplaque hemorrhage have a such situations, it could be challenging to separate
positive longitudinal magnetization. PSR of SNAP calcification from lumen based on black-blood
takes advantage of this difference in sign of longi- images alone. As such, accurate identification of
tudinal magnetization to provide two images from calcification and therefore overall plaque size
the same acquisition: negative part provides an depend on the combined use of bright-blood and
angiogram, while the positive part provides signal black-blood sequences. Most multi-contrast pro-
for intraplaque hemorrhage. Figure 3 shows an tocols used in previous clinical studies have
ulcer on the SNAP angiogram with corresponding involved at least 3D time-of-flight imaging to
intraplaque hemorrhage in the plaque. Since both overcome the problem. Second, the quantification
MR angiography and intraplaque hemorrhage of calcification should be mainly performed on
imaging are obtained from the same SNAP acqui- spin echo sequences since their size tends to be
sition, they are naturally registered together. overestimated on gradient echo sequences due to
increased susceptibility of calcification. Use of
very short TE sequences called ultrashort echo
Calcification time (UTE) sequences renders calcification bright
and is less affected by susceptibility. UTE
Calcification is hypointense on all contrast sequences require special hardware and spiral or
weightings due to its very short T2 relaxation radial readouts in addition to dual echo acquisition
time. This forms the basis for identification of for the identification of calcifications.
7 Imaging of Carotid Atherosclerosis 105

Loose Matrix as gadolinium contrast is injected. Additionally, a

bright blood technique is preferred to measure the
Loose matrix refers to the category of plaque arterial input function. Therefore, a dynamic
tissue rich in proteoglycans, which can be encoun- contrast-enhancement (DCE) MRI sequence is
tered frequently in carotid atherosclerosis of var- started prior to gadolinium injection. In practice,
ious stages. Loose matrix tissue is rich in water this is a T1-WI bright-blood gradient echo
and shows up as hyperintense signals on T2-WI sequence with inflow saturation of both arterial
images. In addition, loose matrix may be perme- and venous flow. After two dynamics are
able to gadolinium and show up as areas of high acquired, half dose of gadolinium contrast bolus
contrast uptake on post-contrast black-blood is injected at a rate of 1 cc/s. Since the T1 relax-
MRI. However, their clinical implications when ation time of blood is drastically reduced, a bright
encountered in advanced lesions are uncertain. blood effect is created post-contrast. About
20 dynamic scans with scan times of 15 s per
dynamic scan are collected. Kinetic modeling
Neovasculature and Inflammation using the Patlak model provides two quantitative
measurements: fractional plasma volume (vp) and
Unlike the previous components, neovasculature permeability rate constant (Ktrans). Both vp and
and inflammation detection require specialized Ktrans have been found to be correlated with his-
modeling of gadolinium contrast uptake for detec- tologically measured neovasculature and macro-
tion. In contrast to post-contrast T1-WI black- phages in a validation study using carotid
blood MRI used for identification of lipid core, endarterectomy specimens [4]. Figure 4 shows a
neovasculature and inflammation detection pseudo-color image with vp in red and Ktrans in
requires fast and continuous acquisition of images green.

Fig. 4 Vasa vasorum

imaging for evaluating
plaque revascularization
and inflammation. Cross-
sectional images of right
carotid bifurcation on
pre-contrast T1-WI QIR (a),
post-contrast T1-WI QIR
(b), and corresponding vasa
vasorum image (c) obtained
from contrast kinetic
modeling of DCE-MRI.
Red areas correspond to
partial volume of blood (vp)
and green areas correspond
to transfer constant (Ktrans).
Asterisk indicates lumen;
circle, jugular vein; arrow
head, lipid-rich necrotic
core (Reprinted with
permission from Balu
et al. [51])
106 J. Sun et al.

Ultrasmall super paramagnetic particles of iron analysis software, CASCADE, developed at the
oxides (USPIO) are a class of iron oxide magnetic University of Washington, has processing rou-
particles that are also used for imaging plaque tines for multi-contrast image registration and
inflammation [25]. After systemic injection, semiautomated lumen and wall detection [27]. In
USPIOs are concentrated in macrophages and order to assist readers in measurement of plaque
can be detected by their strong magnetic components, a morphology-enhanced probabilis-
susceptibility. tic plaque segmentation (MEPPS) algorithm can
Long TE gradient echo sequences can be used be used to automatically segment plaque compo-
to detect the T2 or T2* effect of USPIOs. Similar nents [28]. The component boundaries are reader
to gadolinium contrast injection, areas of plaque correctable if required. Following the analysis,
with macrophages can be identified by comparing total volume of plaque components as well as
pre and post USPIO injection. In contrast to gad- percent volumes relative to plaque volume can
olinium, areas of macrophage concentration are be calculated. A 3D distribution of plaque com-
identified by loss of signal, and the signal differ- ponents and plaque burden can also be used for
ence is maximized at 36 h as compared to assessment as shown in Fig. 5.
5–10 min for gadolinium. Use of positive suscep-
tibility contrast techniques may improve detection
of USPIOs [26]. Current Applications

Imaging-Based Patient Stratification

Measurement of Plaque Morphology
and Composition Patients with carotid atherosclerosis are managed
with stringent risk factor control that covers life-
Several important measurements can be derived style changes and medications. Nonetheless,
from a multi-contrast plaque MRI protocol. Out- revascularization could be highly beneficial to
lines of the lumen and outer wall on black-blood those with a high risk of future stroke. However,
MRI provide a measure of the plaque burden. The this potential benefit needs to be evaluated against
wall area defined as the difference between the non-negligible surgical risk for complications
two outlines provides a first measurement of such as perioperative myocardial infarction and
plaque in the vessel wall. Plaque area normalized distal embolism on an individual basis. In the
for vessel size by the outer wall area provides the NASCET study which compared carotid endarter-
normalized wall index (NWI), which is similar to ectomy with medical treatment alone in patients
percent atheroma volume used in intravascular with transient ischemic attack or nondisabling ische-
ultrasound imaging of coronary atherosclerosis. mic stroke within the previous 6 months, there
While plaque burden can be measured on a single was an absolute risk reduction of 17  3.5 % at
contrast weighting such as pre-contrast black- 2 years favoring surgery in patients with ipsilateral
blood T1-WI sequence, outlining plaque bound- high-grade stenosis (70–99 %) [29]. In contrast,
aries on multi-contrast MRI is more robust for patients with mild stenosis (<50 %) did not benefit
advanced lesions in view of superficial calcifica- from surgery, although the remaining risk with med-
tion and plaque-mimicking flow artifacts. ical therapy was still considerable at 18.7 % over
Measurement of plaque composition requires a 5 years. Risk reduction with carotid endarterectomy
series of steps for multi-contrast image review. in asymptomatic patients was more moderate, even
These typically include multi-contrast image reg- in the presence of high-grade stenosis, as shown in
istration and segmentation of plaque component ACAS and the Asymptomatic Carotid Stenosis
boundaries. Due to the small components and Trial (ACST), which estimated that about 17 patients
complex nature of plaque constituents, special need to be treated with carotid endarterectomy to
algorithms targeted at atherosclerotic plaque anal- prevent one stroke in 5 years [30, 31]. Improvement
ysis are typically used. For example, the plaque in cost-effectiveness of carotid endarterectomy
7 Imaging of Carotid Atherosclerosis 107

Fig. 5 3D volume rendering of vessel wall morphology plaque burden and components. Orange indicates
and plaque components. Segmentation results of vessel intraplaque hemorrhage (arrow); blue, calcification; yel-
wall morphology and plaque components can be displayed low, lipid-rich necrotic core
using 3D volume rendering to illustrate distribution of

in asymptomatic patients is urgent in light of the MRI-detected plaque characteristics to increase

much larger patient population with asymptomatic survival benefits from carotid endarterectomy.
carotid atherosclerosis. In fact, 76.7 % of the In a study comparing symptomatic and asymp-
795,000 strokes that happen each year are first tomatic carotid plaques that cause at least 50 %
attacks. stenosis, multi-contrast MRI was used to compre-
The unmet need to identify who is at risk of hensively characterize plaque morphological and
future stroke and therefore would benefit from compositional features [32]. Fibrous cap rupture
surgical intervention has highlighted the impor- as well as large necrotic core and intraplaque
tance of imaging-based patient stratification in hemorrhage were more prevalent in symptomatic
clinical decision making. The aforementioned plaques, whereas no difference was found in lumi-
clinical trials were performed at a time when the nal stenosis or plaque volumes. As such, plaque
assessment of carotid atherosclerosis was largely compositional features appear to further differen-
based on angiography for measuring luminal tiate symptomatic plaques from those with a
narrowing as the sole criterion to inform surgeons. benign natural history among plaques causing
Since then, the understanding of carotid athero- severe stenosis. The convenience of using a single
sclerosis biology has advanced to appreciate the T1-WI imaging sequence to detect intraplaque
fundamental role of plaque rupture and thrombo- hemorrhage has led to studies that specifically
sis in the pathophysiology of neurological ische- examined the association of intraplaque hemor-
mic events. In the new paradigm built on plaque rhage with cerebrovascular events in broad
instability, luminal stenosis is more appropriately populations. Yamada et al. studied 392 carotid
viewed as one of the many risk factors for plaque arteries from 222 patients with suspected or con-
rupture rather than the primary cause of stroke. A firmed carotid atherosclerosis using the MPRAGE
number of morphological and pathological fea- sequence [33]. Carotid arteries were categorized
tures that differentiate ruptured from as mild (0–29 %), moderate (30–69 %), or severe
non-ruptured plaques can now be characterized (70–99 %) stenosis measured on contrast-
by carotid MRI, which holds the promise to be enhanced MR angiography. They noted that
used in prospective clinical trials that use carotid plaques with high signal intensity,
108 J. Sun et al.

indicative of intraplaque hemorrhage, were more Carotid Artery Stenting

likely to have had ipsilateral ischemic events within
the previous 6 months. The association between Carotid artery stenting has become a reasonable
ischemic events and MRI-detected ipsilateral alternative to carotid endarterectomy when
intraplaque hemorrhage was strongest in the patients are considered for revascularization due
group with moderate stenosis, implying that infor- to its less invasive nature. A few randomized trials
mation on plaque characteristics may be particu- have compared the two approaches in terms of
larly helpful in selecting high-risk patients where clinical outcomes including perioperative compli-
assessment of luminal stenosis is not able to cations and long-term event rates. Although long-
do. Howarth et al. compared ten symptomatic term event rates are comparable between the two,
patients and ten asymptomatic patients with a it has been noted that carotid artery stenting has a
focus on plaque inflammation measured with higher perioperative risk of embolic stroke com-
USPIO-enhanced MRI [25]. More regions with pared to carotid endarterectomy [39]. This has
signal drop, indicative of more extensive macro- raised concerns for carotid artery stenting, as
phage infiltration, were observed in symptomatic those perioperative microembolizations adversely
patients. These retrospective imaging studies affect patients’ cognitive function and has been
have reaffirmed the role of plaque instability in shown to lower the quality of life after procedure.
the pathophysiology of ischemic stroke as noted Plaque composition and thrombogenicity could
in histopathological studies and have extended the be a major determinant of the risk of distal embo-
finding to carotid atherosclerosis without severe lization in the setting of carotid artery stenting.
stenosis. Yoshimura et al. tested the hypothesis in a study
The prognostic value of plaque characteristics involving 112 patients undergoing carotid artery
has been demonstrated in prospective studies of stenting [40]. Diffusion-weighted imaging was
both symptomatic and asymptomatic patient performed before and after stenting to detect
cohorts. Takaya et al. followed 154 consecutive new ischemic lesions due to intervention.
patients with asymptomatic 50–79 % stenosis. Pre-operation time-of-flight MR angiography
Baseline plaque characteristics, including thin/ images were used to identify ipsilateral plaques
ruptured fibrous cap, intraplaque hemorrhage, with intraplaque hemorrhage. Both new ischemic
larger hemorrhage area, large necrotic core, and lesions and perioperative symptoms were found to
maximal wall thickness, were predictive of subse- be more frequent in patients with intraplaque
quent ischemic events [34]. Most of these features hemorrhage. Carotid MRI performed prior to sur-
appear to predict recurrent strokes as well, as gical intervention therefore may contribute to an
shown by Kwee et al. in a study involving accurate evaluation of the risk of perioperative
126 symptomatic patients with 30–69 % stenosis microembolization, which can be translated into
[35]. In particular, MRI detection of intraplaque improved clinical outcomes of both carotid artery
hemorrhage has been shown in various clinical stenting and carotid endarterectomy.
populations to be associated with increased
atherothrombotic risk. Three independent meta-
analyses were recently published which estimated Cryptogenic Stroke
the predictive value of MRI-detected intraplaque
hemorrhage for planning future observational As many as a third of ischemic strokes are con-
studies and clinical trials [36–38]. Intraplaque sidered cryptogenic, which means that no definite
hemorrhage alone appears to be associated with etiology can be determined after clinical workup.
a more than tenfold increased risk of neurological Imaging carotid atherosclerosis with MRI may
ischemic events in symptomatic patients. enable visualization of telltale features of culprit
Leveraging such prognostic information assess- lesions, which will facilitate targeted secondary
able via carotid MRI could prove to be invaluable prevention strategies for individual patients by
in future clinical practice. revealing or excluding large artery atherosclerosis
7 Imaging of Carotid Atherosclerosis 109

as the underlying etiology. Freilinger classification based on plaque morphology and

et al. performed a preliminary study in a series of composition. Several studies have utilized MR
32 consecutive patients with clinically diagnosed imaging to look into the association of traditional
cryptogenic stroke and nonstenosing (<50 %) and novel atherosclerotic risk factors with plaque
carotid atherosclerosis [41]. A multi-contrast composition, such as calcification, necrotic core,
MRI protocol detected complicated lesions, and intraplaque hemorrhage, in population-based
including fibrous cap rupture, intraplaque hemor- study samples. Wasserman et al. studied 214 sub-
rhage, and mural thrombus, in 12 (37.5 %) jects with carotid atherosclerosis enrolled in the
plaques ipsilateral to the cryptogenic stroke, Multi-Ethnic Study of Atherosclerosis (MESA)
whereas none was found on the contralateral side study [42]. After adjustment for potential
which showed MR signals consistent with confounding factors, subjects in the middle and
fibroatheroma and pathological intimal thickening highest tertile of total plasma cholesterol had a
as the predominant lesion type. higher likelihood of showing a lipid core com-
For the purpose of differential diagnosis in pared to the lowest tertile (odds ratio: 2.76 and
patients presenting with cryptogenic stroke, it is 4.63, respectively). The MRI study of the
preferable to use a protocol that affords extended Rotterdam study participants found that
coverage as patients can have multiple tandem intraplaque hemorrhage was associated with a
lesions and the location of culprit features, if number of clinical risk factors including age,
any, is not known a priori. The newly developed sex, hypertension, and smoking, whereas necrotic
3D gradient echo-based sequences are showing core was associated with sex and
capability to cover both extracranial and intracra- hypercholesterolemia [43].
nial carotid arteries, which may be a viable solu- Lesions in animal models are usually in the
tion in the near future. early stage of atherosclerosis and lack important
advanced features seen in human lesions. This has
led to a significant knowledge gap in understand-
In Vivo Insights on Pathobiological ing how atherosclerosis progresses from subclin-
Mechanisms ical lesions to clinically overt, culprit lesions.
Asymptomatic carotid atherosclerosis coupled
As an increasing number of biological targets with in vivo imaging offers a unique opportunity
implicated in the pathogenesis of atherosclerosis for clinical researchers to advance understanding
can now be successfully imaged with advances in of important pathological events leading to clini-
MR sequence design and post-processing, carotid cal events. In a case-control study comparing
MRI has become a powerful tool for studying carotid plaques with and without intraplaque hem-
disease mechanisms. The advantages of carotid orrhage, an 18-month imaging follow-up demon-
MRI as a research tool for clinical studies include strated distinctively different progression patterns
its noninvasive nature, absence of ionizing radia- between the two groups [44]. While plaques with-
tion, high reproducibility facilitating serial imag- out intraplaque hemorrhage tended to be station-
ing design, and the opportunity of combining ary over the study period, those with intraplaque
different contrast weightings for better plaque hemorrhage showed a larger increase in wall vol-
characterization. ume and necrotic core volume, which was trans-
Epidemiological studies have traditionally lated mainly into luminal narrowing. The ability
used carotid ultrasound or coronary calcium scan of MRI to serially track changes of carotid athero-
to measure subclinical atherosclerosis. Recent sclerosis was shown in a recent study, where mul-
studies have attempted to introduce MR plaque tiple MRI scans with a constant time interval were
imaging to epidemiological studies. This is a retrospectively retrieved from subjects with
novel concept as MR plaque imaging not only intraplaque hemorrhage to examine the immediate
provides measurement of subclinical atheroscle- and long-term effects of intraplaque hemorrhage
rotic burden but also allows phenotypic on plaque progression [45]. Images spanning 4–5
110 J. Sun et al.

years successfully captured how plaques Efficacy Markers for Testing

progressed from the stage without intraplaque Therapeutic Strategies
hemorrhage to the stage with intraplaque hemor-
rhage. Rather than a transient mass effect on The ability of MRI to track morphological and
plaque volume, the development of intraplaque compositional changes of atherosclerotic plaques
hemorrhage was found to have gradually altered has also led to research efforts using quantitative
the trajectory of plaque progression by posing an MRI measurements of carotid atherosclerosis
immediate and long-lasting promoting effect on as efficacy markers for testing therapeutic
plaque progression. Given that animal models do strategies that aim to stabilize atherosclerotic
not mimic advanced, clinical relevant plaques, plaques (Fig. 6). As a surrogate for systemic ath-
knowledge so-obtained from imaging human erosclerosis, information on carotid atherosclero-
carotid atherosclerosis is invaluable to the under- sis should give clues to the underlying coronary
standing of the natural history of high-risk atherosclerosis, in which vessel wall information
plaques, which may point to promising directions currently obtainable is still limited due to
for reducing residual cardiovascular risk in con- technical constraints. This may be particularly
temporary medicine. important in the current era due to several reasons.

Fig. 6 Serial MR scans showing plaque stabilization same plaque, consistent with plaque stabilization. The
under pharmacotherapy. Top panel: three consecutive distinct lumen (L ) and outer wall boundaries (long arrows)
cross sections from the baseline scan show a plaque with suggest a reduction in plaque volume. However, the most
a big necrotic core, which is delineated as non-enhanced dramatic change is the decrease in non-enhanced areas,
areas on contrast-enhanced T1-WI (CE-T1) images. Bot- indicative of smaller necrotic core size, at follow-up. A
tom panel: Matched cross sections from the follow-up scan region with thin fibrous cap at baseline had increased cap
show morphological and compositional changes of the thickness at follow-up (hollow arrows)
7 Imaging of Carotid Atherosclerosis 111

First, the prevalence of atherosclerosis is high, rosuvastatin on carotid plaque morphology and
but the incidence of atherothrombotic complica- composition in more advanced lesions causing
tions of any given population is low, especially in 16–79 % stenosis [48]. A significant reduction in
the primary prevention setting. Consequently, necrotic core size was noted at 2 years, but not in
while a high volume of myocardial infarction wall volume, which suggests plaque lipid deple-
and ischemic strokes are seen clinically, the tion, and thus stability, may be one of the under-
large number of subjects needed to power a lying reasons for clinical benefits with statin
clinical trial has been daunting. Thus, early sig- therapy. The time course of plaque lipid depletion
nals from surrogate markers that can display under lipid-lowering treatment was later
the efficacy of experimental agents are highly addressed by Zhao et al. [49]. In 33 subjects
desirable. Second, novel surrogate markers, with documented necrotic core at baseline,
to certain extent, target biological activities or necrotic core size gradually decreased over
pathways in atherogenesis. Characterizing 3 years. While a significant reduction in necrotic
changes in those biomarkers will help clarify core could be seen in year 1, reduction in wall
the mechanisms of actions exerted by the volume was not apparent until year 2, lagging
experimental agent, which often deviate behind the change in necrotic core. When plaque
from what are known from preclinical studies. inflammation is targeted, previous studies have
Third, it is common for patients with different shown a reduction in imaging signals within a
demographics or clinical conditions to have het- few weeks [50]. Tang et al. compared two doses
erogeneous responses to the same treatment, of atorvastatin treatment in 47 patients with over
which has led to the common practice of 40 % carotid stenosis and USPIO uptake at base-
predefined subgroup analysis in clinical line. The high-dose group showed a significant
trials. Given the dramatic morphological and reduction in USPIO-defined inflammation at
pathological differences, plaques at different 6 weeks [50].
stages or with different phenotypes are likely to The aforementioned studies are all single cen-
respond heterogeneously to the same medication, ter studies. However, in cases of large-scale
which, however, is less understood. Intraplaque recruitment, short recruitment period, or clinical
hemorrhage appears to be one such factor conditions with low prevalence, multicenter
influencing therapeutic response [46]. In this design is usually the norm. In view of the substan-
regard, plaque characteristics may not only tial variations in technique instrumentation
serve as efficacy markers but also enable a between imaging sites, the recent completion of
deeper understanding of the efficacy of a few multicenter carotid MRI studies represents a
antiatherogenic medications in terms of individual solid step toward the standardization of MRI
responses. protocol and quantitative image analysis. A mul-
A number of studies have been conducted ticenter clinical trial involving 13 imaging
using MRI-based plaque measurements as end sites and two vendor platforms at 3-T was carried
points, predominantly using statins. Corti out to examine change in necrotic core size of
et al. studied 32 carotid plaques under the carotid atherosclerosis over a short 6-month
treatment of simvastatin [47]. Reduction in wall period [46]. Under current standard-of-care to
area but not lumen area was seen at 1 year. When medically control modifiable cardiovascular risk
follow-up was extended to 2 years, there was a factors, it was shown that plaques without
further decrease in wall area and a slight intraplaque hemorrhage had a small but signifi-
increase in lumen area. These findings indicate cant reduction in necrotic core size. In contrast,
that carotid atherosclerosis is subject to rapid plaques with intraplaque hemorrhage had an
progression or regression. Change in wall area increase in necrotic core size and wall volume
appeared to precede change in lumen area. and a decrease in lumen volume – a progression
Another study by Underhill et al. used pattern consistent with previous natural history
multi-contrast MRI to examine the effect of studies.
112 J. Sun et al.

Summary References

Carotid atherosclerosis is a common clinical find- 1. de Weerd M, Greving JP, de Jong AW, Buskens E, Bots
ML (2009) Prevalence of asymptomatic carotid artery
ing, which almost always involves the bifurcation stenosis according to age and sex: systematic review
segment where recirculating flow is most and metaregression analysis. Stroke 40:1105–1113
severe. Vessel wall MRI, with high soft tissue 2. Cai JM, Hatsukami TS, Ferguson MS, Kerwin WS,
contrast and flexibility of sequence design, affords Saam T, Chu BC, Takaya N, Polissar NL, Yuan C
(2005) In vivo quantitative measurement of intact
inherent advantages for imaging carotid
fibrous cap and lipid-rich necrotic core size in athero-
atherosclerosis, yet several key technical points sclerotic carotid plaque: comparison of high-
must be addressed to achieve satisfactory visual- resolution, contrast-enhanced magnetic resonance
ization of vessel wall pathology, including flow imaging and histology. Circulation 112:3437–3444
3. Moody AR, Murphy RE, Morgan PS, Martel AL,
and fat suppression, submillimeter spatial resolu-
Delay GS, Allder S, MacSweeney ST, Tennant WG,
tion, and sufficient SNR within clinically applica- Gladman J, Lowe J, Hunt BJ (2003) Characterization
ble scan times. Traditional 2D imaging has gained of complicated carotid plaque with magnetic resonance
success in providing clinically relevant informa- direct thrombus imaging in patients with cerebral
ischemia. Circulation 107:3047–3052
tion of atherosclerosis, demonstrated by a broad 4. Kerwin WS, O’Brien KD, Ferguson MS, Polissar N,
spectrum of applications in previous literature. Hatsukami TS, Yuan C (2006) Inflammation in carotid
But newly developed 3D sequences hold the atherosclerotic plaque: a dynamic contrast-enhanced
promise for future clinical practice due to short- MR imaging study. Radiology 241:459–468
5. Trivedi RA, Mallawarachi C, U-King-Im JM, Graves
ened scan times and increased scan coverage.
MJ, Horsley J, Goddard MJ, Brown A, Wang L,
Atherosclerotic plaques display considerable Kirkpatrick PJ, Brown J, Gillard JH (2006) Identifying
morphological and pathological variations. inflamed carotid plaques using in vivo USPIO-
Appropriate interpretation of image findings may enhanced MR imaging to label plaque macrophages.
Arterioscler Thromb Vasc Biol 26:1601–1606
depend on a sensible understanding of both
6. Virmani R, Kolodgie FD, Burke AP, Farb A, Schwartz
pathology and the capability and limitations SM (2000) Lessons from sudden coronary death – a
of MRI. comprehensive morphological classification scheme
With the changing landscape in the field of for atherosclerotic lesions. Arterioscler Thromb Vasc
Biol 20:1262–1275
atherosclerosis management including pharmaco-
7. Leitinger N, Schulman IG (2013) Phenotypic polariza-
therapy and surgical intervention, clinicians are tion of macrophages in atherosclerosis. Arterioscler
facing either a shortage of data or data that are Thromb Vasc Biol 33:1120–1126
not generalizable to current practice, when 8. Virmani R, Ladich ER, Burke AP, Kolodgie FD (2006)
Histopathology of carotid atherosclerotic disease. Neu-
selecting between medical treatment, carotid end-
rosurgery 59:S219–S227, S3-S13
arterectomy, and carotid artery stenting. Further- 9. Spagnoli LG, Mauriello A, Sangiorgi G, Fratoni S,
more, the limitations of luminal stenosis in Bonanno E, Schwartz RS, Piepgras DG, Pistolese R,
predicting patients’ future risk of ischemic stroke Ippoliti A, Holmes DR (2004) Extracranial
thrombotically active carotid plaque as a risk factor
are increasingly recognized. New opportunities
for ischemic stroke. JAMA 292:1845–1852
have emerged that leverage technical develop- 10. Fisher M, Paganini-Hill A, Martin A, Cosgrove M,
ments in MRI of carotid atherosclerosis to assist Toole JF, Barnett HJ, Norris J (2005) Carotid plaque
clinical decision making. As such, management of pathology: thrombosis, ulceration, and stroke patho-
genesis. Stroke 36:253–257
patients with symptomatic or asymptomatic
11. Redgrave J, Lovett JK, Gallagher PJ, Rothwell PM
carotid atherosclerosis is awaiting significant par- (2006) Histological assessment of 526 symptomatic
adigm shift. With continued efforts in clinical carotid plaques in relation to the nature and timing of
studies using direct MR imaging of carotid ath- ischemic symptoms – the Oxford plaque study. Circu-
lation 113:2320–2328
erosclerosis, the clinical outcomes of carotid ath-
12. Narula J, Nakano M, Virmani R, Kolodgie FD,
erosclerosis may eventually be improved to a new Petersen R, Newcomb R, Malik S, Fuster V, Finn AV
level. (2013) Histopathologic characteristics of
7 Imaging of Carotid Atherosclerosis 113

atherosclerotic coronary disease and implications of atherosclerotic disease evaluation. Magn Reson Med
the findings for the invasive and noninvasive detection 69:337–345
of vulnerable plaques. J Am Coll Cardiol 25. Howarth SP, Tang TY, Trivedi R, Weerakkody R,
61:1041–1051 U-King-Im J, Gaunt ME, Boyle JR, Li ZY, Miller SR,
13. Yarnykh VL, Terashima M, Hayes CE, Shimakawa A, Graves MJ, Gillard JH (2009) Utility of USPIO-
Takaya N, Nguyen PK, Brittain JH, McConnell MV, enhanced MR imaging to identify inflammation and
Yuan C (2006) Multicontrast black-blood MRI of the fibrous cap: a comparison of symptomatic and
carotid arteries: comparison between 1.5 and 3 Tesla asymptomatic individuals. Eur J Radiol 70:555–560
magnetic field strengths. J Magn Reson Imaging 26. Stuber M, Gilson WD, Schar M, Kedziorek DA, Hof-
23:691–698 mann LV, Shah S, Vonken EJ, Bulte JW, Kraitchman
14. Balu N, Yarnykh VL, Scholnick J, Chu B, Yuan C, DL (2007) Positive contrast visualization of iron oxide-
Hayes C (2009) Improvements in carotid plaque imag- labeled stem cells using inversion-recovery with
ing using a new eight-element phased array coil at 3 T. ON-resonant water suppression (IRON). Magn Reson
J Magn Reson Imaging 30:1209–1214 Med 58:1072–1077
15. Parker DL, Goodrich KC, Masiker M, Tsuruda JS, 27. Kerwin WS, Xu D, Liu F, Saam T, Underhill HR,
Katzman GL (2002) Improved efficiency in double- Takaya N, Chu BC, Hatsukami TS, Yuan C (2007)
inversion fast spin-echo imaging. Magn Reson Med Magnetic resonance imaging of carotid atherosclerosis:
47:1017–1021 plaque analysis. Top Magn Reson Imaging
16. Yarnykh VL, Yuan C (2003) Multislice double 18:371–378
inversion-recovery black-blood imaging with simulta- 28. Liu F, Xu DX, Ferguson MS, Chu BC, Saam T,
neous slice reinversion. J Magn Reson Imaging Takaya N, Hatsukami TS, Yuan C, Kerwin WS
17:478–483 (2006) Automated in vivo segmentation of carotid
17. Wang J, Yarnykh VL, Hatsukami T, Chu B, Balu N, plaque MRI with morphology-enhanced probability
Yuan C (2007) Improved suppression of plaque- maps. Magn Reson Med 55:659–668
mimicking artifacts in black-blood carotid atheroscle- 29. Barnett HJ, Taylor DW, Eliasziw M, Fox AJ, Ferguson
rosis imaging using a multislice motion-sensitized GG, Haynes RB, Rankin RN, Clagett GP, Hachinski
driven-equilibrium (MSDE) turbo spin-echo (TSE) VC, Sackett DL, Thorpe KE, Meldrum HE, Spence JD
sequence. Magn Reson Med 58:973–981 (1998) Benefit of carotid endarterectomy in patients
18. Wang J, Yarnykh VL, Yuan C (2010) Enhanced image with symptomatic moderate or severe stenosis. North
quality in black-blood MRI using the improved American Symptomatic Carotid Endarterectomy Trial
motion-sensitized driven-equilibrium (iMSDE) Collaborators. N Engl J Med 339:1415–1425
sequence. J Magn Reson Imaging 31:1256–1263 30. Walker MD, Marler JR, Goldstein M, Grady PA, Toole
19. Li L, Miller KL, Jezzard P (2012) DANTE-prepared JF, Baker WH, Castaldo JE, Chambless LE, Moore
pulse trains: a novel approach to motion-sensitized and WS, Robertson JT, Young B, Howard VJ, Purvis S,
motion-suppressed quantitative magnetic resonance Vernon DD, Needham K, Beck P, Celani VJ,
imaging. Magn Reson Med 68:1423–1438 Sauerbeck L, von Rajcs JA, Atkins D (1995) Endarter-
20. Balu N, Yarnykh VL, Chu B, Wang J, Hatsukami T, ectomy for asymptomatic carotid artery stenosis.
Yuan C (2011) Carotid plaque assessment using fast 3D JAMA 273:1421–1428
isotropic resolution black-blood MRI. Magn Reson 31. Halliday A, Mansfield A, Marro J, Peto C, Peto R,
Med 65:627–637 Potter J, Thomas D (2004) Prevention of disabling
21. Fan Z, Zhang Z, Chung YC, Weale P, Zuehlsdorff S, and fatal strokes by successful carotid endarterectomy
Carr J, Li D (2010) Carotid arterial wall MRI at 3T in patients without recent neurological symptoms:
using 3D variable-flip-angle turbo spin-echo (TSE) randomised controlled trial. Lancet 363:1491–1502
with flow-sensitive dephasing (FSD). J Magn Reson 32. U-King-Im JM, Tang TY, Patterson A, Graves MJ,
Imaging 31:645–654 Howarth S, Li ZY, Trivedi R, Bowden D, Kirkpatrick
22. Yarnykh VL, Yuan C (2002) T-1-insensitive flow sup- PJ, Gaunt ME, Warburton EA, Antoun NM, Gillard JH
pression using quadruple inversion-recovery. Magn (2008) Characterisation of carotid atheroma in symp-
Reson Med 48:899–905 tomatic and asymptomatic patients using high resolu-
23. Ota H, Yarnykh VL, Ferguson MS, Underhill HR, tion MRI. J Neurol Neurosurg Psychiatry 79:905–912
DeMarco JK, Zhu DC, Oikawa M, Dong L, Zhao 33. Yamada N, Higashi M, Otsubo R, Sakuma T,
XH, Collar A, Hatsukami TS, Yuan C (2010) Carotid Oyama N, Tanaka R, Iihara K, Naritomi H,
intraplaque hemorrhage imaging at 3.0-T MR imaging: Minematsu K, Naito H (2007) Association between
comparison of the diagnostic performance of three signal hyperintensity on T1-weighted MR imaging of
T1-weighted sequences. Radiology 254:551–563 carotid plaques and ipsilateral ischemic events. AJNR
24. Wang J, Bornert P, Zhao H, Hippe DS, Zhao X, Balu N, Am J Neuroradiol 28:287–292
Ferguson MS, Hatsukami TS, Xu J, Yuan C, Kerwin 34. Takaya N, Yuan C, Chu BC, Saam T, Underhill H, Cai
WS (2013) Simultaneous noncontrast angiography and JM, Tran N, Polissar NL, Isaac C, Ferguson MS, Gar-
intraPlaque hemorrhage (SNAP) imaging for carotid den GA, Cramer SC, Maravilla KR, Hashimoto B,
114 J. Sun et al.

Hatsukami TS (2006) Association between carotid Determinants of magnetic resonance imaging detected
plaque characteristics and subsequent ischemic cere- carotid plaque components: the Rotterdam study. Eur
brovascular events: a prospective assessment with MRI Heart J 33:221–229
– initial results. Stroke 37:818–823 44. Takaya N, Yuan C, Chu BC, Saam T, Polissar NL,
35. Kwee RM, van Oostenbrugge RJ, Mess WH, Prins Jarvik GP, Isaac C, McDonough J, Natiello C,
MH, van der Geest RJ, Ter Berg JW, Franke CL, Small R, Ferguson MS, Hatsukami TS (2005) Presence
Korten AG, Meems BJ, van Engelshoven JM, of intraplaque hemorrhage stimulates progression of
Wildberger JE, Kooi ME (2013) MRI of carotid ath- carotid atherosclerotic plaques: a high-resolution mag-
erosclerosis to identify TIA and stroke patients who are netic resonance imaging study. Circulation
at risk of a recurrence. J Magn Reson Imaging 111:2768–2775
37:1189–1194 45. Sun J, Underhill HR, Hippe DS, Xue Y, Yuan C,
36. Saam T, Hetterich H, Hoffmann V, Yuan C, Hatsukami TS (2012) Sustained acceleration in carotid
Dichgans M, Poppert H, Koeppel T, Hoffmann U, atherosclerotic plaque progression with intraplaque
Reiser MF, Bamberg F (2013) Meta-analysis and sys- hemorrhage: a long-term time course study. J Am
tematic review of the predictive value of carotid plaque Coll Cardiol Img 5:798–804
hemorrhage on cerebrovascular events by magnetic 46. Sun J, Balu N, Hippe DS, Xue Y, Dong L, Zhao X, Li F,
resonance imaging. J Am Coll Cardiol 62:1081–1091 Xu D, Hatsukami TS, Yuan C (2013) Subclinical
37. Hosseini AA, Kandiyil N, Macsweeney ST, Altaf N, carotid atherosclerosis: short-term natural history of
Auer DP (2013) Carotid plaque hemorrhage on mag- lipid-rich necrotic core–a multicenter study with MR
netic resonance imaging strongly predicts recurrent imaging. Radiology 268:61–68
ischemia and stroke. Ann Neurol 73:774–784 47. Corti R, Fuster V, Fayad ZA, Worthley SG, Helft G,
38. Gupta A, Baradaran H, Schweitzer AD, Kamel H, Smith D, Weinberger J, Wentzel J, Mizsei G,
Pandya A, Delgado D, Dunning A, Mushlin AI, Sanelli Mercuri M, Badimon JJ (2002) Lipid lowering by
PC (2013) Carotid plaque MRI and stroke risk: a sys- simvastatin induces regression of human atheroscle-
tematic review and meta-analysis. Stroke rotic lesions: two years’ follow-up by high-resolution
44:3071–3077 noninvasive magnetic resonance imaging. Circulation
39. Brott TG, Hobson RW, Howard G, Roubin GS, Clark 106:2884–2887
WM, Brooks W, Mackey A, Hill MD, Leimgruber PP, 48. Underhill HR, Yuan C, Zhao XQ, Kraiss LW, Parker
Sheffet AJ, Howard VJ, Moore WS, Voeks JH, Hop- DL, Saam T, Chu B, Takaya N, Liu F, Polissar NL,
kins LN, Cutlip DE, Cohen DJ, Popma JJ, Ferguson Neradilek B, Raichlen JS, Cain VA, Waterton JC,
RD, Cohen SN, Blackshear JL, Silver FL, Mohr JP, Lal Hamar W, Hatsukami TS (2008) Effect of rosuvastatin
BK, Meschia JF (2010) Stenting versus endarterec- therapy on carotid plaque morphology and composi-
tomy for treatment of carotid-artery stenosis. N Engl J tion in moderately hypercholesterolemic patients: a
Med 363:11–23 high-resolution magnetic resonance imaging trial. Am
40. Yoshimura S, Yamada K, Kawasaki M, Asano T, Heart J 155:581–584
Kanematsu M, Takamatsu M, Hara A, Iwama T 49. Zhao XQ, Dong L, Hatsukami T, Phan BA, Chu B,
(2011) High-intensity signal on time-of-flight magnetic Moore A, Lane T, Neradilek MB, Polissar N,
resonance angiography indicates carotid plaques at Monick D, Lee C, Underhill H, Yuan C (2011) MR
high risk for cerebral embolism during stenting. Stroke imaging of carotid plaque composition during
42:3132–3137 lipid-lowering therapy: a prospective assessment of
41. Freilinger TM, Schindler A, Schmidt C, Grimm J, effect and time course. J Am Coll Cardiol Img
Cyran C, Schwarz F, Bamberg F, Linn J, Reiser M, 4:977–986
Yuan C, Nikolaou K, Dichgans M, Saam T (2012) 50. Tang TY, Howarth SP, Miller SR, Graves MJ, Patterson
Prevalence of nonstenosing, complicated atheroscle- AJ, U-King-Im JM, Li ZY, Walsh SR, Brown AP,
rotic plaques in cryptogenic stroke. J Am Coll Cardiol Kirkpatrick PJ, Warburton EA, Hayes PD, Varty K,
Img 5:397–405 Boyle JR, Gaunt ME, Zalewski A, Gillard JH (2009)
42. Wasserman BA, Sharrett AR, Lai S, Gomes AS, The ATHEROMA (Atorvastatin Therapy: Effects on
Cushman M, Folsom AR, Bild DE, Kronmal RA, Reduction of Macrophage Activity) study. Evaluation
Sinha S, Bluemke DA (2008) Risk factor associations using ultrasmall superparamagnetic iron oxide-
with the presence of a lipid core in carotid plaque of enhanced magnetic resonance imaging in carotid dis-
asymptomatic individuals using high-resolution MRI: ease. J Am Coll Cardiol 53:2039–2050
the multi-ethnic study of atherosclerosis (MESA). 51. Balu N, Wang J, Dong L, Baluyot F, Chen H, Yuan C
Stroke 39:329–335 (2009) Current techniques for MR imaging of
43. van den Bouwhuijsen QJ, Vernooij MW, Hofman A, atherosclerosis. Top Magn Reson Imaging
Krestin GP, van der Lugt A, Witteman JC (2012) 20(4):203–215
Carotid Artery Dissection
Olivier Naggara, Myriam Edjlali-Goujon, Eric Bodiguel,
Marie Pierre Gobin-Metteil, Denis Trystram,
Christine Rodriguez-Regent, Jean-Louis Mas,
Jean Francois Meder, and Catherine Oppenheim

Contents Mural Hematoma Consequences . . . . . . . . . . . . . . . . . . . . 119

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 Clinical Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Local Symptoms and Signs . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 Ischemic Symptoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Pathogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 Outcome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Traumatic Carotid Artery Dissection . . . . . . . . . . . . . . . . 117 Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
Spontaneous Carotid Artery Dissection . . . . . . . . . . . . . 117 Digital Subtraction Angiography . . . . . . . . . . . . . . . . . . . . 122
Insights on the Mural Hematoma . . . . . . . . . . . . . . . . . 118 Ultrasound . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
Mural Hematoma Constitution . . . . . . . . . . . . . . . . . . . . . . 118 Cross-Sectional Imaging: CT or MRI? . . . . . . . . . . . . . . 123
Mural Hematoma Location . . . . . . . . . . . . . . . . . . . . . . . . . . 119 Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Medical Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Mechanical Revascularization Strategies . . . . . . . . . . . . 135

O. Naggara (*) Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

Service de Neuroradiologie, Centre Hospitalier Sainte- References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Anne, INSERM UMR894, Université Paris Descartes,
Sorbonne Paris Cité, Paris, France
Service de Radiologie pédiatrique, Faculté de Médecine,
Hôpital Necker Enfants Malades, Université Paris
Descartes, Sorbonne Paris Cité, Paris, France
M. Edjlali-Goujon • M.P. Gobin-Metteil • D. Trystram •
C. Rodriguez-Regent • J.F. Meder • C. Oppenheim
Service de Neuroradiologie, Centre Hospitalier Sainte-
Anne, INSERM UMR894, Université Paris Descartes,
Sorbonne Paris Cité, Paris, France
E. Bodiguel • J.-L. Mas
Service de Neurologie, Faculté de Médecine, Centre
Hospitalier Sainte-Anne, Université Paris Descartes,
Sorbonne Paris Cité, Paris, France

# Springer Science+Business Media New York 2016 115

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_16
116 O. Naggara et al.

Abstract Keywords
Carotid artery dissections (CAD) are a leading Carotid artery dissection • Mural hematoma •
cause of nonatherosclerotic stroke in young High resolution MRI • Stroke
adults, responsible for up to 25 % of ischemic
strokes. The cornerstone of CAD pathophysiol-
ogy and diagnosis is the presence of a mural Introduction
hematoma of unknown etiology, possibly
caused by an intimal tear, a primary rupture of Carotid artery dissection (CAD) is caused by the
vasa vasorum, or an underlying arteriopathy formation of a mural hematoma, of the subpetrous
impairing vasomotion. This mural hematoma segment of the internal carotid artery (ICA).
occurs within the media layer extending distally Although the clinical presentation of CAD can
and circumferentially. The diagnosis of CAD is be benign, cervical dissections are a major cause
generally based on a suggestive clinical presen- of ischemic stroke in middle-aged adults [1]. The
tation, exclusion of atherosclerosis, and support- increased diagnosis of this condition in the last
ive radiologic evidence. An early and reliable decade can be attributed to the availability of
diagnosis is important as anticoagulation or diagnostic imaging techniques, such as color
antithrombotic treatment is recommended in Doppler ultrasound (US), computed tomography
order to reduce the risk of an early thromboem- (CT), magnetic resonance (MR) imaging, and,
bolic event. The optimal imaging method for the more recently, high-resolution MR imaging of
diagnosis of CAD is still debated. Doppler ultra- the arterial wall (HR-MRI). Although several
sound examination (DUS), the first-line screen- comprehensive reviews have focused on different
ing tool, may show normal findings whenever aspects of CAD such as risk factors or treatment,
the mural hematoma results in subtle lumen the present review proposes a global overview of
alterations or is located in a segment that cannot the epidemiology, pathophysiology, and
readily be displayed by DUS. Despite the ability predisposing factors, clinical and imaging charac-
of CT angiography to reveal imaging findings of teristics, and therapeutic options. Additionally,
CAD, brain CT is known to be poorly sensitive published data on the outcome of CAD, including
for the detection of ischemic lesions. Thus, cer- mortality rates and recurrences, were summa-
vical contrast-enhanced MR angiography rized. This review deliberately focuses on sponta-
(CE-MRA) coupled with T1-WI fat-suppressed neous (nontraumatic) CADs in adults because
axial sequences has gradually replaced conven- these are managed by stroke physicians, whereas
tional angiography for the diagnosis of cervical traumatic ICAD and cases in children are more
artery dissection. However, the diagnosis of commonly handled by surgeons and pediatricians
CAD often remains challenging, and it is not and can require different management strategies.
rare for the diagnosis of CAD to remain pre- Moreover, purely intracranial dissections or aortic
sumptive despite an extensive imaging work- dissections extending to the cervical arteries are
up. The diagnosis can be retrospectively con- outside the field of this review because their
firmed on imaging follow-up by monitoring the mechanisms, symptoms, and management are
lumen healing or progression in response to different.
treatment. High-resolution MRI (HR-MRI) of
the arterial wall is a noninvasive imaging tech-
nique that provides a delineation of the lumen Epidemiology
and arterial wall. This technique, available at 1.5
and 3-T, which requires standard or dedicated The incidence of CAD in the general population is
surface coils, has been extensively used for low, around 2,6 (95 % CI 1.9–3.3) per 100,000
carotid atherosclerosis and, more recently, for inhabitants per year [2]. Vertebral artery dissec-
carotid dissection. tions (VAD) are less common than CAD, although
8 Carotid Artery Dissection 117

this difference has been reduced by the increased The risk of ICAD is increased after major thoracic
availability of MRI, which enables an accurate injuries, whereas the risk of vertebral artery dis-
diagnosis of vertebral artery dissection [3]. sections, over the scope of this review, is
In the North American study, CAD occurred in increased in case of cervical spine fractures [6].
the community at a mean age of 45, 8 years, which
is similar to reports from two large published Minor Traumatism or Mechanical Trigger
multicenter series of ICAD in European Event
populations (44,0 years [459 patients] and 45,3 A dissection is termed spontaneous whenever
years [696 patients]) [2, 4]. there is no history of major trauma. Major pene-
A slight predominance in men was reported in trating or nonpenetrating trauma to the artery is
the European multicenter hospital-based series often lacking, but precipitating events involving
(53–57 %), whereas 0–52 % of the patients were hyperextension, rotation, or lateroversion of the
women in the North American population-based neck are often reported by patients with ICAD
study [2, 4]. CADs seem to occur more frequently [7]. These include whiplash injury, chiropractic
in men and at an older age (47,0 vs. 43,4 years) manipulation, various sporting activities, neck
than do vertebral artery dissections [2, 4]. These movements, and severe coughing. The exact role
age and gender differences are difficult to explain. of these precipitating events in the pathophysiol-
However, it is possible that risk factors or trigger- ogy of ICAD is unclear but suggests a mechanical
ing events, such as traumas in young patients, are contribution. The styloid process and hyoid bone,
more involved in the pathogenesis of VAD both in the vicinity to the ICA, have been impli-
than ICAD. cated in the pathogenesis of CAD with shorter
distances between the styloid process and hyoid
bone and dissected ICA than in nondissection
Pathogenesis carotid pathology [8, 9].
However, such traumas are recurrent and result
The pathophysiology of CAD remains controver- in ICAD in only a few individuals. If anatomic
sial. Patients with ICAD may have a constitu- factors contribute so strongly to CAD risk, one
tional, at least in part, genetically determined could wonder why ICAD recurrence is rare. If
weakness of the vessel wall. Environmental fac- prior cervical trauma seems to be an important
tors such as minor trauma or acute infection could environmental determinant of CAD, it is not an
act as triggers [5]. ICADs are usually classified as independent outcome predictor [7]. Because of
either traumatic in case of major trauma preceding the characteristics of most prior cervical trauma,
the onset of clinical symptoms or spontaneous, the term mechanical trigger event rather than
otherwise. trauma may be more appropriate [7].

Traumatic Carotid Artery Dissection Spontaneous Carotid Artery Dissection

Major Traumatism In the absence of a traumatic trigger, the etiology

Traumatic ICADs can occur as a result of major often remains undefined, likely to be multifactorial,
penetrating or nonpenetrating traumas. The most including infections, environmental exposures, and
common cause of a traumatic dissection is blunt collagen abnormalities of the arterial wall.
(severe hyperextension) or penetrating (stab
wound) injury to the neck. CAD has been shown Infection
to occur in about 1–2 % of patients who have had An association of ICAD with history of recent
blunt trauma, and the risk is also increased with infection has been reported [10, 11], potentially
trauma-associated injuries, such as facial or skull- predisposing individuals to CAD via endothelial
base fractures, and traumatic brain injury [6]. damage or prothrombotic mechanisms [10–14].
118 O. Naggara et al.

A possible causal role of recent infection is Transient Arteriopathy

supported by a seasonal peak of CAD in autumn. The epidemiology of CAD with an incidence peak
This pattern has been attributed to an increased in autumn, the clinical observation that some
occurrence of infection and weather-related patients with CAD develop multiple consecutive
changes in blood pressure, coagulation parame- dissections within weeks [18, 19], and the associ-
ters, and physical activity [13]. The failure to ation with hereditary connective tissue diseases,
prove a specific infective agent led to the hypoth- as well as ultrastructural abnormalities of connec-
esis that the inflammatory response rather than a tive tissue in skin biopsy specimens, support the
specific infective agent might be responsible for a hypothesis that a general susceptibility state is
transient arteriopathy nonspecific inflammation- involved in the pathogenesis of CAD.
mediated injury of the arterial wall linked to The presence of an underlying vasculopathy is
CAD [15]. Reports of elevated C-reactive protein also suggested by the fact that CAD patients com-
(CRP) and peripheral leukocyte counts in CAD monly present with concomitant arterial abnor-
support this hypothesis [16]. malities, such as fibromuscular dysplasia, aortic-
root dilation, hyper-distensibility [20], and
Genetic Risk Factors increased stiffness of the arterial wall, as well as
Genetic factors might also have a role in the impaired spontaneous and endothelial-dependent
pathophysiology of CAD, mainly as part of a vasodilation [21]. About 50 % of patients with
multifactorial predisposition. The prevailing the- CAD have connective tissue aberrations in their
ory is that CAD is a multifactorial disease caused skin, including composite collagen fibrils and
by several genetic variants and environmental fragmentation of elastic fibers [22]. In addition,
factors, each probably having a modest and signs of tissue weakening along the tunica media
potentially synergistic effect. The high preva- and the tunica adventitia junction in superficial
lence of multiple dissection events and of long- temporal artery biopsy specimens of patients
term (>1 year) recurrent dissections in patients with CAD but not of controls suggest the presence
with a familial history of CAD indicates that a of a generalized arteriopathy leading to impair-
specific predisposition for familial cervical dis- ment of the stability of the arterial wall in patients
sections may exist [15]. Less often (<2 % of with CAD [23]. Association of CAD with visceral
cases), ICAD has been associated with a known artery dissection, or reversible cerebral vasocon-
monogenic connective tissue disease, such as striction syndrome [24], strengthens this inflamma-
vascular Ehlers-Danlos syndrome type IV [15], tory hypothesis [5, 15]. This latter hypothesis is
or, more rarely, in patients with Marfan’s further reinforced by high-resolution MRI findings
syndrome or other known monogenic connective (see “High-Resolution MRI” section) [25, 26].
tissue disorders, such as osteogenesis imperfect
[15]. About half of patients with cervical dissec-
tions have abnormalities in skin connective tissue Insights on the Mural Hematoma
that follow an autosomal-dominant inheritance
pattern [15]. Mural Hematoma Constitution
Although these findings suggest that genetic
factors play a role in CAD [15], all published As in aortic or coronary dissections, CAD is the
genetic association studies on ICAD have been mere consequence of a wall hematoma, either
underpowered. More accurate data on genetic secondary to an intimal tear or to direct bleeding
risk factors for CAD can only be obtained from within the arterial wall caused by ruptured vasa
much larger multicenter genetic association stud- vasorum. Current insight into the pathogenesis of
ies, such as the ongoing genome-wide associa- coronary arteries and thoracic aorta dissection
tion study within the Cervical Artery Dissections suggests that dissecting aneurysms are disorders
and Ischemic Stroke Patients (CADISP) of the outer arterial layers, involving the vasa
consortium [17]. vasorum. Histological evidences of pathological
8 Carotid Artery Dissection 119

alterations of the tunica adventitia and the tunica temporal bone [29]. Since intracranial arteries
media suggest that CAD is due to an “outside-in” have no external elastic limitans and have a thin-
process starting in the outer arterial layers rather ner media and adventitia compared with cervical
than an “inside-out” injury starting from the endo- arteries, intracranial extension of a CAD has a
thelium leading to endothelial rupture as the pri- high risk of intracranial rupture.
mary event [27].
Regardless of its mechanism, the mural hema-
toma can expand toward the intima or the adven- Mural Hematoma Consequences
titia, resulting in a stenosis or pseudoaneurysm
formation. Mural blood then dissects longitudi- Dissections can be divided into two groups. The
nally. Dissections can tear through the intima, first group is mainly subintimal, and the dissec-
with the consequence for mural blood to enter tions compromise the arterial lumen or cause
the lumen of the artery. Increased thickness of thrombus to enter the lumen. The second group
the arterial wall by mural blood accumulation is subadventitial, and these dissections either per-
causes compression of the lumen, causes pertur- turb structures adjacent to the outer arterial wall.
bation of the blood flow stream and of the vascular Extracranial subintimal CAD causes symptoms
endothelium, and, finally, causes activation of primarily through luminal compromise and pres-
platelets and the coagulation cascade. These ence of luminal clot. Ischemic symptoms and
changes contribute to the formation of a luminal infarction are caused by reduced perfusion in the
thrombus. CAD may start from the luminal side at brain territory supplied by the artery or embolism.
the intimal surface and, through an intimal flap, Subadventitial cervical CAD can perturb the sym-
dissect into the media. The mural hematoma can pathetic fibers located along the dilated artery and
create a false lumen that might reconnect with the might compress or cause ischemia to the lower
true lumen, creating parallel circulatory channels, cranial nerves (IX–XII) that exit near the
separated by the intimal flap [28]. This false skull base.
lumen may either remain patent, resolve
completely, or thrombose and cause narrowing
of the true lumen. The mural hematoma some- Clinical Features
times extends between the media and the adven-
titia, leading to the formation of an aneurysmal The consequences of the mural hematoma are
lesion. local symptoms and signs including headache
and neck pain (supposedly attributable to a dis-
tension of the artery by the mural hematoma stim-
Mural Hematoma Location ulating pain-sensitive receptors), Horner’s
syndrome, or cranial-nerve palsies resulting from
The commonest location for a spontaneous CAD the stretching of sympathetic-nerve and cranial-
is the cervical segment 2–3 cm distal to the carotid nerve fibers by an enlarged carotid artery. The
bulb, an area that is probably subject to most of the mechanism by which CAD leads to cerebral or
stretch during extension or rotation of the neck. retinal ischemia is thought to be embolic, from
Although several studies have described the prox- intraluminal thrombi forming at the site of the
imal anatomical location of spontaneous CAD, intimal tear. Hemodynamic infarcts are also pos-
data on the cranial extension of the mural hema- sible. Occasionally, CAD can lead to subarach-
toma are scarce. One study reported that the mural noid hemorrhage when the dissection extends
hematoma extended in the cranial direction, intracranially.
involving the petrous segment of the ICA in The classical triad of presentation of a CAD is
nearly 75 % of CAD [29]. If CADs extend for a ipsilateral headache, facial or neck pain, and
variable length cranially, they do not usually past Horner’s syndrome, followed hours or days later
the point of entry of the ICA into the petrous by cerebral or retinal ischemia. This triad of
120 O. Naggara et al.

symptoms is found in less than one third of attributable to other factors. Although sometimes
patients, but the presence of two elements of this inaugural, it is most often preceded by local symp-
triad strongly suggests the diagnosis [1, 5]. toms that started shortly before the ischemic
event, a finding that is highly suggestive of CAD.
Cerebral infarction and TIA are inaugural in
Local Symptoms and Signs respectively 10–15 % and 10–20 % of cases
[30]. Middle cerebral artery territory is involved
The following symptoms and signs can occur in more than 80 % of cases [30]. In rare instances,
separately or in combination: [2, 19, 30] Horner’s carotid artery dissection can (MCA) lead to retinal
syndrome, unusual neck pain or headache, ischemia [19, 32].
cranial-nerve palsy, tinnitus, and, rarely, The literature provides support for several con-
cervical-root injury. Horner’s syndrome and cepts, regarding the primary cervical lesion and
cranial-nerve palsy occur in carotid artery dissec- the secondary intracranial lesion in CAD. An
tions. Sudden-onset Horner’s syndrome, particu- autopsy case has provided the only direct demon-
larly if associated with headache or neck pain or stration of secondary intracranial emboli
with an ipsilateral ischemic stroke in the carotid [33]. Three studies described indirect signs of
territory, can be considered to be specific to CAD emboli on transcranial Doppler monitoring stud-
and should lead to urgent imaging of the cervical ies [34–36], but the sample size of these studies
arteries. limits the significance of these results. According
The characteristics of pain associated with to current concepts relating to mechanism and
CAD are not specific and can sometimes mimic stroke pattern, strokes in pial or perforating artery
migraine or even cluster headache. CAD with territory are more likely to be embolic, whereas
isolated pain might be more common than border-zone infarcts are more likely to be hemo-
expected [31] and is more often caused by extra- dynamic. Using this simplification scheme, sev-
cranial vertebral artery dissection but can also be eral authors discussed the mechanism of
caused by carotid artery dissections. The onset infarction in CAD patients, with conflicting
type ranged from thunderclap headache to results [37, 38]. Based on the stroke pattern on
progressive pain. CT, one group demonstrated an embolic mecha-
Cranial-nerve palsies are rare, representing less nism in 92.2 % of 65 patients [39], whereas others
than 7 % of CAD cases in large hospital-based reported a similar frequency of embolic and
series. The hypoglossal nerve is the most com- hemodynamic infarcts in 11 patients
monly affected, followed by the CN IX and CN X [38]. Recently, the occlusive intracranial throm-
cranial nerves, which are topographically close to bus in vivo was directly demonstrated using T2*
the carotid artery in its cervical trajectory. The sequence. Other findings that substantiate the con-
most likely mechanism is compression of the cept of artery-to-artery embolism in the pathogen-
nerves by an enlarged carotid artery. Cranial- esis of stroke in CAD were the multiple acute
nerve ischemia is another potential mechanism, DWI lesions in the majority of patients and the
particularly in very rare cases of upper cranial- occurrence of pial artery or perforating artery ter-
nerve palsy. ritory stroke in 96 % of patients [40–43].
If the simplification of the relationship between
the mechanism and stroke pattern provides an
Ischemic Symptoms easy approach to the presumed mechanisms of
cerebral infarction in most patients, one should
Stroke (transient ischemic attack [TIA] or cerebral bear in mind that, even at the individual level,
infarction) is the most frequent and severe mani- both mechanisms may be at play and may rein-
festation of extracranial CAD. The clinical pre- force each other in causing the ischemic lesion.
sentation of cerebral ischemia caused by CAD Border-zone infarcts might result from mixed
does not differ from that of cerebral ischemia mechanisms by impaired clearance of emboli in
8 Carotid Artery Dissection 121

hypoperfused regions. Indeed, during the consti- accounted for in published series. The rate of
tution of the mural hematoma, only the embolic ischemic recurrences ranges from 0 % to 13,3 %
mechanism may explain the occurrence of an at 1 year, with the largest recurrence rates
infarction, whereas the hemodynamic mechanism observed when recurrent events that occurred
may appear progressively with the narrowing of before the diagnosis of CAD are taken into
the lumen due to the growth of the mural hema- account [19, 44]. Recurrent ischemic events usu-
toma. Once the lumen is severely stenosed, hemo- ally occur during the first weeks after the dissec-
dynamic impairment may encourage the tion. Factors associated with an increased risk of
formation of secondary clots in border-zone recurrent ischemic events are multiple dissections
regions, impede the clearance of distal clots, and and a history of hypertension [19]. Although iso-
increase the impact on cerebral tissues. The fact lated cases of ischemic strokes caused by chronic
that border-zone infarction rarely occurred in iso- dissecting aneurysms of the carotid artery have
lation but mainly occurred in association with pial been reported, two prospective series of aneurys-
or perforating artery territory infarction supports mal CAD found no ischemic events after about
the hypothesis of an impaired clearance of small 3 years of follow-up [44]. Similar rates of ische-
emboli [40]. mic recurrences have been reported in 46 patients
with CAD with a transient stenosis or occlusion
and in 46 patients with CAD with a permanent
Outcome stenosis or occlusion [45]. However, in another
series of 130 consecutive patients with CAD,
Clinical Outcome ischemic recurrences were attributed to a worsen-
Mortality rates in the acute phase of CAD are ing carotid stenosis in five of six cases [46].
generally low (<5 %) [2, 19], although higher
rates of up to 23 % have been reported in previous Arterial Lesion Evolution
series of subsets of patients with severe CAD. In The proportion of patients with complete resolu-
general, mortality after CAD might be tion of arterial abnormalities varies between stud-
underestimated, as patients with severe forms of ies: 46 % for stenoses, 33 % for occlusions, and
CAD sometimes die before arterial imaging and 12 % for dissecting aneurysms in the general
etiological diagnosis can be done. In support of population [46–48]. Complete resolution or stable
this possibility, data from studies on patients with residual luminal irregularity was documented
acute complete MCA infarction have suggested after a median duration of 0,29 years and in
that CAD could be a major cause of “malignant” 82 % of cases within the first year [49]. The like-
cerebral infarction. Functional outcome is good in lihood of complete recanalization seems higher in
about three-quarters of patients with CAD, but patients with CAD who present with only local
impact on quality of life and socioprofessional symptoms and signs [50]. Occlusions can lead to
integration can be important [1, 32]. The func- residual stenoses, and residual aneurysms can
tional outcome after CAD does not seem to be appear after the acute phase in initially stenotic
better than in other types of ischemic strokes in or occluded arteries [44].
young individuals. Factors associated with a poor
functional outcome are presence of cerebral ische- Recurrences of Dissections
mia, arterial occlusion, carotid location, older age, Recurrences of dissections are defined by a new
and a severe National Institutes of Health stroke mural hematoma, remote the initial lesion. Early
scale score at onset. recurrences of dissection in the days or weeks
after the initial event could be a manifestation of
Recurrent Ischemic Events a unique transient disorder, whereas late recur-
Recurrent ischemic events seem to be rare, rences occurring several months or years later
although early recurrences (before the patient is could indicate an underlying connective tissue
discharged from the stroke unit) are not always weakness. Both early and late recurrences are
122 O. Naggara et al.

rare; they seem to be most frequent within the first since it does not allow direct visualization of the
2 months after the initial event [19], but some vessel wall, whenever the false lumen does not
dissections that were diagnosed as multiple opacify with contrast medium or when the lumen
might have occurred sequentially within a short contains thrombus. It has been replaced by non-
time frame. Late recurrences are possibly invasive imaging modalities, both for the primary
underestimated as studies with long-term follow- diagnosis and the follow-up of dissections.
up are scarce [51]. Recurrent dissections were not Pathognomonic signs are double lumen and inti-
reported in the only population-based series with mal flap. These are detected in less than 10 % of
follow-up data of 48 patients with CAD (mean cases [52], while the following patterns are found
duration of follow-up: 7–8 years). In hospital- more commonly. The string sign, the angiographic
based series, the rate of recurrent dissections hallmark of CAD, is a long, tapered, usually eccen-
ranged between 0,6 % and 25,0 %. The recurrence tric, and irregular stenosis that begins beyond the
rate was probably overestimated in tertiary refer- ICA bulb and might extend to the skull base
ral series. Risk factors for recurrent CAD are [52, 53]. The “string and pearl” sign corresponds
younger age [18], a familial history of CAD, to a focal narrowing with a distal dilatation. The
vascular Ehlers-Danlos syndrome, and flame-shaped tapering of the lumen, a tapered
fibromuscular dysplasia [32]. The prognosis of occlusion that spares the carotid bulb, is suggestive
recurrent CAD has been described as benign [51]. of dissection. Although these patterns are indicative
of dissection, they are not specific. DSA findings of
spontaneous CAD may also include a dissecting
Imaging aneurysm, alone or associated with a tapered steno-
sis [52, 53], with changing lumen patterns over
Dissection of the craniocervical arteries, which time, often returning to a normal caliber.
was once considered as an uncommon diagnosis,
has become increasingly recognized as a cause of
stroke in young and middle-aged patients. In part, Ultrasound
this increased recognition is due to increased use
of noninvasive imaging studies. The latter has Noninvasive imaging is required for the detection
allowed the screening of larger numbers of of possible CAD. Ultrasound imaging provides
patients. In some populations, the rate of dissec- direct visualization of pathological findings that
tion of the craniocervical arteries has been diag- may be related to CAD and demonstrates
nosed at rates 3–10 times greater than that impaired lumen patency [54]. Pathognomonic
determined before the use of MRI, MR angiogra- findings on gray-scale US include mural hema-
phy (MRA), and, more recently, CT angiography toma, an intimal flap, or coexistence of a true and
(CTA) studies. a false lumen (Fig. 1) [55]. Unlike DSA, US is
able to demonstrate a false lumen even if this
lumen is thrombosed.
Digital Subtraction Angiography Spectral Doppler US waveforms in CAD can
be normal or may demonstrate nonspecific fea-
In the past, the diagnosis of arterial dissection tures also occurring in high-grade stenosis and
relied on standard catheter angiography (DSA). occlusions (damped spectral waveform with
As some authors have pointed out, DSA is an lower amplitude and biphasic pattern, high-
imperfect reference standard because of question- resistance spectral waveform, or absence of flow
able interobserver variability. Indeed, a number of with no spectral Doppler waveform) [54].
entities can mimic arterial dissection and DSA is The diagnostic sensitivity of US decreased if
an invasive procedure. In addition, if DSA may CAD results in a low-grade stenosis [54].
detect vessel wall irregularities, such as Sturzenegger et al. showed high accuracy in
fibromuscular dysplasia, it may miss dissections, detecting a dissection with US in case of ICA
8 Carotid Artery Dissection 123

Fig. 1 Longitudinal color

(a, b) and power (c)
Doppler ultrasound image
demonstrating a
hypoechoic thickened wall,
a finding consistent with a
mural hematoma (a)
narrowing and displacing
the true lumen (b) of the
ICA with enlargement of its
external diameter to 7 mm
(normal 5 mm). An
abnormal high-resistance
spectral Doppler US
waveform is demonstrated
in the dissected lumen (d)

occlusion or high-grade stenosis but a 20 % sen- Magnetic Resonance Imaging

sitivity in low-grade stenosis [49]. Another limi- Dissections of the craniocervical arteries are best
tation is that US is less reliable for CAD outside assessed with cross-sectional fat-suppressed T1
the “window” of the US examination such as sequence with cranial and caudal saturation pulses
dissections located in the subpetrous segment and with MRA techniques. The former can pro-
and the carotid canal or in case of small mural vide direct visualization of the mural hematoma,
hematoma [54, 55]. which is the pathological hallmark of a dissection,
and the latter allows noninvasive visualization of
Cross-Sectional Imaging: CT or MRI?
Lumen Imaging: MR Angiography
MRI, including MR angiography (MRA) and high- Contrast-enhanced MRA (CE-MRA) is the tech-
resolution MRI (HR-MRI), and CT angiography nique of choice, thanks to recent improvements in
(CTA) are the main noninvasive cross-sectional MR gradients and software, which allow rapid
imaging tools used for assessing CAD. The use of acquisitions with centric k-space filling and auto-
MRI and CTA is twofold: (1) demonstration of a matic bolus triggering. Like DSA, CE-MRA can
mural hematoma and (2) evaluation of the conse- demonstrate luminal irregularity; long narrowing
quence of the mural hematoma on the lumen of the vessel, the so-called string sign; or occlu-
patency. MRI is currently the recommended imag- sion and pseudoaneurysm formation (Fig. 2).
ing technique [5], thanks to its higher ability to CE-MRA also has the advantage of being easily
demonstrate arterial wall abnormalities and its bet- preceded with fat-suppressed T1-WI axial images
ter specificity as compared with CTA. CTA can be through narrowed segments to confirm or refute a
performed as a first-line screening tool, particularly suspected dissection. The combination of cross-
in case of trauma. sectional MRI and MRA is not only a powerful
124 O. Naggara et al.

Fig. 2 Luminal patterns of carotid artery dissection on (arrow) that begins beyond the ICA bulb, with a distal
contrast-enhanced MR angiography (CE-MRA). (a, b) dilatation, the “string and pearl” sign. (c) Occlusive pattern
Stenotic type of CAD. In (a) CE-MRA demonstrates a of CAD. Flame-shaped tapering followed by an occlusion
long, tapered, eccentric, and irregular stenosis (arrows) of the lumen that spares the carotid bulb. (d)
that begins beyond the ICA bulb and that extends to the Pseudoaneurysmal pattern of CAD (arrow) associated
skull base, the “string” sign. In (b) focal eccentric stenosis with a distal long-tapered stenosis (double arrows)

diagnostic modality but can also be used effec- hematoma after 6 months, as it becomes isointense
tively for noninvasive follow-up of CAD to mon- with the surrounding soft tissue. These patterns are,
itor lumen patency evolution, resolution of a however, not specific for a dissection as fresh
mural hematoma, or development of complica- thrombus due to an atheromatous or embolic occlu-
tions (Figs. 3, 4, and 5). sion that may exhibit a similar evolution of MR
signal intensity. It may, indeed, be impossible to
Mural Hematoma Imaging distinguish an intraluminal clot from an occlusive
Carotid artery dissection is characterized, on mural hematoma. In such cases, a crescentic thick-
fat-suppressed T1 sequence, by a narrowed eccen- ening with eccentric signal void proximal to com-
tric flow void, which is surrounded by a crescent- plete occlusion should suggest dissection rather
shaped, hyperintense area expanding the vessel than atheromatous occlusion (Fig. 9).
diameter (Fig. 6) [53]. MR signal of the mural At least four causes of false-positive MRI or
hematoma has a similar temporal evolution than MRA findings have been reported [57].
intracerebral counterpart. [56] An acute mural Atheromatous plaques causing subtotal occlu-
hematoma can be hypointense on T2- and T1-WI sion can be hyperintense on T2-WI images and
images and therefore be difficult to distinguish mimic a CAD [57]. This rim is, however, typically
from an area of flow void. The mural hematoma isointense on T1-WI images, whereas the
may therefore be missed on MRI within the first periarterial rim in subacute CAD is usually
48 h after CAD [53]. Thereafter, the hematoma hyperintense on T1-WI images [56].
becomes of intermediate signal intensity on Structures adjacent to arteries can simulate a
T1-WI images and hyperintense on proton density periarterial rim of hyperintense signal which can
and T2-WI images (Fig. 7). After a few days, the mimic a mural hematoma or a pseudoaneurysm.
hematoma remains hyperintense on T1-WI images For instance, carotid arteries are surrounded
for approximately 2 months (Figs. 7 and 8). Sub- by fat. Systematic fat-saturation imaging is the
sequent loss of signal intensity can mask the best solution.
8 Carotid Artery Dissection 125

Fig. 3 Evolution of a pseudoaneurysm type of carotid the stenosis associated with a progression of the
artery dissection. The pseudoaneurysm (arrow) was initially pseudoaneurysm 2 months after onset of the CAD (b).
associated with a focal-tapered stenosis (a). Resolution of Progressive normalization seen at 6 and 12 months (c, d)

Fig. 4 Evolution of a stenotic type of carotid artery dissection. The long, irregular stenosis (arrow) improves with
progressive normalization seen at 12 months

Turbulent or slow flow on MRI and MRA can can produce hyperintense slow flow distally,
be another confounding factor for CAD by pro- which can be manifested by loss of part or all of
ducing hyperintense signal in the periphery of the the intracranial flow void. This signal can be mis-
arterial flow void, along the surface of the arterial taken for mural hematoma. Inflow phenomena
lumen. For example, marked narrowing of the generally produce signal abnormalities that are
extracranial segment of the internal carotid artery located centrally within the flow void rather than
126 O. Naggara et al.

Fig. 5 Twelve-month evolution of a pseudoaneurysm type of carotid artery dissection with true and false lumen
remaining patent

Fig. 6 Right carotid artery

dissection on contrast-
enhanced MR angiography
(CE-MRA) (a) and axial
fat-suppressed T1-WI
image (b). Carotid artery
dissection is characterized,
on fat-suppressed T1-WI by
a narrowed eccentric flow
void, which is surrounded
by a crescent-shaped,
hyperintense mural
hematoma expanding the
outer vessel diameter

peripherally. Homogeneity of the hyperintense mistaken for dissection. The recognition that the
signal on all slices associated with vessel expan- image is obtained at the extreme of the imaging
sion supports dissection rather than slow flow. volume (i.e., first or second image in a stack of
Flow-related enhancement in arteries and veins images) and correlation with the signal of the
can simulate mural hematoma. Bright signal in arteries on other imaging sequences resolve the
arteries at the extreme of an imaging volume due issue.
to entry of unsaturated spins (i.e., the so-called A false-negative diagnosis of dissection on
entry slice phenomenon) has long been recog- MR is most likely at the very early stage, when
nized as a cause of abnormal arterial signal on the mural hematoma is not yet hyperintense.
selected images. This phenomenon can simulate Indeed, at this stage, the blood products within
mural abnormal signal or complete absence of the the hematoma have not yet progressed to the stage
expected flow void within an artery and be of methemoglobin. In this case, the signal
8 Carotid Artery Dissection 127

Fig. 7 Temporal evolution

of MR signal of the mural
hematoma on T1 and T2
sequences. An acute mural
hematoma, e.g., within
2 days after onset, is
isointense on T2- and
T1-WI images. The
hematoma becomes of
intermediate signal
intensity on T1-WI images
and hyperintense on T2-WI
images after day 2. The
hematoma remains
hyperintense on both T2-
and T1-WI images for
approximately 2 months.
Subsequent loss of signal
intensity can mask the
hematoma after 6 months,
as it becomes isointense
with the surrounding soft
tissue (not shown)

Fig. 8 Temporal evolution of MR signal of the mural isointense to the surrounding tissue with bright heteroge-
hematoma on T1 sequence. Contrast-enhanced MR angi- neous signal of its peripheral part (b). It becomes homo-
ography (a) demonstrated a long irregular stenosis sparing geneously hyperintense at day 3 (c) and was no longer
the right ICA bulb. The mural hematoma was initially visible on fat-suppressed T1-WI image at day 60 (d)
128 O. Naggara et al.

Fig. 9 Occlusive cases of

carotid artery dissection. In
the first patient (a, b),
CE-MRA demonstrated
right ICA occlusion and
long stenosis of the left
ICA. On fat-suppressed
T1-WI image (b), it was
difficult in the right ICA
lumen to distinguish an
intraluminal clot from an
occlusive mural hematoma
(arrow). CAD diagnosis
was done, thanks to left ICA
mural hematoma (double
arrows). In the second
patient (c, d), CE-MRA
demonstrated left ICA
occlusion (arrow). On
fat-suppressed T1-WI
image (d), it was also
difficult distinguishing an
intraluminal clot from an
occlusive mural hematoma.
In such a case, the
crescentic hyperintense
thickening associated with
increase outer external
diameter of the occluded
ICA suggested dissection
rather than atheromatous

intensity of the hematoma does not differ from The mural hematoma frequently extends cra-
that of background fat-suppressed tissue. Other nially to the petrous carotid segment in the case of
signs, such as the presence of a pseudoaneurysm, carotid artery dissection. This location is within
or an increase in external diameter may help in the limits of the field of view of standard brain
these difficult cases. MRI. Consequently, apart from dedicated cervical
In addition, whenever a CAD produces arterial fat-suppressed T1 sequences, brain MRI may be
occlusion (as opposed to stenosis), the mural useful to demonstrate CAD. More than three-
hematoma may not be conspicuous against the quarters of such acute CAD can be diagnosed
background of the hyperintense signal abnormal- using stroke brain MRI protocol only, including
ity of an occluded arterial lumen on DWI, FLAIR, T2* sequences, and time-of-flight
fat-suppressed T1-WI images. In such cases, MRA (Fig. 10) [29]. This result might have clin-
although the arterial occlusion is recognized, the ical implications for patients who are not initially
cause may not be discernible. suspected of having CAD based on available
8 Carotid Artery Dissection 129

Fig. 10 Illustration of carotid artery dissection on brain circle of Willis (d). Sagittal T1-WI imaging (e). Increased
MRI sequences in five patients. Axial gradient-recalled external diameter, crescentic mural thickening (arrows),
echo T2 (a), diffusion-weighted imaging (b), FLAIR (c), and eccentric lumen (arrows) were clearly seen in all
and native slices of 3D time-of-flight angiography of the patients

history at the time of protocol and at institutions visualization of the cervical arterial mural hema-
where cervical CE-MRA is not coupled with MR toma on both carotid and vertebral arteries [59].
brain imaging. The interpretation of direct signs of
CAD on brain MRI is a learnable skill and High-Resolution MRI Using Surface Coils
requires little specialized training and can also be at 1.5-T
achieved in regular clinical practice. The aim of MRI of the arterial wall of cervical
arteries is to obtain in plane, submillimetric
High-Resolution MRI voxels. The loss of signal to noise induced by
High-resolution magnetic resonance imaging reducing the voxel size (500  500 μm  2–3
(HR-MRI) is an attractive tool for the assessment mm) is compensated for by the use of phased-
of carotid artery diseases. Using dedicated surface array surface coils, which increases the signal-to-
radiofrequency coils at 1.5-T or standard coils at noise ratio by 40 %. The so-called HR-MRI cor-
3-T, it provides high-resolution images of the responds to images with submillimetric voxels
arterial wall in vivo. Optimal image quality relies that imply the use of dedicated surface coils.
on correct patient’s preparation, on adequate coil These coils are commercially available and com-
positioning, and on the use of pulse sequences patible with most MR manufacturers. They con-
with signal suppression of the flowing blood. sist of a pair of multichannel coils. The coils,
HR-MRI has initially been applied for the made of flexible material, are placed bilaterally
in vivo analysis of carotid atherosclerosis with around the neck. The carotids, being superficial
now solid evidence that HR-MRI can identify structures, are well suited for surface coil imaging.
the major components of atherosclerotic plaque, To explore the patient with suspected CAD,
i.e., lipid core, hemorrhage, calcifications, as well the radiologist should ensure that the coil is at the
as the fibrous cap [58]. Beyond atherosclerosis, level of the carotid-presumed lesion. A multicentric
HR-MRI can be used in case of suspicion of study reports a rate of 90 % of interpretable MRI
cervical artery dissection, providing an excellent exams performed for the purpose of carotid
130 O. Naggara et al.

bifurcation analysis. For a given magnetic field examination. The surface coils should be posi-
when using the same surface coil, image quality tioned anteriorly and laterally, facing the mandib-
seems stable across different platforms [60]. ular bone for the petrous portion of internal
The MR protocol comprises multiple carotid artery, whereas it should be placed poste-
sequences in the axial plane, covering about rior and under the mastoid for the V3 segment;
3 cm in height. Consequently, the slices must be (2) acquire an oblique scout view (3D phase con-
focused on the index lesion or artery. trast, 2D time of flight, 1 min) of the entire index
As in atherosclerotic plaque exploration, it is artery so as to visualize the plaque distribution and
currently recommended to combine multiple con- the level of the carotid lesion; (3) acquire HR
trasts to fully characterize potential mural hema- images (voxel <500  500 μm in plane); (4) use
toma. Pulse sequences designed for vascular multicontrasts, by combining black and bright
imaging are classed as “bright blood” or “black blood sequences, to identify the mural hematoma
blood,” depending upon the flowing blood signal. presence and age; (5) keep the total acquisition
Both of these sequences provide a good contrast time to less than 30 min. Indicative parameters at
between the lumen and the arterial wall. Bright 1.5-T are detailed in Table 1 [58].
blood sequence corresponds to 3D TOF gradient
echo sequences used for MR angiography. On this High-Resolution MRI at 3-T
sequence, the flowing blood signal is enhanced, Although most of the HR-MRI studies have been
and the lumen is brighter than the vessel’s wall. performed on 1.5-T MR units, 3-T units are now
The lack of 180 refocusing pulse creates T2*- more widely available. Surface coils for carotid
sensitive signal. The aim of black blood imaging are commercially available for 3-T MR
(BB) sequences is to cancel the signal of the units. These are slightly different from those used
flowing blood so as to increase the contrast on 1.5-T units (tuned to a different frequency and
between the lumen (in black) and the wall. For adapted to the 3-T connectors and/or interfacing).
that reason, the acquisition rate is synchronized to They provide significant improvement in signal-
the heartbeat and is preceded by a double- to-noise and contrast-to-noise ratios [61],
inversion-recovery (IR) magnetization preparation allowing reduction in both the acquisition time
pulse. This process maximizes flow suppression and the in-plane resolution. For a given acquisi-
due to outflow and minimizes artifacts due to vessel tion time, thinner slices or larger anatomic cover-
motion. The time of inversion (TI) for the double- age in the z-axis can be obtained. At 3-T, longer
IR preparatory pulse is fixed as close to the null repetition times are advised for PD- and T2-WI
point of the blood signal. The double-IR prepara- images. Double-IR blood suppression techniques
tion and rapid SE acquisition are typically repeated are available at 3-T, but they require an adjustment
using 2RR cardiac gating until all the rapid SE of the TI for the blood T1 at 3-T, i.e., slightly
shots are acquired. Black blood sequence can be T2 longer TI for 3-T compared to that used for
or proton density (PD) weighted (2RR) or T1-WI 1.5-T [61]. Demonstration of strong agreement
(1RR). By applying the double-IR preparation between 3 and 1.5-T HR-MRI of the carotid
just after the cardiac trigger, the slice-selective plaques, particularly in the identification of ath-
re-inversion pulse is placed in the “end-diastolic” erosclerotic plaque composition, including hem-
part of the cardiac cycle. Because the TI is usually orrhage, supports the translation of histologically
close to 650 ms, the rapid SE acquisition is also validated 1.5-T criteria to 3-T carotid images
played out during diastole. A third inversion pulse [62]. However, there are differences in the mea-
can be added for fat suppression. sured size of arterial wall hemorrhage measures,
Several principles should be followed when larger at 1.5-T than at 3-T, suggesting that patients
designing an MR protocol for imaging cervical enrolled in serial prospective trials of hematoma
arterial wall in CAD: (1) the coil positioning is of measurement in CAD should be imaged at the
importance and should be guided by lumen abnor- same field strength in order to minimize interstudy
malities seen on CE-MRA or ultrasonographic errors during analysis [62].
8 Carotid Artery Dissection 131

Table 1 Indicative imaging parameters for 1.5-T high-resolution plaque MR imaging using dedicated surface coils [60]
Parameters 3D TOF T1 BB DP BB T2 BB
TR (ms) 30 1 RR 2 RR 2 RR
TE (ms) 6.9 9 16–20 50
FOV (mm) 130 130 130 130
Slice thickness (mm) 2.2 3 3 3
Matrix 288  224 352  256 256  512 256  512
NEX 2 3 2 2
Synchronization – Cardiac Cardiac Cardiac
Number of slices 20 8 8 8
Resolution (μm)a 254  254 254  254 254  254 254  254
Scan time (min)b 4 4–5b 4–6b 4–6b
PD indicates proton density, TR repetition time, TE echo time, TI inversion time, NEX no. of excitations, BB black blood
sequence with double-inversion-recovery preparation pulse
After zero filling interpolation
Reported scan times are based on a heart rate of 70 beats/min. Times shorten with faster rates

Fig. 11 1.5-T high-

resolution MRI using
surface coils. Axial
fat-suppressed T1-WI
images showing the mural
hematoma (a) and an
intimal flap (b) creating two
parallel circulatory
channels. Although most
often seen under CTA, the
intimal flap with the double
lumen sign can be seen on
HR-MR images

High-Resolution MRI Findings in CAD excellent depiction of the structural characteristics

Using standard MRI, an early and reliable identi- of the vessel wall and the vessel lumen in acute
fication of acute CAD might be impaired by the CAD. The eccentric vessel lumen (occlusion, ste-
limited spatial resolution, the tortuous course of nosis, luminal thrombus) and the vessel wall
the arteries, the great variability in normal vessel (crescent mural hematoma, pseudoaneurysm,
caliber, and the presence of the thick bony cover- double lumen, intimal tear) of the dissected artery
ing and adjacent veins. This is the case for dissec- can then be exquisitely analyzed using HR-MRI
tion of the upper portions of the vertebral arteries (Fig. 11) [63–65]. The mural hematoma is well
(V2 and V3 segments), outside the field of this defined between intimal and adventitial wall
review, and for dissection of petrous ICA. In some boundary. The signal intensity of mural hematoma
cases, the diagnosis of CAD remains presumptive, varies with the time elapsed since the onset of
further confirmed by sequential follow-up MR symptoms [56]. More specifically, the age of the
showing signs of healing or progression in mural hematoma may be classified into acute,
response to treatment. early subacute, late subacute, and chronic hema-
1.5-T HR-MRI has demonstrated its usefulness tomas according to its signal intensities on the T1-
for the assessment of CAD [59] and permits an and T2-WI sequences. One study suggested that
132 O. Naggara et al.

the MR signal changes of mural hematoma had correspond to a localized inflammation of the
only moderate agreement with the dynamics seen arterial wall. Spontaneous CAD is more fre-
in cerebral hematomas, even if it can be difficult to quently associated with the presence of PAE com-
determine the exact delay between the occurrence pared to traumatic CAD [25]. Another study
of CAD and the HR-MRI scan. Initially, the mural demonstrated that a subset of patients with spon-
hematoma is isointense to the sternocleidomastoid taneous CAD showed signs of a generalized tran-
muscle on T1-WI and hypointense on T2 and TOF sient inflammatory arteriopathy in contrast-
imaging. At a mean delay of 5,8 days, mural enhanced HR-MRI and [18F]-
hematoma may appear hyperintense on all fluorodeoxyglucose positron emission tomogra-
sequences (Figs. 7 and 8). Progressively, the sig- phy CT (PET-CT) [26]. This subset of patients
nal drops on all pulse sequences [56]. may be more prone to multiple dissections. Inter-
HR-MRI can help in distinguishing between estingly, such HR-MRI pattern is similar to those
dissection and atherosclerosis in case of cervical observed in inflammatory diseases, such as
artery occlusion [66]. The signal of the mural Takayasu’s disease and giant cell arteritis. PAE
hematoma is usually homogeneous in CAD and inflammation could simply result from com-
while being heterogeneous in hemorrhagic ath- pression by the mural hematoma. However, such
erosclerotic plaques. In addition, unequivocal dis- HR-MRI pattern was more frequently found in
tinction between mural hematoma and thrombus spontaneous CAD and rarely observed in trau-
is allowed by HR-MRI. matic CAD despite a larger hematoma in this latter
Recent studies have shown the presence of group in one study [25]. This minimizes the like-
perivascular soft tissue signal changes, related to lihood of periarterial abnormalities simply
periarterial edema (PAE) on 3-T HR-MRI in resulting from a mechanical compression. How-
patients with recent CAD. PAE can be detected ever, we cannot exclude the possibility that
on a 1.5-T MR unit (Fig. 12), with a high preva- inflammatory HR-MRI pattern could be associ-
lence (83.3 %) of PAE in the spontaneous CAD ated with thrombogenicity after CAD rather than
group including 18 patients. These changes may with the CAD mechanism itself.
Because of the limited accessibility of HR-MRI
using dedicated surface coils and the long scanning
time inherent to the technique, the idea is not to
perform HR-MRI in all patients suspected of CAD
but rather to improve our imaging protocols to get
supportive imaging evidence in most patients. This
is already feasible using 3-T MR magnets with
standard multichannel head and neck coils. The
3-T’s increase signal-to-noise ratio can be targeted
at improving spatial resolution without the need of
dedicated surface coil. In daily practice, 3-T
fat-suppressed T1- and T2-WI sequences with
3 mm slice thickness challenge the high resolution
obtained using dedicated surface coil with 1.5-T
MR unit to image the cervical artery wall.
Recent technological innovations, including
parallel imaging, optimized k-space trajectories,
Fig. 12 1.5-T high-resolution MRI using surface coils. and variable flip angles, allow the acquisition, at
Periarterial edema in spontaneous carotid artery dissection.
High-resolution fat-suppressed T2-WI MR image. Note
3-T, of fat-suppressed tridimensional T1-WI
the signal changes in the soft tissues (arrows) surrounding sequences (3D T1), with neck and head coverage,
the right dissected ICA submillimetric voxels (spatial resolution of
8 Carotid Artery Dissection 133

Fig. 13 3-T fat-suppressed 3D FSE T1 sequence, with of contrast-enhanced MR angiography and 3D FSE T1
luminal black blood effect. Axial (a), coronal (b), and images on coronal (d, f) and axial (e) views: the
sagittal views (c) showing a bright T1 hematoma and its hyperintense mural hematoma is facing luminal
extension. Fusion images combining a volume rendering irregularities

0.8  0.8  1 mm [3] before and 0.4  0.4  submillimetric quasi isotropic voxels, allowing
0.5 mm [3] after interpolation), and acceptable scan for multiplanar reformations; (5) a large field of
duration for clinical practice (total scan time in view in the z-axis, allowing whole neck coverage.
between 5 and 7 min). In a preliminary study, it This potentially allows 3D T1 to be used as a first-
was demonstrated that fat-suppressed 3D T1 line diagnostic tool, when the CAD location is
sequence can demonstrate the mural hematoma unknown, as opposed to HR-MRI using dedicated
[64]. The 3D T1 sequence may offer several advan- surface coil sequences that are centered on the
tages over standard axial T1 sequence: (1) absence abnormal segment of the dissected artery because
of inflow artifacts, thanks to a 3D acquisition and to of a limited number of axial sections in clinically
an outer volume suppression pulse; (2) a compatible acquisition time (Fig. 13). Another
noncardiac gated black blood effect due to the group confirmed the added value of the 3D T1
presence of different intravoxel velocities in the sequence, compared to the fat-suppressed 2D T1
blood vessels, causing a phase dispersion, which sequence, in patients with acute or subacute carotid
results in signal loss. This phenomenon is ampli- or vertebral artery dissection using a standard
fied with the use of long echo trains and low multi-array surface coil at 1.5-T, with the limitation
refocusing flip angles; (3) homogeneous fat sup- of incomplete fat saturation in the lower cervical
pression; (4) a tridimensional acquisition with region [67].
134 O. Naggara et al.

CT Angiography present is from a systematic meta-analysis of non-

Multisection CT angiography (CTA) provides randomized studies that reported on outcomes
high-resolution and high-contrast images of the with antiplatelets versus anticoagulants [70].
arterial lumen and vessel wall. Axial source There was no significant difference in the rates
images provide excellent visualization of the of death or disability between both treatment
craniocervical arteries as a consequence of the groups. Thus, at present, empirical results are
contrast difference between the enhanced arterial used to choose between anticoagulants and
lumen and the surrounding structures. antiplatelet drugs, in a case-by-case-based deci-
Postprocessing of the source data as MIPs or sion. Anticoagulation is sometimes preferred over
curved multiplanar reformats MPR produces antiplatelet drugs when there is a severe stenosis,
images that closely resemble those obtained with an occlusion (with a risk of embolization before
DSA or MRA, and voxel-rendering techniques recanalization), or a pseudoaneurysm, provided
allow a 3D display of the vessels and surrounding that the infarct is not too large. Anticoagulants
soft tissues. CTA of CAD has shown excellent might also be preferred whenever a thrombus is
agreement with DSA [68]. A 100 % sensitivity seen in the arterial lumen or when the presence of
and specificity was achieved using the presence of multiple ischemic lesions in the same arterial terri-
a narrowed eccentric lumen in association with tory suggests an embolic mechanism. Conversely,
enlargement of the overall vessel diameter as a antiplatelet drugs are commonly favored when
criterion for acute carotid dissection. Other signs there is a general contraindication to anticoagulants
of extracranial carotid artery dissection on CT or when CAD is associated with a large infarct,
include stenosis, mural thickening, occlusion, exposing the individual to a high risk of hemor-
aneurysm formation, and thin annular contrast rhagic transformation. Whether antiplatelets
enhancement [69]. should be preferred in cases of local symptoms
Another advantage of CTA examinations over only is still a matter of debate. Antiplatelet drugs
MRI resides in identification of the rare but patho- might also be preferred to anticoagulants in the rare
gnomonic intimal flap. The higher sensitivity of case of intracranial extension of CAD that could
CTA is likely related to both the faster acquisition theorically lead to a subarachnoid hemorrhage.
time and higher spatial resolution. The information of evolution of the mural
hematoma during the first week after initiation of
the treatment may be important. Since CAD is
Treatment characterized by a mural accumulation of blood,
the dissecting hematoma may enlarge under
Medical Treatment anticoagulation, with subsequent lumen
narrowing. It was recently demonstrated, using
Anticoagulants or antiplatelets are usually HR-MRI, that heparin did not promote enlarge-
recommended in the acute phase of CAD to pre- ment of the mural hematoma, progression of lumi-
vent primary or recurrent stroke. However, no nal narrowing, or dissection recurrence much
randomized trial has successfully compared the more than aspirin [71].
efficacy of these treatments. The Cervical Artery The currently available observational studies
Dissection in Stroke Study (CADISS) is an ongo- suggest that thrombolytic therapy should not be
ing randomized multicentric prospective study withheld in patients in whom CAD is associated
comparing antiplatelet therapy with with an acute ischemic stroke. However, the lack
anticoagulation for patients with carotid and ver- of any trend toward a benefit of thrombolysis in
tebral artery dissection. Randomization must be such patients explains the research for more effi-
within 7 days of onset of symptoms. In the cient treatment options including mechanical
absence of randomized data, the best evidence at revascularization strategies.
8 Carotid Artery Dissection 135

Mechanical Revascularization with standard multichannel head and neck coils

Strategies and volumic fat-suppressed T1-WI sequence.

Recently, stent-supported angioplasty has been

used to treat CAD in case of neurological symp- References
toms while on anticoagulation or as an alternative
to the traditionally accepted use of anticoagulation 1. Schievink WI (2001) Spontaneous dissection of the
carotid and vertebral arteries. N Engl J Med
[72]. Stenting permitted resolution of the stenosis
with immediate recanalization of the artery. In a 2. Lee VH, Brown RD Jr, Mandrekar JN, Mokri B (2006)
recent systematic review of 140 patients Incidence and outcome of cervical artery dissection: a
(153 CAD), including traumatic (48 %, n = 64), population-based study. Neurology 67:1809–1812
3. von Babo M, De Marchis GM, Sarikaya H, Stapf C,
spontaneous (37 %, n = 49), and iatrogenic (16 %,
Buffon F, Fischer U, Heldner MR, Gralla J, Jung S,
n = 21) CAD, the technical success rate of Simonetti BG, Mattle HP, Baumgartner RW, Bousser
stenting was 99 %, and the procedural complica- MG, Arnold M (2013) Differences and similarities
tion rate was 1.3 %. At a mean angiographic between spontaneous dissections of the internal carotid
artery and the vertebral artery. Stroke 44:1537–1542
follow-up of 12.8 months, in-stent stenosis or
4. Touze E, Oppenheim C, Zuber M, Meary E, Meder JF,
occlusion was encountered in 2 % of patients, Mas JL (2003) Early asymptomatic recurrence of cer-
and at a mean clinical follow-up of 17.7 months, vical artery dissection: three cases. Neurology
neurological events were seen in 1.4 % of patients 61:572–574
5. Debette S, Leys D (2009) Cervical-artery dissections:
[73]. Further trials comparing stenting with med-
predisposing factors, diagnosis, and outcome. Lancet
ical management are needed to elucidate the role Neurol 8:668–678
of stenting in selected situations. 6. Majidi S, Hassan AE, Adil MM, Jadhav V, Qureshi AI
(2014) Incidence and outcome of vertebral artery dis-
section in trauma setting: analysis of national trauma
data base. Neurocrit Care 21(2):253–258
Summary 7. Engelter ST, Grond-Ginsbach C, Metso TM, Metso AJ,
Kloss M, Debette S, Leys D, Grau A, Dallongeville J,
An early and reliable identification of CAD is of Bodenant M, Samson Y, Caso V, Pezzini A, Bonati LH,
Thijs V, Gensicke H, Martin JJ, Bersano A, Touze E,
importance in order to prevent ischemic brain
Tatlisumak T, Lyrer PA, Brandt T, Cervical Artery D,
lesion. The diagnosis of carotid artery dissection Ischemic Stroke Patients Study G (2013) Cervical
is currently based on (1) a suggestive clinical artery dissection: trauma and other potential mechani-
presentation, (2) exclusion of atherosclerosis, cal trigger events. Neurology 80:1950–1957
8. Renard D, Azakri S, Arquizan C, Swinnen B,
and (3) supportive radiologic evidence. Obtaining
Labauge P, Thijs V (2013) Styloid and hyoid bone
radiological evidence of CAD may be a real chal- proximity is a risk factor for cervical carotid artery
lenge. The emergence of high-resolution, rapid dissection. Stroke 44:2475–2479
imaging methods has enabled MRI to noninva- 9. Raser JM, Mullen MT, Kasner SE, Cucchiara BL,
Messe SR (2011) Cervical carotid artery dissection is
sively image the fine internal structure of the
associated with styloid process length. Neurology
cervical arterial walls. HR-MRI can be used in 77:2061–2066
case of suspicion of cervical artery dissection, 10. Grau AJ, Brandt T, Buggle F, Orberk E, Mytilineos J,
providing an excellent visualization of the cervi- Werle E, Conradt, Krause M, Winter R, Hacke W
(1999) Association of cervical artery dissection with
cal arterial mural hematoma. Because of the lim-
recent infection. Arch Neurol 56:851–856
ited accessibility of 1.5-T HR-MRI and the long 11. Guillon B, Berthet K, Benslamia L, Bertrand M,
scanning time inherent to the technique, the idea is Bousser MG, Tzourio C (2003) Infection and the risk
not to perform HR-MRI in all patients with CAD of spontaneous cervical artery dissection: a case-
control study. Stroke 34:e79–e81
but rather to improve our imaging protocols to get
12. Grau AJ, Brandt T, Forsting M, Winter R, Hacke W
supportive imaging evidence in most patients. (1997) Infection-associated cervical artery dissection.
This is already feasible using 3-T MR magnets Three cases. Stroke 28:453–455
136 O. Naggara et al.

13. Lindsberg PJ, Grau AJ (2003) Inflammation and infec- Nikolaou K, Bartenstein P, Reiser M, Hacker M,
tions as risk factors for ischemic stroke. Stroke Dichgans M (2011) Vessel wall inflammation in spon-
34:2518–2532 taneous cervical artery dissection: a prospective, obser-
14. Rubinstein SM, Peerdeman SM, van Tulder MW, vational positron emission tomography, computed
Riphagen I, Haldeman S (2005) A systematic review tomography, and magnetic resonance imaging study.
of the risk factors for cervical artery dissection. Stroke Stroke 42:1563–1568
36:1575–1580 27. Volker W, Dittrich R, Grewe S, Nassenstein I, Csiba L,
15. Debette S, Markus HS (2009) The genetics of cervical Herczeg L, Borsay BA, Robenek H, Kuhlenbaumer G,
artery dissection: a systematic review. Stroke 40: Ringelstein EB (2011) The outer arterial wall layers are
e459–e466 primarily affected in spontaneous cervical artery dis-
16. Genius J, Dong-Si T, Grau AP, Lichy C (2005) section. Neurology 76:1463–1471
Postacute c-reactive protein levels are elevated in cer- 28. Caplan LR (2008) Dissections of brain-supplying
vical artery dissection. Stroke 36:e42–e44 arteries. Nat Clin Pract Neurol 4:34–42
17. Debette S, Metso TM, Pezzini A, Engelter ST, Leys D, 29. Naggara O, Soares F, Touze E, Roy D, Leclerc X,
Lyrer P, Metso AJ, Brandt T, Kloss M, Lichy C, Pruvo JP, Mas JL, Meder JF, Oppenheim C (2011) Is
Hausser I, Touze E, Markus HS, Abboud S, Caso V, it possible to recognize cervical artery dissection on
Bersano A, Grau A, Altintas A, Amouyel P, stroke brain mr imaging? A matched case-control
Tatlisumak T, Dallongeville J, Grond-Ginsbach C, study. AJNR Am J Neuroradiol 32:869–873
Group C (2009) Cadisp-genetics: an international pro- 30. Biousse V, D’Anglejan-Chatillon J, Touboul PJ,
ject searching for genetic risk factors of cervical artery Amarenco P, Bousser MG (1995) Time course of
dissections. Int J Stroke 4:224–230 symptoms in extracranial carotid artery dissections. A
18. Schievink WI, Mokri B, O’Fallon WM (1994) Recur- series of 80 patients. Stroke 26:235–239
rent spontaneous cervical-artery dissection. N Engl J 31. Biousse V, Woimant F, Amarenco P, Touboul PJ,
Med 330:393–397 Bousser MG (1992) Pain as the only manifestation of
19. Touze E, Gauvrit JY, Moulin T, Meder JF, Bracard S, internal carotid artery dissection. Cephalalgia
Mas JL (2003) Risk of stroke and recurrent dissection 12:314–317
after a cervical artery dissection: a multicenter study. 32. Leys D, Bandu L, Henon H, Lucas C, Mounier-Vehier-
Neurology 61:1347–1351 F, Rondepierre P, Godefroy O (2002) Clinical outcome
20. Calvet D, Boutouyrie P, Touze E, Laloux B, Mas JL, in 287 consecutive young adults (15 to 45 years) with
Laurent S (2004) Increased stiffness of the carotid wall ischemic stroke. Neurology 59:26–33
material in patients with spontaneous cervical artery 33. Steinke W, Schwartz A, Hennerici M (1996) Topogra-
dissection. Stroke 35:2078–2082 phy of cerebral infarction associated with carotid artery
21. Baumgartner RW, Lienhardt B, Mosso M, Gandjour J, dissection. J Neurol 243:323–328
Michael N, Georgiadis D (2007) Spontaneous and 34. Molina CA, Alvarez-Sabin J, Schonewille W,
endothelial-independent vasodilation are impaired in Montaner J, Rovira A, Abilleira S, Codina A (2000)
patients with spontaneous carotid dissection: a case- Cerebral microembolism in acute spontaneous internal
control study. Stroke 38:405–406 carotid artery dissection. Neurology 55:1738–1740
22. Brandt T, Orberk E, Weber R, Werner I, Busse O, 35. Srinivasan J, Newell DW, Sturzenegger M, Mayberg
Muller BT, Wigger F, Grau A, Grond-Ginsbach C, MR, Winn HR (1996) Transcranial doppler in the eval-
Hausser I (2001) Pathogenesis of cervical artery dis- uation of internal carotid artery dissection. Stroke
sections: association with connective tissue abnormal- 27:1226–1230
ities. Neurology 57:24–30 36. Droste DW, Junker K, Stogbauer F, Lowens S,
23. Volker W, Besselmann M, Dittrich R, Nabavi D, Besselmann M, Braun B, Ringelstein EB (2001) Clin-
Konrad C, Dziewas R, Evers S, Grewe S, Kramer SC, ically silent circulating microemboli in 20 patients with
Bachmann R, Stogbauer F, Ringelstein EB, carotid or vertebral artery dissection. Cerebrovasc Dis
Kuhlenbaumer G (2005) Generalized arteriopathy in 12:181–185
patients with cervical artery dissection. Neurology 37. Bonati LH, Wetzel SG, Gandjour J, Baumgartner RW,
64:1508–1513 Lyrer PA, Engelter ST (2008) Diffusion-weighted
24. Mawet J, Boukobza M, Franc J, Sarov M, Arnold M, imaging in stroke attributable to internal carotid artery
Bousser MG, Ducros A (2013) Reversible cerebral dissection: the significance of vessel patency. Stroke
vasoconstriction syndrome and cervical artery dissec- 39:483–485
tion in 20 patients. Neurology 81:821–824 38. Weiller C, Mullges W, Ringelstein EB, Buell U, Reiche
25. Naggara O, Touze E, Marsico R, Leclerc X, Nguyen T, W (1991) Patterns of brain infarctions in internal
Mas JL, Pruvo JP, Meder JF, Oppenheim C (2009) carotid artery dissections. Neurosurg Rev 14:111–113
High-resolution mr imaging of periarterial edema asso- 39. Lucas C, Moulin T, Deplanque D, Tatu L, Chavot D
ciated with biological inflammation in spontaneous (1998) Stroke patterns of internal carotid artery dissec-
carotid dissection. Eur Radiol 19(9):2255–2260 tion in 40 patients. Stroke 29:2646–2648
26. Pfefferkorn T, Saam T, Rominger A, Habs M, Gerdes 40. Morel A, Naggara O, Touze E, Raymond J, Mas JL,
LA, Schmidt C, Cyran C, Straube A, Linn J, Meder JF, Oppenheim C (2012) Mechanism of
8 Carotid Artery Dissection 137

ischemic infarct in spontaneous cervical artery dissec- 54. Benninger DH, Georgiadis D, Gandjour J,
tion. Stroke 43:1354–1361 Baumgartner RW (2006) Accuracy of color duplex
41. Naggara O, Morel A, Touze E, Raymond J, Mas JL, ultrasound diagnosis of spontaneous carotid dissection
Meder JF, Oppenheim C (2012) Stroke occurrence and causing ischemia. Stroke 37:377–381
patterns are not influenced by the degree of stenosis in 55. Dittrich R, Dziewas R, Ritter MA, Kloska SP,
cervical artery dissection. Stroke 43:1150–1152 Bachmann R, Nassenstein I, Kuhlenbaumer G,
42. Naggara O, Touze E, Seiller N, Gobin-Metteil MP, Mas Heindel W, Ringelstein EB, Nabavi DG (2006) Nega-
JL, Meder JF, Oppenheim C (2008) Asymmetry of tive ultrasound findings in patients with cervical artery
intracranial internal carotid artery on 3d tof mr angiog- dissection: negative ultrasound in cad. J Neurol
raphy: a sign of unilateral extracranial stenosis. Eur 253:424–433
Radiol 18:1038–1042 56. Habs M, Pfefferkorn T, Cyran CC, Grimm J,
43. Iancu-Gontard D, Oppenheim C, Touze E, Meary E, Rominger A, Hacker M, Opherk C, Reiser MF,
Zuber M, Mas JL, Fredy D, Meder JF (2003) Evalua- Nikolaou K, Saam T (2011) Age determination of
tion of hyperintense vessels on flair mri for the diagno- vessel wall hematoma in spontaneous cervical artery
sis of multiple intracerebral arterial stenoses. Stroke dissection: a multi-sequence 3t cardiovascular mag-
34:1886–1891 netic resonance study. J Cardiovasc Magn Reson 13:76
44. Touze E, Randoux B, Meary E, Arquizan C, Meder JF, 57. Provenzale JM, Sarikaya B, Hacein-Bey L, Wintermark
Mas JL (2001) Aneurysmal forms of cervical artery M (2011) Causes of misinterpretation of cross-sectional
dissection: associated factors and outcome. Stroke imaging studies for dissection of the craniocervical arter-
32:418–423 ies. AJR Am J Roentgenol 196:45–52
45. Kremer C, Mosso M, Georgiadis D, Stockli E, 58. Oppenheim C, Naggara O, Touze E, Lacour JC,
Benninger D, Arnold M, Baumgartner RW (2003) Schmitt E, Bonneville F, Crozier S, Guegan-
Carotid dissection with permanent and transient occlu- Massardier E, Gerardin E, Leclerc X, Neau JP,
sion or severe stenosis: long-term outcome. Neurology Sirol M, Toussaint JF, Mas JL, Meder JF (2009)
60:271–275 High-resolution mr imaging of the cervical arterial
46. Arauz A, Hoyos L, Espinoza C, Cantu C, wall: what the radiologist needs to know. Radio-
Barinagarrementeria F, Roman G (2006) Dissection graphics 29:1413–1431
of cervical arteries: long-term follow-up study of 59. Bachmann R, Nassenstein I, Kooijman H, Dittrich R,
130 consecutive cases. Cerebrovasc Dis 22:150–154 Kugel H, Niederstadt T, Kuhlenbaumer G, Ringelstein
47. Charbonneau F, Gauvrit JY, Touze E, Moulin T, EB, Kramer S, Heindel W (2006) Spontaneous acute
Bracard S, Leclerc X, Mas JL, Meder JF (2005) Diag- dissection of the internal carotid artery: high-resolution
nosis and follow-up of cervical arterial magnetic resonance imaging at 3.0 Tesla with a dedi-
dissections–results of the sfnv-sfnr study. J cated surface coil. Invest Radiol 41:105–111
Neuroradiol 32:255–257 60. Touze E, Toussaint JF, Coste J, Schmitt E,
48. Leclerc X, Lucas C, Godefroy O, Nicol L, Moretti A, Bonneville F, Vandermarcq P, Gauvrit JY, Douvrin F,
Leys D, Pruvo JP (1999) Preliminary experience using Meder JF, Mas JL, Oppenheim C (2007) Reproducibil-
contrast-enhanced mr angiography to assess vertebral ity of high-resolution mri for the identification and the
artery structure for the follow-up of suspected dissec- quantification of carotid atherosclerotic plaque compo-
tion. AJNR Am J Neuroradiol 20:1482–1490 nents: consequences for prognosis studies and thera-
49. Sturzenegger M, Mattle HP, Rivoir A, Baumgartner peutic trials. Stroke 38:1812–1819
RW (1995) Ultrasound findings in carotid artery dis- 61. Yarnykh VL, Terashima M, Hayes CE, Shimakawa A,
section: analysis of 43 patients. Neurology 45:691–698 Takaya N, Nguyen PK, Brittain JH, McConnell MV,
50. Nedeltchev K, Bickel S, Arnold M, Sarikaya H, Yuan C (2006) Multicontrast black-blood mri of
Georgiadis D, Sturzenegger M, Mattle HP, carotid arteries: comparison between 1.5 and 3 Tesla
Baumgartner RW (2009) R2-recanalization of sponta- magnetic field strengths. J Magn Reson Imaging
neous carotid artery dissection. Stroke 40:499–504 23:691–698
51. Dittrich R, Nassenstein I, Bachmann R, Maintz D, 62. Underhill HR, Yarnykh VL, Hatsukami TS, Wang J,
Nabavi DG, Heindel W, Kuhlenbaumer G, Ringelstein Balu N, Hayes CE, Oikawa M, Yu W, Xu D, Chu B,
EB (2007) Polyarterial clustered recurrence of cervical Wyman BT, Polissar NL, Yuan C (2008) Carotid
artery dissection seems to be the rule. Neurology plaque morphology and composition: initial compari-
69:180–186 son between 1.5- and 3.0-t magnetic field strengths.
52. Houser OW, Mokri B, Sundt TM Jr, Baker HL Jr, Radiology 248:550–560
Reese DF (1984) Spontaneous cervical cephalic arte- 63. Naggara O, Louillet F, Touze E, Roy D, Leclerc X, Mas
rial dissection and its residuum: angiographic spec- JL, Pruvo JP, Meder JF, Oppenheim C (2010) Added
trum. AJNR Am J Neuroradiol 5:27–34 value of high-resolution mr imaging in the diagnosis of
53. Provenzale JM, Morgenlander JC, Gress D (1996) vertebral artery dissection. AJNR Am J Neuroradiol
Spontaneous vertebral dissection: clinical, conven- 31:1707–1712
tional angiographic, ct, and mr findings. J Comput 64. Edjlali M, Roca P, Rabrait C, Naggara O, Oppenheim C
Assist Tomogr 20:185–193 (2012) 3d fast spin-echo t1 black-blood imaging for the
138 O. Naggara et al.

diagnosis of cervical artery dissection. AJNR Am J imaging of cervical artery dissection. AJNR Am J
Neuroradiol 34(9):E103–E106 Neuroradiol 29:1753–1760
65. Naggara O, Oppenheim C, Toussaint JF, Calvet D, 70. Kennedy F, Lanfranconi S, Hicks C, Reid J,
Touze E, Mas JL, Meder JF (2007) Asymptomatic Gompertz P, Price C, Kerry S, Norris J, Markus HS,
spontaneous acute vertebral artery dissection: diagno- Investigators C (2012) Antiplatelets vs anticoagulation
sis by high-resolution magnetic resonance images with for dissection: cadiss nonrandomized arm and meta-
a dedicated surface coil. Eur Radiol 17:2434–2435 analysis. Neurology 79:686–689
66. Bachmann R, Nassenstein I, Kooijman H, Dittrich R, 71. Machet A, Fonseca AC, Oppenheim C, Touze E,
Stehling C, Kugel H, Niederstadt T, Kuhlenbaumer G, Meder JF, Mas JL, Naggara O (2013) Does
Ringelstein EB, Kramer S, Heindel W (2007) High- anticoagulation promote mural hematoma growth or
resolution magnetic resonance imaging (mri) at 3.0 delayed occlusion in spontaneous cervical artery dis-
Tesla in the short-term follow-up of patients with proven sections? Cerebrovasc Dis 35:175–181
cervical artery dissection. Invest Radiol 42:460–466 72. Lavallee PC, Mazighi M, Saint-Maurice JP,
67. Cuvinciuc V, Viallon M, Momjian-Mayor I, Sztajzel R, Meseguer E, Abboud H, Klein IF, Houdart E,
Pereira VM, Lovblad KO, Vargas MI (2013) 3d Amarenco P (2007) Stent-assisted endovascular throm-
fat-saturated t1 space sequence for the diagnosis of bolysis versus intravenous thrombolysis in internal
cervical artery dissection. Neuroradiology 55:595–602 carotid artery dissection with tandem internal carotid
68. Leclerc X, Godefroy O, Salhi A, Lucas C, Leys D, and middle cerebral artery occlusion. Stroke
Pruvo JP (1996) Helical ct for the diagnosis of extra- 38:2270–2274
cranial internal carotid artery dissection. Stroke 73. Pham MH, Rahme RJ, Arnaout O, Hurley MC, Bern-
27:461–466 stein RA, Batjer HH, Bendok BR (2011) Endovascular
69. Vertinsky AT, Schwartz NE, Fischbein NJ, stenting of extracranial carotid and vertebral artery
Rosenberg J, Albers GW, Zaharchuk G (2008) Com- dissections: a systematic review of the literature. Neu-
parison of multidetector ct angiography and mr rosurgery 68:856–866, discussion 866
Nonatherosclerotic Nondissection
Diseases of Carotid Artery 9
Santosh Kumar Kannath and T. R. Kapilamoorthy

Contents Carotid Aneurysms and Carotid Blowout . . . . . . . . 154

Carotid Aneurysm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140 Carotid Blowout Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . 155
Fibromuscular Dysplasia (FMD) . . . . . . . . . . . . . . . . . . 140 Miscellaneous . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
Natural History and Clinical Features . . . . . . . . . . . . . . . 140 Carotidynia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
Pathology and Imaging-Pathologic Correlation . . . . . 141 Eagle Syndrome or Stylohyoid Syndrome . . . . . . . . . . 157
Vasculitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
Takayasu Arteritis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
Giant Cell Arteritis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
Other Uncommon Vasculitis . . . . . . . . . . . . . . . . . . . . . . . 145
HIV Vasculopathy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Connective Tissue Diseases . . . . . . . . . . . . . . . . . . . . . . . . 147
Arterial Tortuosity Syndrome . . . . . . . . . . . . . . . . . . . . . . . . 147
Ehler-Danlos Syndrome Type 4 (EDS4) . . . . . . . . . . . . 147
Loeys Dietz Syndrome (LDS) . . . . . . . . . . . . . . . . . . . . . . . 148
Pseudoxanthoma Elasticum (PXE) . . . . . . . . . . . . . . . . . . 150
Marfan Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
Other Systemic Disease with Carotid
Involvement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
Neurofibromatosis Type 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
PHACE(S) Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Tuberous Sclerosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Behcet Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154

S.K. Kannath (*)

Neurointervention Center, Department of Imaging
Sciences and Interventional Radiology, Neurointervention
Center, Sree Chitra Institute of Medical Sciences and
Technology, Trivandrum, Kerala, India
T.R. Kapilamoorthy
Department of Imaging Sciences and Interventional
Radiology, Sree Chitra Institute of Medical Sciences and
Technology, Trivandrum, Kerala, India

# Springer Science+Business Media New York 2016 139

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_38
140 S.K. Kannath and T.R. Kapilamoorthy

Fibromuscular Dysplasia (FMD)
Nonatherosclerotic and nondissection diseases
Fibromuscular dysplasia is an idiopathic, segmen-
of carotid are often unrecognized clinical
tal, nonatherosclerotic, and noninflammatory dis-
entity, primarily due to nonspecific imaging
ease involving the small- and medium-sized
findings and unavailability of diagnostic tests.
arteries, commonly affecting renal and carotid
This entity comprises of wide spectrum of clin-
arteries. The exact prevalence of the disease is
ical conditions such as nonspecific arteritis,
unknown. The prevalence of the symptomatic
vasculitis, connective tissue diseases, infective
disease based on earlier angiographic studies is
processes, and some acquired conditions.
estimated to be between 0.3 and 3 %, though
Oftentimes, carotid involvement may be the
obviously this could not be extrapolated to general
first and only manifestation of these diseases.
population due to the evident bias and its actual
Ancillary clinical and examination findings
incidence may even be much lower [1–3].
lend clue to the underlying disease process
and help the clinician to further investigate
into the specific pathology. The list is exhaus-
Natural History and Clinical Features
tive; however, some of the important disease
entities and their clinical implications are
The etiology of FMD is not known, and various
reviewed in this chapter.
factors such as familial predisposition, hormonal
influence, smoking, and mechanical forces are
Keywords attributed to the causation of FMD. FMD com-
Carotid aneurysms • Carotid stenosis • monly affects the truncal portion of the internal
Takayasu arteritis • Fibromuscular dysplasia • carotid artery which is prone for mechanical
Giant cell arteritis • Vasculitis • Connective stretching and shearing over the cervical verte-
tissue disorders • Marfan syndrome • Ehler- brae. As this segment relatively lacks vasa
Danlos syndrome • Neurofibromatosis • vasorum, an ischemic cause has also been attrib-
Arterial tortuosity syndrome • Carotidynia • uted. A congenital theory presumes that the arte-
Eagle syndrome rial lesions are persistent embryonic intimal
cushions that normally appear in the main trunk
during fetal life. The fact that proximal cervical
Introduction ICA is formed from dorsal aortic arch lends cre-
dence to this theory [1, 4].
Though atherosclerotic manifestation of the carotid Cervicocranial FMD presents at an older age,
artery is a more common and important vascular though classical FMD is commonly discovered in
pathology in middle and elderly patients, nonather- younger and female population [4]. Carotid and
osclerotic vascular diseases assume prominence in vertebral arteries are nearly as frequently affected
younger population and remain an important cause as renal arteries. FMD may be detected inciden-
for stroke in this age group. Nonatherosclerotic tally or may have ischemic or hemorrhagic pre-
diseases may remain asymptomatic and are inci- sentation. Stroke or transient ischemic attacks
dentally detected or may present with catastrophic (TIA) can occur due to thromboembolic phenom-
events. Due to its variable age of presentation, it ena, dissection, or hemodynamic compromise.
may be confused with atherosclerotic pathologies Spontaneous carotid artery dissections are a
if the latter presents prematurely. Thus understand- known complication, and FMD changes may be
ing of this group of pathologies is vital to design identified in up to 15 % of spontaneous dissec-
appropriate investigations for optimal therapeutic tions [5]. Recent registry data showed incidence
management. In this chapter, some of the important of carotid dissections and aneurysms to be 75 %
nonatherosclerotic and nondissection diseases of and 21 %, respectively. Besides, vertebral
carotid artery are reviewed. artery dissections are also not uncommon [6].
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery 141

The incidence of cerebral aneurysms is reported to the distal carotid artery sparing the carotid
be 20–30 % in FMD, higher than the general bulb. These lesions are stable and do not pro-
population. Subarachnoid hemorrhage is a gress beyond 40 years.
dreaded complication of aneurismal rupture. 3. Perimedial fibroplasias: Occurs in 10 %.
Besides, rarer manifestations such as intracerebral Lesions have a collagen collar of variable
hemorrhage, carotidocavernous fistula or thickness in the outer half or the media or at
vertebrovenous fistulas are also described. Rarely, media-adventitia junction. It can be localized
FMD can involve intracranial arteries such as or diffused. Producing a corrugated appear-
middle cerebral artery, basilar artery, and distal ance, the caliber of outpouching does not
carotid artery in isolation or as extension of cervi- exceed the diameter of the normal proximal
cal FMD. Intracranial FMDs are observed in pedi- artery.
atric population and usually have fatal outcome 4. Medial hyperplasia : Occurs in 1–2 % . There is
[1]. In some follow-up studies, the risk of recur- smooth muscle hyperplasia without fibrosis.
rent TIAs or stroke is reported up to 5 % per year, Beads are less numerous than other subtypes.
and hence it is prudent to keep such patients on 5. Periarterial fibroplasia: Fibrous thickening of
long strict clinical surveillance. Natural history of adventitia with cicatrix formation causes local/
incidental intracranial aneurysms in FMD is diffuse area of narrowing in the artery . This is a
unknown [1, 7]. very rare type occurring in <1 % of cases.

Traditionally, angiography has been used to

Pathology and Imaging-Pathologic diagnose FMD. The frequently affected segments
Correlation include mid and distal parts of ICA and vertebral
artery at C1 and C2 levels. Bilateral involvement
Three types of FMDs are identified histologically is common, and carotid and vertebral artery
depending on the arterial layer involvement. The pathologies may coexist. The most common
angiographic appearance can be correlated with angiographic abnormality is the appearance of a
histological features; however, categorization string of beads which is a characteristic of FMD
into different subtypes is successful only in few and usually suggests subtype medial FMD
entities. The various pathological subtypes (Fig. 1). Fine bristling of lumen may be noted
and its angiographic appearances are described indicating the presence of microaneurysms.
below. Standing columns of spastic contractions are the
differential diagnosis. Tubular stenosis accounts
1. Intimal fibroplasias: Occurs in less than 10 %. for 20 % of angiographic observations which may
There is circumferential thickening of intima be focal or diffuse. It is not related to specific
by collagenous tissue. It is usually seen in subtypes and can be confused with other vascular
children and young adults with no gender pre- diseases such as atherosclerosis or vasculitis.
dilection. The lesions are localized or a long Rarely, a weblike defect at the origin of internal
segment; however, this feature is not specific to carotid artery is also described. Other rare atypical
this entity. manifestations include beading of external carotid
2. Medial hyperplasia: Occurs in 60–70 %. There branches, diverticular outpouching of carotid
are areas of thickened fibrotic ridges alternate artery [1, 2, 4, 8].
with thinned media producing a beaded Intracranial FMD is described in conjunction
appearance due to alternate hyperplasic change with carotid involvement or as isolated findings.
and local aneurysm formation. The beads are Angiography may show a typical string of bead
aneurismal outpouchings that exceed the nor- appearance or multiple aneurysms and arterial
mal proximal caliber of the artery, and constric- tortuosities. Intima is as commonly affected as
tions represent fibrotic ridges. The appearance media, and most cases are reported in pediatric
is a diagnostic of FMD and usually occurs in population [2].
142 S.K. Kannath and T.R. Kapilamoorthy

Fig. 1 Diagnostic angiogram of bilateral common carotid artery (a, b). MRI revealed bilateral parieto-occipital
artery in a 3-year-old boy reveals tapering and occlusion of region acute infarcts (not shown). Volumetric reconstruc-
the right internal carotid artery and typical arterial beading tion shows moderate focal stenosis of midsegment of the
and moderate stenosis of proximal left internal carotid main renal artery (c)

Though typical angiographic appearance is and symptoms can significantly hamper the daily
confirmatory of FMD, the test is invasive, and activities in more than two thirds of TA
similar findings can be observed by other nonin- patients [9].
vasive studies such as US, CTA, or MRA. Distal The cerebral manifestations of TA include
carotid, vertebral artery, and intracranial circula- headache, dizziness, visual disturbances, stroke,
tion are better evaluated with CT or MRI, and in TIA, or rarely subarachnoid hemorrhage and
the current scenario, angiography is only intracerebral bleeds. The reported incidence of
performed if revascularization is planned. Besides these symptoms is quite variable with minor
diagnosis, imaging also helps in the surveillance symptoms noted in more than 50 % of patients
of intracranial aneurysms. and dreaded events such as TIA or stroke
observed in 10–35 % of patients [10, 11]. Ischemic
manifestations may be due to hemodynamic
Vasculitis insufficiency or thromboembolic phenomenon.
Intracranial stenosis is a rare important cause for
Takayasu Arteritis stroke in TA [12]. Subarachnoid hemorrhage sec-
ondary to aneurismal rupture is also rarely
Takayasu arteritis (TA) is a chronic inflammatory reported in TA. The aneurysms are common in
disease of large arteries affecting predominantly the posterior circulation, and persistent hyperten-
aorta, subclavian, carotid, renal, and pulmonary sion and altered hemodynamic factors are thought
arteries. Though reported from all over the world, to play a role in its genesis. Necrosis and destruc-
the disease is more prevalent in Asian population. tion of wall and elastic lamina are also reported;
The affected individuals are usually young and however, primary arteritis seems unlikely [13].
female, and since the initial symptoms are Vasa vasorum (VV) is assumed to play an
nonspecific, the time to diagnosis is often delayed important role in the pathogenesis of
by months to years. Initial inflammatory symp- TA. Inflammatory cells infiltrate the VV and incite
toms include anorexia, weight loss, malaise, fever, the chronic and granulomatous inflammatory
myalgia, arthralgia, or carotidynia. The disease reaction that spreads across to involve the media
shows progressive or relapsing remitting course and adventitia layer of the artery. The adventitia
and finally burns out. The clinical presentation is and media are thickened, and intimal layer shows
often related to the extent of compensatory collat- hyperplasia. The smooth muscles and elastin of
eral development in the affected vascular territory, media are destroyed, and the panarterial layer is
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery 143

gradually replaced by diffuse or nodular fibrotic imaging modalities such as CT/MRI and
lesions resulting in the development of stenosis or US. The angiography is performed as a part of
eventual occlusion of the artery [9, 14, 15]. Exten- interventional procedure or to assess the hemody-
sive destruction followed by inadequate fibrosis namic status of cerebral circulation in patients
response leads to the formation of arterial aneu- presenting with TIA/stroke or hemorrhage.
rysms. The prognosis of TA has improved over the Since carotid artery is superficial, US can
last decade primarily due to earlier diagnosis, uti- detect many of these changes involving common
lization of imaging to assess disease progression, carotid artery and internal carotid artery such as
and advances in medical therapy [16]. The major stenosis, occlusion, or aneurysm formation.
morbidity or mortality is related to the develop- Besides, it can demonstrate the extraluminal arte-
ment of cardiac failure or stroke. Premature rial wall thickening and helps in the diagnosis as
atherosclerosis further adds to the increased cere- well as monitoring of further progression. It also
bral risks. Diagnostic criteria introduced by provides information regarding the direction of
Ishikawa and the American College of Rheuma- flow in carotid and vertebral artery, which is use-
tology are purely clinical and demonstrated more ful while planning therapy. Limitations include
than 90 % sensitivity and specificity for TA. inability to visualize the arteries in entirety.
European consortium for diagnosis of childhood Cross-sectional imaging such as MRI and CT
TA includes imaging cognates as well and had can demonstrate mural thickening of the arterial
reported very high sensitivity and specificity wall, extent of thickening, intraluminal thrombo-
nearing 100 % [9]. sis, steno-occlusive lesions, collateral pathways,
The gold standard imaging study for the eval- and aneurysm formation. MRI has the advantage
uation of TA has been a conventional angiogram, of superior tissue characterization and helps to
though recently, CT and MRI have been increas- delineate inflamed/fibrosed media and adventitia-
ingly employed to study the luminal and periadventitial regions and edematous intima [16].
extraluminal features with high degree of accu- Disease activity monitoring: Activity is best
racy. Typical changes described in cerebral angi- followed up by hematological parameters includ-
ography include: ing erythrocyte sedimentation rate and C-reactive
protein assessment. Recent reports suggest good
1. Irregularity of contour of artery: This is the correlation between the disease activity and the
earliest change seen and is mainly due to inti- delayed gadolinium enhancement of the wall of
mal fibrous thickening. the artery especially aorta. Thus MRI could
2. Stenosis caused by circumferential thickening become an important modality in assessing the
indicating advanced disease. disease progression of the patient. Caution needs
3. Complete occlusion due to thrombosis in a to be exercised; however, while interpretation as
stenosed segment or due to the disease process MRI lacks specificity, wall edema and enhance-
itself. Stenosis or occlusion often occurs at or ment may be observed in clinically dormant phase
close to the origin of the branches of aorta. as well [17, 18].
4. Diffuse dilatation or ectasias due to destruction The imaging studies show stenosis and/or
of media leading to irregular uniform widening occlusion of common carotid arteries, aneurysms
of carotid. of unilateral or bilateral carotid arteries,
5. Aneurysm saccular or fusiform : As there is a intraluminal thrombosis, and rarely dissection of
thick adventitial covering rupture of blood ves- the carotid artery. Extensive collateralization of
sel which is uncommon. external carotid artery through ascending cervical
6. Combination of lesions: Stenosis and and thyrocervical arteries may be visualized
aneurysms. reforming the internal carotid artery. Tortuous
vascular channels may be visualized within the
The angiograms are infrequently performed arterial wall which is in fact the vasa vasorum that
and are increasingly replaced by noninvasive develops secondary to ischemia of the arterial wall
144 S.K. Kannath and T.R. Kapilamoorthy

Fig. 2 This young lady presented with multiple events of b). Thin tortuous channels are seen along the course of the
acute cerebral ischemia. Aortogram demonstrates tapering right common carotid artery that reconstitutes antegrade
of brachiocephalic artery, occlusion of bilateral subclavian flow in the right internal carotid artery (c, d). MRI shows
artery, and moderate intermediate length smooth stenosis old infarct in the right corona radiate (e)
of proximal left common carotid artery (white arrows in a,

or cerebral parenchyma. Intracranial arteries may or may develop later [22, 23]. Ischemic cranial
show occlusion or arterial wall thickening symptoms include headache, scalp tenderness, jaw
(Fig. 2). Rarely nonaneurysmal SAH has been claudication, amaurosis fugax (AF), and partial or
reported [19–21]. complete visual loss. AF is an important premoni-
tory symptom portending worse visual outcomes if
not recognized and intervened at earliest. The
Giant Cell Arteritis visual loss occurs due to narrowing or occlusion
of posterior ciliary arteries and rarely due to central
Giant cell arteritis (GCA), also known as temporal retinal artery occlusion, posterior ischemic optic
arteritis or Horton disease, is the most common neuritis, or retrobulbar neuritis [24]. Cerebral
primary systemic vasculitis affecting people older ischemic events are rare and are reported to occur
than 50 years. The disease incidence increases in 2–4 % of GCA patients [22, 25]. Ischemic events
with age and shows mild female predominance. are common in posterior circulation and may pre-
Vasculitic involvement involves large- and sent as TIAs or stroke. Atypical presentation
medium-sized arteries and frequently affects includes the presence of nonspecific constitutional
cervicocranial arteries such as superficial tempo- symptoms, unexplained fever, claudication pain of
ral artery, posterior ciliary artery, and vertebral extremities due to large vessel GCA involvement,
artery. Aorta, proximal large aortic arteries, and carotid tenderness, or development of aortic aneu-
upper limb arteries are also involved in disease rysm, which in the absence of systemic biomarkers
process; however, intracranial arteries are not typ- can delay timely diagnosis and institution of appro-
ically affected by this disease [22]. priate management.
Clinical symptoms are variable and are related Pathological studies demonstrated infiltration
to developing ischemia or nonspecific inflamma- of all the three layers of arterial wall by chronic
tory processes. Polymyalgic symptoms may be inflammatory cells such as lymphocytes, macro-
seen in 30–50 % of GCA at the time of diagnosis phages, and giant cells. The internal elastic lamina
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery 145

shows fragmentation, and smooth muscles are of aortic or large arterial wall and has been found
destroyed. Typical histological findings such as to be useful in the diagnosis of GCA in its early
the presence of giant cells at the transition of phase or when biopsy is negative. Also this
intima and media are found only in 50 % of the increased uptake in acute phase of GCA has
cases. The hyperplasic changes of intima narrow been correlated with the development of aortic
or obliterate the vascular lumen and contribute to dilatation at late follow-up [30].
the ischemic pathogenesis. However, these
changes are neither specific nor sensitive, and
often a false negative result may occur due to Other Uncommon Vasculitis
segmental nature of the disease [22].
Diagnosis of GCA is made as per the guide- HIV Vasculopathy
lines of the American College of Rheumatology
1990 classification criteria. Ultrasonography and Owing to improved survival rates among HIV
Doppler studies are valuable in the diagnosis of patients, the primary vascular pathology due to
GCA, though they are not commonly employed. HIV infection is increasingly being documented
Ultrasonography may show a dark hypoechoic in clinical practice. Young males are predomi-
circumferential arterial wall thickening termed as nantly affected, and often the patients who are
“halo sign” that disappears 2–3 weeks after the immunocompetent may manifest the disease. Vas-
initiation of steroid therapy. A recent meta- cular pathology may be aneurismal or occlusive in
analysis revealed a sensitivity of 69 % and spec- nature and involve extracranial carotid as well as
ificity of 91 % for unilateral halo sign. For bilat- intracranial arteries. Carotid involvement is noted
eral halo sign, the specificity approaches 100 %, in 25 % of all HIV-related aneurysms, and majority
and some investigators initiate therapy without of these aneurysms present with clinical features
confirmatory evidence from temporal artery such as painful enlarging mass lesions, hoarseness
biopsy. Besides halo sign, stenosis or occlusion of voice, dysphagia, or stridor [31, 32]. Intracranial
of temporal artery could be identified; however, aneurysms may be multiple large or giant fusiform
these findings are neither specific nor sensitive. aneurysms, reported usually in children or isolated
Ultrasonography is operator dependent, and exper- saccular aneurysms in adults [33]. Among the
tise, knowledge of temporal artery anatomy, and patients presenting with stroke, HIV vasculopathy
availability of high-resolution probes can affect the has been found to be the cause in 20 % of the
diagnostic accuracy of GCA [22, 26, 27]. A recent patients [34].
small study reported improved visualization of Exact pathogenesis in unclear and probably
carotid wall, and its vascularization and luminal involves multiple factors. Pathological studies
border by contrast enhanced ultrasonography had demonstrated infiltration of vasa vasorum by
[28]. High-resolution MRI and MRA with 1.5T neutrophils, macrophages, and plasma cells. The
and 3T may show inflammation of temporal artery resultant vasa vasoritis leads to ischemia and
visualized as rim of thickening and enhancement destruction of arterial wall and incites the devel-
of the arterial wall with central dark lumen due to opment of arterial aneurysms. There is fragmen-
spin flow (Fig. 3). The performance of MRI/MRA tation and loss of internal elastic lamina, thinning
was found to be comparable to USG and of outer arterial wall, and intimal fibrosis. Bland
Doppler [29]. thrombus occluding the lumen may also be dem-
Angiography is rarely performed these days onstrated. Opportunistic bacterial infection of
and is largely superseded by noninvasive modal- arterial wall is a distant possibility and should be
ities. MRA and CTA are useful in detecting extra- evaluated for if tissue specimen is available for
cranial vascular abnormalities of aorta or other analysis [32, 34–36].
large arteries (Fig. 4). FDG-PET study has the Vascular abnormalities are detected during rou-
potential to detect early features of inflammation tine surveillance or while evaluating specific
146 S.K. Kannath and T.R. Kapilamoorthy

Fig. 3 Biopsy has proven giant cell arteritis. Extensive thrombus is visualized in the proximal left common carotid
aortic and arch vessel wall thickening and severe luminal artery (arrow in panel c)
narrowing and irregularity are noted (a, b, d). Intraluminal

clinical manifestations. Ultrasound is widely better delineate these pathologies and are useful
available and can easily demonstrate pseudo- in detecting intracranial vascular abnormalities
aneurysms, stenosis, or occlusion of extracranial as well. Intracranial aneurysms are common in
carotid artery. Pseudoaneurysms are commonly the circle of Willis, first- or second-order
localized to distal common carotid artery or branches of MCA, ACA, or PCA. Besides diag-
carotid bifurcation, and it is visualized as blow- nosis, imaging is also useful in monitoring the
out defect in the arterial wall with turbulent color growth of the aneurysm and in assessing the
flow. The adjacent arterial wall shows thickening response to retroviral therapy as some of the
and hyperechoic spotting [37]. CT and MRI can lesions resolve with treatment [31, 34, 38].
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery 147

Fig. 4 Giant cell arteritis. Fat saturated T1W contrast arterial wall is subtly appreciable in source images of TOF
enhanced MRI (a, b) shows soft tissue enhancement in MRA (c, d), highlighted in panel e (arrow). The thickened
the right temporal fossa (arrow in a) which is readily arterial wall shows contrast enhancement (f)
apparent when compared to opposite side. Thickening of

features include dysmorphic facies, skin laxity,

Connective Tissue Diseases hyperextensible joints, arachnodactyly, high
arched palate, and bilateral inguinal hernia
Arterial Tortuosity Syndrome which may mimic other CTDs such as Ehler-
Danlos or Marfan syndrome [39–41].
Arterial tortuosity syndrome is a rare autosomal
recessive connective tissue disorder of large- and
medium-sized arteries due to mutations of
SL2A10 gene encoding the glucose transporter Ehler-Danlos Syndrome Type 4 (EDS4)
receptor GLUT10. Impaired GLUT10 activity
leads to upregulation of TGF beta activity in the Also known as vascular EDS, EDS4 is an autoso-
arterial wall and results in poor elastogenesis and mal dominant CTD caused by heterogeneous
extracellular matrix formation. The hallmark mutation of COL3A1 gene located in the chromo-
findings include arterial tortuosity and elonga- some 2 and encodes type 3 collagen that predom-
tion, aneurysms, and dissections of large- and inates the walls of arteries and abdominal hollow
middle-sized arteries, focal coarctation of aorta, visceral organs. This entity is rare and constitutes
stenosis of great vessels of the arch, and pulmo- <10 % of all EDS patients. The arterials walls are
nary artery (Fig. 5). Associated extravascular fragile and thin and are easily susceptible to shear
148 S.K. Kannath and T.R. Kapilamoorthy

Fig. 5 Arterial tortuosity syndrome in a 3-month-old child. Arch, brachiocephalic artery, left common carotid artery, and
descending aorta are tortuous and elongated (a–c). The left brachial artery is small in caliber and appears quite tortuous (d)

stress injuries especially in large- and medium- vessel wall, the arteries are easily damaged while
sized arteries [42–45]. The four cardinal features obtaining arterial access or during the cerebral
that help in the diagnosis of EDS4 include dys- angiography. Catastrophic and life-threatening
morphic facial features, thin skin with visible complications are reported during the procedure,
subcutaneous veins, bruises, echymosis or hema- and hence invasive diagnostic tests are best
tomas, and features of fragility or rupture of artery, avoided unless performed as a part of interven-
digestive tract, or uterus. The typical features of tional procedure [48–50]. Imaging surveillance by
EDS such as skin and joint laxity are however CT/MRI or US is advocated to detect or monitor
uncommon in this entity. The diagnosis is con- the progress of vascular disease.
firmed by demonstration of deficient synthesis of
type 3 procollagen by cultured skin fibroblasts or
detection of mutation of COL3A1 gene. Loeys Dietz Syndrome (LDS)
Major catastrophic events such as visceral or
arterial rupture are uncommon in childhood; how- LDS is an autosomal dominant connective tissue
ever, nearly 80 % suffer from at least one such disease caused by the heterozygous mutation of
episode by 40 years. The average life expectancy genes encoding the receptors such as TGFBR1,
is around 50 years, and cause of death is due to TGFBR2, SMAD3, and TGFB2 that are involved
vascular complications in more than 90 % of the in the transformation growth factor signaling
patients [45, 46]. The carotid and intracranial pathway. Phenotypic features include aortic aneu-
events are reported to occur in about 15 % of rysms, arterial tortuosity, hypertelorism, bifid
patients. The typical complications include arte- uvula, or cleft palate. The aortic aneurysms are
rial dissection, carotidocavernous fistula, carotid frequently observed in LDS patients, and it
or intracranial aneurysms, and arterial rupture remains the major cause of morbidity and mortal-
(Figs. 6 and 7). The EDS4 remains an important ity [51, 52]. Besides, generalized arterial tortuos-
differential diagnosis for ischemic stroke in youn- ity, tortuosity of head and neck arteries,
ger patients [44, 47]. Due to inherent weakness of dissections, and intracranial aneurysms are also
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery 149

Fig. 6 A middle-aged lady presenting with left carotico- proximal cavernous internal carotid artery (arrow in a).
cavernous fistula with clinical features of EDS. Dilatation Panel b shows direct carotico-cavernous fistula with sig-
of distal cervical internal carotid artery and old dissection nificant cortical venous reflux. Also note fusiform dilata-
sequelae are noted (Arrow in a, c). Note small aneurysm of tion of V4 segment of the left vertebral artery (arrow in d)

Fig. 7 Another patient

with EDS. Extensive
tortuosity and mild ectasias
of bilateral cervical
vertebral arteries are evident
150 S.K. Kannath and T.R. Kapilamoorthy

reported. The other neurological features include long-term structural stability, transfer and reorient
Chiari 1 malformation, craniosynostosis, and the stress load, stabilize the TGF beta binding
hydrocephalus. The aneurysms are large and proteins, and regulate its signaling. Fibrillin 1
symptomatic, reported to occur in 10 % of the mutation is not identified in one third of the
patients causing a mortality rate due to cerebral patients with marfanoid features, and instead,
hemorrhage of 7 % [51, 53]. Among the several mutations of TGFBR1 and TGFBR2 may be
factors that are taken into consideration in the detected. These patients are grouped as MS type
management of these aneurysms, the presence of 2, and ocular features in this subset are uncommon
arterial tortuosity cannot be overlooked as it sig- [58–60]. Revised Ghent criterion enables the
nificantly hampers the distal access of interven- establishment of MS based on cardinal features,
tional devices. Both neurosurgical clipping and family history, genetic screening tests, and
endovascular coiling have been reported for the weighted systemic scores of ancillary clinical
management of cerebral aneurysms [51, 53, 54]. findings [61].
Typical pathological features of large caliber
arteries include cystic medial necrosis, elastin
Pseudoxanthoma Elasticum (PXE) fragmentation, smooth muscle necrosis, and apo-
ptosis. Cystic medial necrosis is uncommon in
PXE is an autosomal recessive disease due to the intracranial arteries. A few pathological studies
mutation of ABCC6 gene on chromosome 16, the reported intimal hyperplasia, medial degenera-
defect of which results in the accumulation of tion, and internal elastic lamina fragmentation of
mineralized and fragmented elastic fibers in the intracranial arteries – findings that may predispose
skin, eyes, and arterial wall. The intima of the to the formation of aneurysms [62, 63]. The vas-
artery is thickened, calcified, and fragmented. cular findings of carotid arteries include arterial
The typical manifestations of the disease include tortuosity (similar to LDS), intracranial and extra-
skin laxity, angioid streaks in retina, and cardio- cranial aneurysms, and dissection. The isolated
vascular changes. The disease is rare and cerebral involvement of the neck and intracranial arteries
complications are reported in 10 % of the PXE is extremely uncommon, limited to small series or
patients. The neurological manifestations include case reports. Dissecting aneurysms secondary to
cerebral ischemia due to carotid or vertebral artery extension of aortic dissection may be more fre-
occlusion, carotid artery calcifications, and cervi- quent observation and may present with neurolog-
cal or intracranial aneurysms. Dissections and ical deficits. The saccular aneurysms are observed
association of rete mirabile of carotid artery are in intracranial circulation, while fusiform type is
also rarely reported [43, 55–57]. common in extracranial carotid arteries. The
patients may be asymptomatic or may present
with pulsatile swelling in the neck, TIA or strokes,
Marfan Syndrome or rarely subarachnoid hemorrhage [62, 64–66].

Marfan syndrome (MS) is a relatively common

connective tissue disease characterized by typical Other Systemic Disease with Carotid
cardiovascular, musculoskeletal, spinal menin- Involvement
geal, and ocular features. The disease is autosomal
dominant inherited; however, in one third of Neurofibromatosis Type 1
patients, new mutations are observed. The defect
lies in the synthesis of fibrillin 1, an extracellular Neurofibromatosis type 1 is a rare autosomal
matrix protein that forms one of the important dominant disease caused by the mutation of
components of elastic microfibril. The microfi- tumor suppressor gene located in the chromosome
brils support elastic membrane and provide 17 that encodes the protein neurofibromin.
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery 151

Neurofibromin regulates cellular mitogenesis by pediatric population was found to be 2.5 % in a large
inhibiting the reticular activating system (ras) institutional-based study. Besides moyamoya pat-
pathways, and hence a loss function leads to tern, isolated stenosis, occlusions, or ectasias of
overexpression of mitogenic signals within the intracranial arteries are observed (Figs. 8 and
cell leading to cellular proliferation or differenti- 9) [75].
ation [67]. The clinical hallmarks of the disease Vertebral arteriovenous fistula is a well recog-
are café-au-lait spots, cutaneous neurofibromas, nized complication of neurofibromatosis with
bone changes, and iris hamartomas. The vascular fewer than 50 cases reported in the literature
involvement is uncommon and is reported to (Fig. 10). The etiology of arteriovenous fistula is
occur in 0.4–6.4 % of patients with NF1 not clearly understood. Spontaneous rupture
[68]. NF1 vasculopathy affects large-, medium-, predisposed by arterial wall thinning and fragility
and small-sized arteries. The arteries may be sec- to adjacent venous plexus is thought to be one of
ondary involved when its wall is weakened by the mechanisms. A congenital theory assumes
extrinsic compression of extravascular neurofi- preexisting abnormal congenital vascular commu-
bromas or due to the proliferation of Schwann nications within the arterial wall which develops
cell within the arterial layers. Recent studies dem- into the fistula. Clinical features of this condition
onstrated a marked intimal and smooth muscle include pulsatile mass lesion, radiculo-
cell proliferation, and this observation has been myelopathy, cranial neuropathy, or tinnitus. The
related to decreased expression of neurofibromin fistula may be found involving the upper or lower
in the vascular wall. It is unsure whether the cervical vertebral arteries or may involve one of
vascular changes are congenital or acquired as a the branches of the vertebral artery. The epidural
pathological response to injury [68–71]. The path- venous sacs are often large and demonstrate rapid
ophysiology of vascular abnormality is related to shunting. Flow reversal and arterial steal may
the amount of intimal proliferation, spindle cell often be identified, and sometimes, small commu-
proliferation in the media, and the arterial diame- nication channels open up when the main fistula is
ter. Hence, small-sized arteries are often occluded, occluded, and hence long-term surveillance is
whereas larger arteries show aneurismal changes advocated. Recent reports suggest good clinical
due to weakening of arterial wall [69–72]. outcome with various endovascular strategies
The NF1 vasculopathy is the second leading [76–78].
cause of death among the NF1 patients younger
than 40 years. The carotid, vertebral, or intracra-
nial arteries are affected in one fifth of the patients, PHACE(S) Syndrome
and vascular abnormality may be in the form of
arterial stenosis, aneurismal changes, or arteriove- PHACE(S) is an acronym for a neurocutaneous
nous malformations. The aneurismal changes are syndrome with salient features of Posterior fossa
the commonest pathology and show female pre- malformation, Facial hemangiomas, Arterial
ponderance and predilection to rupture spontane- anomalies, Coarctation of aorta and Cardiac
ously during pregnancy. Extracranial aneurysms defects, Eye anomalies, and ventral developmen-
are large and fusosaccular and may present as pul- tal anomalies such as Sternal clefts or
satile or enlarging neck mass or acute rupture. Intra- Supraumbilical raphe. The vascular lesions are
cranial aneurysms are relatively uncommon and common in PHACES, seen in up to 78 % of
may affect intradural ICA, middle cerebral, anterior patients. The abnormalities are either congenital
cerebral, or basilar arteries [68, 69, 71, 73, 74]. or progressive and are related to the development
The aneurysms are incidentally detected or present or phase of growth of cutaneous hemangioma.
with subarachnoid hemorrhage. Rarely coexistent The arterial anomalies are detected in the location
intracranial and extracranial aneurysms also may be of hemangioma, and this association is hypothe-
found. The incidence of cerebral vasculopathy in sized due to vascular expression factors released
152 S.K. Kannath and T.R. Kapilamoorthy

Fig. 8 A young patient with neurofibromatosis. MRI (a) supraclinoid internal carotid artery (d) and excessive tor-
shows gliosis and atrophy of the right frontal operculum. tuosity of the left superior and inferior divisions of middle
MR angiogram (b, c) shows occlusion of the right internal cerebral artery (e). Cervical and intracranial vertebral arter-
carotid artery and small caliber of the right middle cerebral ies are also tortuous (f)
artery. Diagnostic angiogram reveals occlusion of the right

by the hemangioma either during embryogenesis carotid-vertebral communications [79–81]. Pro-

or proliferative growth. The hemangiomas of face, gressive vasculopathy such as stenosis,
orbit, or scalp are associated with vascular moyamoya pattern, or occlusions is observed in
malformations of intracranial or carotid arteries, 40 % of the patients. These vascular findings are
whereas mandibular, neck, or chest lesions are observed in more than 90 % of the patients
related to cardiac, aortic, or supraaortic abnormal- presenting with clinical features of acute stroke
ities [79, 80]. [82]. Internal carotid arteries are commonly
The congenital vascular anomalies observed affected followed by middle cerebral, anterior
include arterial dysplasias, aberrant course or ori- cerebral, and vertebrobasilar arteries. Associated
gin, hypoplasia, or agenesis of different intracra- brain structural anomalies are evident in the brain
nial arteries, dolichoectasia, saccular intracranial in 41 % of the patients, more commonly in poste-
aneurysms, and presence of persistent embryonic rior fossa [81].
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery 153

Fig. 9 Young girl with neurofibromatosis 1, presenting middle cerebral artery (a, c) with striate collaterals (b)
with acute stroke. Diagnostic angiogram of left internal suggestive of Moyamoya disease
carotid artery reveals occlusion of proximal segment of

Fig. 10 A middle aged

female with pulsatile mass
lesion in the neck. MRI (a)
show tortuous and dilated
left cervical vertebral artery
emptying into a large
pouch. This is confirmed in
the diagnostic angiogram
(b) She was diagnosed to
have neurofibromatosis 1

Tuberous Sclerosis mapped on chromosome 9 and chromosome

16, respectively. Besides typical hamartomatous
Tuberous sclerosis (TS) is an autosomal dominant lesions discovered in different organs, vascular
multisystem disorder caused by the mutation of involvement is extremely rare and is more com-
tumor suppressor genes, TSC1 and TSC2, monly observed in the aorta. To date, only three
154 S.K. Kannath and T.R. Kapilamoorthy

extracranial carotid aneurysms and 19 intracranial mass or neurological symptoms such as TIA or
aneurysms are reported in the literature. Intracra- stroke. Indeed, luminal thrombus is noted in one
nial aneurysms are more often reported with the fifth of such lesions. Detailed discussion of this
most commonly affected being the internal carotid entity is dealt with elsewhere. Other causes of true
artery. The aneurysms are giant and fusiform and aneurysms are rare and are described in the pre-
may affect multiple arteries. Vascular wall disin- ceding sections. Pseudoaneurysms are relatively
tegration due to accumulation of mucopolysac- uncommon and accounts for 14 % of all extracra-
charides is posited in its genesis [83–85]. nial carotid aneurysms. The most common cause
of such aneurysm includes blunt or penetrating
trauma and previous carotid surgeries. Direct pen-
Behcet Disease etrating injury by sharp objects, shrapnels, or iat-
rogenic carotid puncture either intentionally or
Behcet disease (BD) is a multisystem inflamma- while central venous access can injure the vessel
tory disease of unknown etiology that presents wall and produce localized or expanding hema-
with typical clinical triad of uveitis and oral and toma and pseudoaneurysm. Blunt injuries damage
genital ulcers. Vascular involvement is observed the arterial wall, when the shear or rotary forces
in 7–29 % of patients, and venous systems are compress and stretch the arterial wall against the
predominantly affected. Arterial involvement is cervical vertebra or skull base and promote aneu-
noted in 3–5 % of the patients, and aneurismal rysm formation. Also the dissecting process of
rupture remains to be the most common cause of carotid artery splits the internal elastic lamina
mortality in patients with BD. Pathomechanism of and muscular media and weakens the wall which
aneurysm formation is thought to be due to vasa may eventually turn aneurismal in one third of
vasoritis and vasculitis of the arterial wall. Carotid dissection [92, 93]. The isolated aneurysms of
artery involvement is very rare and may affect the external carotid artery are very rare [94].
common carotid or internal carotid artery. The Untreated extracranial aneurysms carry poor
patient may present with enlarging pulsatile prognosis with high mortality [93, 95]. Besides
lump in the neck or subarachnoid hemorrhage central neurological features, aneurysms may
[86–88]. Intradural aneurysms are rarer and bulge into throat or compress adjacent structures
show no segmental predilection. Multiple aneu- such as the esophagus and lower cranial nerves.
rysms may be found, and these patients show The most feared complication is aneurysm expan-
tendency to develop aneurysms at different loca- sion and rupture resulting in a massive exsangui-
tions on follow-up [89, 90]. nation and fatal outcome. Initial diagnostic
evaluation is done with ultrasonography and
Doppler which confirm the clinical diagnosis,
Carotid Aneurysms and Carotid delineate the aneurysm boundary and extent, and
Blowout demonstrate intraluminal thrombosis. Larger
aneurysms, aneurysms close to skull base and
Carotid Aneurysm presence of calcifications hinder detailed analysis.
However, it is a useful tool in following the
Aneurysms of extracranial carotid artery are growth of the aneurysm in susceptible individuals
uncommonly observed and constitute about 4 % or in patients on conservative therapy. CT angiog-
of all peripheral aneurysms and less than 5 % of all raphy is currently the investigation of choice for
carotid surgeries [91]. Carotid aneurysms are cat- the evaluation of cervical aneurysms. CTA delin-
egorized into true aneurysms or pseudoaneurysms eates the extent and boundary of the aneurysm, its
depending on whether the layers of arterial wall relation to other vascular and airway structures,
are intact or not. Atherosclerosis remains the most presence of thrombosis or soft tissue hematoma,
common cause of true aneurysm, seen in elderly and status of intracranial vasculature. Besides
hypertensives, and presents with pulsatile neck diagnosis, it provides wealth of information
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery 155

Fig. 11 Carotid aneurysms in two patients (a, b, and c, respectively) with no identifiable etiology. Luminal irregularity
and thrombosis with calcifications are present (not shown)

influencing the choice of treatment as well as over of characteristic signal abnormalities of hematoma
anticipated difficulties. Access vessel tortuosities, helps to clinch the diagnosis. Due to rarity of the
presence of kinks, thrombosis, and disrupted cir- disease, the role of CT and MRI as a comprehen-
cle of Willis are some of the factors considered sive imaging study is not explored, and the current
while planning the therapeutic approach [96, 97]. practice is dependent on local institutional protocol
Rare involvement of external carotid artery or and availability of resources. Conventional angiog-
concomitant intra-extracranial disease can also raphy is an invasive modality, usually performed as
be detected by CT angiography. Moreover, a part of interventional procedure (Fig. 11) [100].
craniocervical CTA as part of whole-body CT
protocol in trauma patients enables early recogni-
tion of vascular abnormalities and allows institu- Carotid Blowout Syndrome
tion of appropriate therapy at the earliest [98, 99].
Disadvantages such as contrast overload and radi- The carotid blowout syndrome (CBS) is a life-
ation concerns are largely superseded by the over- threatening emergency caused by the rupture of
all benefits. Potential pitfalls include overlooking the carotid artery or its branches due to the weak-
subtle imaging features such as vessel wall irreg- ening of adventitial layer. The important factors
ularity, spasm, or small peripheral arterial aneu- predisposing CBS include trauma, prior neck irra-
rysms especially when images are degraded by diation for head and neck malignancies, or radical
swallowing artifacts. neck surgeries. Radiation increases the risk by
MRI and MRA are complementary to CT in the 7.6-fold and can occur anytime between 2 and
diagnosis of these aneurysms. MRI is useful in the 20 years after the therapy. Radiation induces
diagnosis and characterization of luminal thrombus inflammatory changes, fibrosis, and ischemia of
and delineation of extent of the aneurysm and adventitia and results in the weakening of arterial
status of adjacent soft tissue structures. Besides wall. The patient may be clinically silent or may
distinct advantage of absent ionizing radiation, present with acute massive exsanguination. The
CT scores over MRI in the imaging workup of latter often carries high morbidity and mortality.
carotid aneurysms due to its easy availability and Despite successful treatment by endovascular par-
reduced scanning time except in situations of clin- ent artery occlusion or stent graft placement,
ical dilemma where CT is inconclusive and strong recurrent CBS and procedural or postprocedural
suspicion of pseudoaneurysm exists. Identification complications are reported to occur in more than
156 S.K. Kannath and T.R. Kapilamoorthy

one fourth of the patients, and overall prognosis is [106]. In the appropriate clinical setting, this
poor due to natural progression of the disease or imaging appearance should raise the suspicion
treatment-related issues [101, 102]. of CBS.
Clinically, CBS is categorized into three Angiogram is performed as a diagnostic pro-
groups – acute, impending, and threatened. cedure prior to intervention or when patients pre-
Acute CBS presents with massive hemorrhage sent with acute CBS. Angiography may
that is not self-limiting and cannot be controlled demonstrate acute contrast extravasation,
by packing or pressure. The artery ruptures freely pseudoaneurysm formation, focal luminal
into the communicating wound or into the airway. narrowing, or luminal irregularity. Findings may
Without immediate resuscitation and treatment, involve common carotid, internal carotid, or
the patient succumbs to rapid exsanguination. external carotid artery branches and may involve
The impending CBS bleeds intermittently through more than one site too. In threatened CBS, the
the oral cavity or surgical wound indicating a angiographic findings are often normal [101, 103,
weakened or ruptured artery with pseudo- 105]. Diffuse blush in the soft tissue suggests
aneurysm formation. Without further treatment, inflammatory or telangiectatic changes secondary
full-blown arterial rupture may ensue anytime. to radiation and merits treatment if patient is
Exposed artery as a result of wound infection, symptomatic.
flap necrosis, or tissue breakdown suggests inev-
itable rupture without any treatment, and it is
termed as threatened CBS [103]. Miscellaneous
Recent study has shown that CTA is very use-
ful in identifying the patients presenting with the Carotidynia
clinical features of impending CBS [104] The
CTA findings include extravasation of contrast, The idiopathic carotidynia (CD) is characterized by
pseudoaneurysms, recurrence of tumor, necrosis, unilateral neck pain and tenderness over the carotid
and exposed main or branches of carotid arteries. artery bifurcation. Due to the diagnostic confusion
The pseudoaneurysm and extravasation may be with other diseases that have similar clinical symp-
observed in the common carotid artery, internal toms, carotidynia was removed by the International
carotid artery, or external carotid artery branches. Headache Society from the 2004 classification of
Petrous ICA is a common location of aneurysm in headache disorders [107, 108]. However, there are
patients who have received radiotherapy for naso- many isolated reports that describe radiological and
pharyngeal carcinoma. The artery is defined as structural changes of the carotid wall, thus empha-
exposed if necrosis contacts more than half of its sizing its existence as a unique entity possibly
circumference. These findings showed good cor- induced by inflammation. The symptoms of
relation with angiographic features and have carotidynia include unilateral neck pain, radiating
value in prognosticating the outcome following to face and ear, aggravated by lateral head move-
intervention. The patients with normal CTA find- ment, swallowing, or chewing. The disease is self-
ings and impending CBS had better clinical out- limited and may be preceded by flu-like symptoms.
come as compared to the group of patients who CD commonly affects young and middle-aged
demonstrated these findings. Disadvantages of individuals and women. Imaging shows the
CTA include its inapplicability in acute CBS, absence of structural vascular and nonvascular
decreased sensitivity in detecting small causes, though many reports suggest specific imag-
pseudoaneurysms, or contrast extravasations ing appearances in carotidynia [109–115]. Acute
[101, 104, 105]. Imaging appearance of CBS in phase reactants such as C-reactive protein (CRP)
MRI is also reported. Pseudoaneurysm was visu- and serum amyloid antigen (SAA) are reported to
alized as enhancing lesion close to the carotid correlate with disease activity [115, 116]. A gamut
artery that appeared isointense on T1-WI of differential diagnosis need to be ruled out before
sequences and hyperintense on T2-WI sequences confirming the diagnosis of CD.
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery 157

Pathological study was reported only in one bone, forms stylohyoid complex (SHC), and a
report where a patient with classical carotidynia reduced distensibility of this complex with sub-
was found to have a thickened carotid adventitia sequent irritation of neurovascular structures is
intraoperatively and histological analysis implicated in the pathogenesis of Eagle syn-
revealed adventitial infiltration of lymphocytes, drome (ES) [118]. The embryological develop-
vascular and fibroblasts proliferation, and early ment of SHC is controversial and is thought to
fibrosis – picture suggesting chronic inflamma- differentiate from Reichert cartilage which is the
tory process [112]. Noninvasive imaging modal- cartilage of the second branchial arch. One the-
ities detect structural abnormalities in the ory focuses on the disintegration of this cartilage
periarterial carotid space, though these changes into five components such as tympanohyal,
are not correlated with histological analysis. stylohyal, ceratohyal, hypohyal, and basohyal.
Ultrasonography and Doppler studies showed Tympanohyal and stylohyal contribute to the
eccentric hypoechoic wall thickening of carotid base and styloid process, respectively; hypohyal
bulb and distal common carotid artery which may and basohyal form the hyoid bone; and
or may not be associated with mild non-flow- ceratohyal in humans is thought to degenerate
limiting luminal narrowing. The hypoechoic and forms the stylohyoid ligament. The second
soft tissue may extend outwardly and corroborate theory assumes the division of the cartilage: the
with the location of tenderness or nodule proximal and larger part contributes to the for-
[109–111]. CT is not sensitive to detect subtle mation of styloid process, distal segment into
soft tissue changes due to lower resolution. PET hyoid bone, and the intervening mesenchyme
CT has shown increased [18 F] fluorodeox- develops into stylohyoid ligament [119–122].
yglucose (18 F-FDG) uptake in the region of The styloid process is related to important
enhancing the soft tissue around the carotid neuro and vascular structures, and thus a deviant
[117]. MRI study may reveal intermediate signal and angled styloid process may impinge upon
intensity tissue around the distal common carotid these structures and cause clinical symptoms.
and carotid bulb limited to carotid sheath on The styloid process is related to the 9th to 12th
T1-WI sequence. These abnormal tissues show cranial nerves, cervical sympathetic chain,
a marked homogenous enhancement, more con- branches of external carotid artery, internal carotid
spicuous on fat-saturated T1-WI sequence. The artery, and internal jugular vein. The syndrome is
luminal caliber of the artery is almost always commonly observed in women in the fourth or
normal. Clearly, MRI delineates the abnormal fifth decades. Typical symptoms include unilat-
tissue as well as the extent of involvement and eral sharp or dull aching pain in the lateral neck,
also helps to rule out other causes that may mimic angle of mandible, oral cavity, or base of the
CD [109, 111, 113–115]. tongue which are precipitated or aggravated by
neck movement, swallowing, or yawning. Com-
pression of carotid artery may cause dull aching
Eagle Syndrome or Stylohyoid pain in the parietal region, TIA, stroke, and
Syndrome Horner syndrome exacerbated by rotary move-
ments of head to contralateral side or it may
The long styloid process defined as more than even cause sudden death [119, 120, 123]. Direct
3 cm in length is seen in 4 % of population, mechanical impingement on the cervical carotid
though the typical symptoms attributed to it are artery also has been implicated as an important
seen in just about 4 % of these patients. The risk factor for carotid artery dissection or aneu-
styloid process is a long slender bone attached rysm [124, 125].
to the base of the temporal bone and is related A gamut of possibilities needs to be excluded
closely to many important neurovascular struc- before one considers the diagnosis of ES. Oral
tures. The styloid process, together with examination of tonsillar fossa may reveal unduly
stylohyoid ligament and lesser cornua of hyoid elongated styloid process. Palpation of tonsillar
158 S.K. Kannath and T.R. Kapilamoorthy

fossa may provoke the pain which is abolished by Gornik L (2012) The United States registry
instillation of local anesthetic along the styloid for fibromuscular dysplasia: results in the first
447 patients. Circulation 125:3182–3190
process. Radiographic studies such as submento- 7. Wells RP, Smith RR (1982) Fibromuscular dysplasia
vertical view or orthopantomogram are useful to of the internal carotid artery: a long term follow-up.
reveal enlarged styloid process. The styloid pro- Neurosurgery 10:39–43
cess is considered enlarged if its length exceeds 8. Mettinger KL, Ericson K (1982) Fibromuscular dys-
plasia and the brain observations on angiographic,
one third of the length of the ramus of the mandi- clinical and genetic characteristics. Stroke 13:46–52
ble or if the length is >3 cm. CT and CTA studies 9. De Souza AWS, de Carvalho JF (2014) Diagnostic
can accurately depict the length, orientation, and and classification criteria of Takayasu arteritis.
relation of styloid process to major vascular struc- J Autoimmun 48–49:79–83
10. Ishikawa K (1978) Natural history and classification
tures [126]. Though invasive, 3D rotational angi- of occlusive thromboaortopathy. (Takayasu’s dis-
ography can directly visualize the mechanical ease). Circulation 57:27–35
impression on carotid or other vessels by elon- 11. Kim JH, Suh DC, Kim JK, Kim SJ et al (2005) Cor-
gated styloid process, it is a useful tool to confirm relation of neurological manifestations of Takayasu’s
arteritis with cerebral angiographic findings. J Clin
and localize disease in doubtful and confusing Imaging 29:79–85
situations [127]. 12. Ringleb PA, Strittmatter EI, Loewer M, Hartmann M
et al (2005) Cerebrovascular manifestations of
Takayasu arteritis in Europe. Rheumatology
Summary 13. Kim DS, Kim JK, Yoo DS, Huh PW, Cho KS, Kang
JK (2002) Takayasu arteritis presenting with sub-
The clinical disease entities encompassing arachnoid hemorrhage. Report of two cases. J Korean
nonatherosclerotic and nondissection diseases of Med Sci 17:695–698
14. Terao C, Yoshifuji H, Mimori T (2014) Recent
carotid artery are exhaustive and have often advances in Takayasu arteritis. Int J Rheum Dis
nonspecific clinical and imaging findings. The 17:238–247
specific clinical inputs and typical imaging fea- 15. Hotchi M (1992) Pathological studies on Takayasu’s
tures are important clues to suspect these vascular arteritis. Heart Vessels Suppl 7:11–17
16. Hartlage GR, Palios J, Barron BJ, Stillman AE,
diseases and guide in further evaluation and man- Bossone E et al (2014) Multimodality imaging of
agement. Genetic and molecular analysis is aortitis. JACC Cardiovasc Imaging 7:605–619
confirmative in hereditary and connective tissue 17. Ohigashi H, Haraguchi G, Konishi M, Tezuka D,
disorders. Kamiishi T, Ishihara T et al (2012) Improved progno-
sis of Takayasu arteritis over the past decade – com-
prehensive analysis of 106 patients. Circ J
References 18. Alibaz-Oner F, Aydin SZ, Direskeneli H (2013)
Advances in the diagnosis, assessment and outcome
1. Touze E, Oppenheim C, Tyrstram D et al (2010) of Takayasu’s arteritis. Clin Rheumatol 32:541–546
Fibromuscular dysplasia of cervical and intracranial 19. Caballero PE (2011) Common carotid artery aneu-
arteries. Int J Stroke 5:296–305 rysm revealing Takayasu’s arteritis. J Stroke
2. Plouin PF, Perdu J, Batide-Alanore AL et al (2007) Cerebrovasc Dis 20(6):556–558
Fibromuscular dysplasia. Orphanet J Rare Dis 2:28 20. Santhosh K, Jaydadevan ER, Kapilamoorthy TR,
3. Schievink WI, Bjornsson J (1996) Fibromuscular Sylaja PN (2013) Temporal evolution of
dysplasia of the internal carotid artery: a clinicopath- hypertrophied vasa vasorum of common carotid
ological study. Clin Neuropathol 15:2–6 artery triggered by cerebral ischemia: a serial angio-
4. Mettinger KL (1982) Fibromuscular dysplasia and the graphic investigation. Neurol India 61(3):312–314
brain. II. Current concept of the disease. Stroke 21. Shuaib UA, Kate M, Homik J, Jerrakathil T (2013)
13:53–58 Recurrent non-aneurysmal subarachnoid
5. Schievink WI (2001) Spontaneous dissection of the haemorrhage in Takayasu arteritis: is the cause immu-
carotid and vertebral arteries. N Engl J Med nological or mechanical? BMJ Case Rep
344:898–906 22. Borchers AT, Gershwin MG (2012) Giant cell arteri-
6. Olin JW, Froehlich J, Gu X, Bacharach JM, Eagle K, tis: a review of classification, pathophysiology,
Gray BH, Jaff MR, Kim ESH, Mace P, Matsumoto geoepidemiology and treatment. Autoimmun Rev
AH, McBane RD, Kline-Rogers E, White CJ, 11:A544–A554
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery 159

23. Smeeth L, Cook C, Hall AJ (2006) Incidence of HIV-related vascular aneurysms. Br J Radiol
diagnosed polymyalgia rheumatica and temporal 75:884–888
arteritis in the United Kingdom, 1990–2001. Ann 38. Mazzoni P, Chiriboga CA, Millar WS, Rogers A
Rheum Dis 65:1093–1098 (2000) Immunodeficiency virus infection: case report
24. Hayreh SS, Podhajsky PA, Zimmerman B (1998) and literature review. Pediatr Neurol 23:252–255
Ocular manifestations of giant cell arteritis. Am J 39. Ritelli M, Drera B, Vicchio M, Puppini G, Biban P,
Ophthalmol 125:509–520 Pilati M, Prioli MA, Barlat S, Colombi M (2009)
25. Gonzalez-Gay MA, Vazquez-Rodriguez TR, Gomez- Arterial tortuosity syndrome in two Italian paediatric
Acebo I, Pego-Reigosa R, Lopez-Diaz MJ, Vazquez- patients. Orphanet J Rare Dis 4:20
Triñanes MC et al (2009) Strokes at time of disease 40. Callewaert BL, Willaert A, Kerstjens-Frederikse WS,
diagnosis in a series of 287 patients with biopsy- De Backer J, Devriendt K, Albrecht B et al (2008)
proven giant cell arteritis. Medicine (Baltimore) Arterial tortuosity syndrome: clinical and molecular
88:227–235 findings in 12 newly identified families. Hum Mutat
26. Nesher G (2014) The diagnosis and classification of 29:150–158
giant cell arteritis. J Autoimmun 48–49:73–75 41. Nauheim MR, Walcott BP, Nahed BV, Macrae CA,
27. Arida A, Kryprianou M, Kanakis M, Sfikakis PP Levinson JR, Oglivy CS (2011) Arterial tortuosity
(2010) The diagnostic value of ultrasonography- syndrome with multiple intracranial aneurysms.
derived edema of the temporal artery wall in giant Arch Neurol 68(3):369–371
cell arteritis: a second meta-analysis. BMC 42. Germain DP (2007) Ehlers- Danlos syndrome type
Musculoskelet Disord 11:44 4. Orphanet J Rare Dis 2:32
28. Schinkel AF, van den Oord SC, van der Steen AF, van 43. Gdynia HJ, Kuhnlein P, Ludolph AC, Huber R (2008)
Laar JA, Sijbrands EJ (2014) Utility of contrast- Connective tissue disorders in dissections of the
enhanced ultrasound for the assessment of the carotid carotid or vertebral arteries. J Clin Neurosci
artery wall in patients with Takayasu or giant cell 15:489–494
arteritis. Eur Heart J Cardiovasc Imaging 15:541–546 44. Boutouyrie P, Germanin DP, Fiessinger JN, Laloux P,
29. Bley TA, Reinhard M, Hauenstein C, Markl M, Perdu J, Laurent S (2004) Increased carotid wall stress
Warnatz K, Hetzel A, Uhl M, Vaith P, Langer M in vascular Ehlers-Danlos syndrome. Circulation
(2008) Comparison of duplex sonography and high- 109:1530–1535
resolution magnetic resonance imaging in the diagno- 45. Pepin M, Schwarze U, Superti-Furga A, Byers PH
sis of giant cell (temporal) arteritis. Arthritis Rheum (2000) Clinical and genetic features of Ehlers–Danlos
52:2470–2477 syndrome type IV, the vascular type. N Engl J Med
30. Blockmans D, Coudyzer W, Vanderschueren S, 342:673–680
Stroobants S, Loeckx D, Heye S, De Ceuninck L, 46. Oderich GS, Panneton JM, Bower TC, Lindor NM,
Marchal G, Bobbaers H (2008) Relationship between Cherry KJ, Noel AA, Kalra M, Sullivan T, Gloviczi P
fluorodeoxyglucose uptake in the large vessels and (2005) The spectrum, management and clinical out-
late aortic diameter in giant cell arteritis. Rheumatol- come of Ehlers-Danlos syndrome type IV: a 30-year
ogy 47:1179–1184 experience. J Vasc Surg 42:98–106
31. Robbs JV, Paruk N (2010) Management of HIV 47. Schievink WI (2004) Cerebrovascular involvement in
vasculopathy – a South African experience. Eur J Ehlers-Danlos Syndrome. Curr Treat Options
Vasc Endovasc Surg 39:S25–S31 Cardiovasc Med 6:231–236
32. Padayachy V, Robbs JV (2012) Carotid artery aneu- 48. Dohle C, Baehring JM (2012) Multiple strokes and
rysms in patients with human immunodeficiency bilateral carotid dissections: a fulminant case of newly
virus. J Vasc Surg 55:331–337 diagnosed Ehlers–Danlos syndrome type IV. J Neurol
33. Hamilton DK, Kassel NF, Jensen ME, Dumont AS Sci 318:168–170
(2007) Subarachnoid hemorrhage and diffuse 49. Horowitz MB, Purdy PD, Valentine RJ et al (2000)
vasculopathy in an adult infected with HIV. Remote vascular catastrophes after neurovascular
J Neurosurg 106:478–480 interventional therapy for type 4 Ehlers-Danlos syn-
34. Tipping B, de Villiers L, Wainwright H, Candy S, drome. AJNR Am J Neuroradiol 21:974–976
Bryer A (2007) Stroke in patients with human immu- 50. Usinskiene J, Mazighi M, Bisdorff A et al (2006)
nodeficiency virus infection. J Neurol Neurosurg Psy- Fatal peritoneal bleeding following embolization of
chiatry 78:1320–1324 a carotid-cavernous fistula in Ehlers-Danlos syn-
35. Maniker AH, Hunt D (1996) Cerebral aneurysm in the drome type IV. Cardiovasc Intervent Radiol
HIV patient: a report of six cases. Surg Neurol 29:1104–1106
46:49–54 51. Loeys BL, Schwarze U, Holm T et al (2006) Aneu-
36. Chetty R, Batitang S, Nair R (2000) Large artery rysm syndromes caused by mutations in the TGF-beta
vasculopathy in HIV-positive patients: another receptor. N Engl J Med 355(8):788–798
vasculitic enigma. Hum Pathol 31:374–379 52. Maccarrick G, Black JH III, Bowdin S, Hamamsy IE,
37. Woolgar JD, Ray R, Maharaj K, Robbs JV (2002) Frischmeier-Guerrerio PA, Guerrerio AL, Sponseller
Colour Doppler and grey scale ultrasound features of PD, Loeys B, Dietz HC (2014) Loeys–Dietz
160 S.K. Kannath and T.R. Kapilamoorthy

syndrome: a primer for diagnosis and management. 69. Okendi E, Moghaddam MB, Oderich GS (2010)
Genet Med 16:576–587 Internal carotid artery aneurysms in a patient with
53. Rahme RJ, Adel JG, Bendok BR, Bebawy JF, Gupta neurofibromatosis type 1. Vasc Endovascular Surg
DK, Batjer HH (2011) Association of intracranial 44:511–514
aneurysm and Loeys-Dietz Syndrome: case illustra- 70. Xu J, Ismat FA, Wang T et al (2007) NF1 regulates a
tion, management, and literature review. Neurosur- Ras-dependent vascular smooth muscle proliferative
gery 69:E488–E493 injury response. Circulation 116:2148–2156
54. Levitt MR, Morton RP, Mai JC, Ghodke B, Hallam 71. Ku YK, Chen HW, Chen HW, Fu CJ, Chin SC, Liu
DK (2012) Endovascular treatment of intracranial YC (2008) Giant extracranial aneurysms of both
aneurysms in Loeys-Dietz syndrome. J Neurointerv internal carotid arteries with aberrant jugular veins
Surg 4:e37 in a patient with neurofibromatosis type 1. Am J
55. van den Berg JSP, Hennekam RCM, Cruysberg JRM, Neuroradiol AJNR 29:1750–1752
Steijlen PM, Swart J, Tijmes N et al (2000) Prevalence 72. Reubi F (1945) Neurofibromatose et le´sions
of symptomatic intracranial aneurysm and ischaemic vasculaires. Schweiz Med Wochenschr 75:463–465
stroke in pseudoxanthoma elasticum. Cerebrovasc 73. Schievink WI, Riedinger M, Maya MM (2005) Fre-
Dis 10:315–319 quency of incidental intracranial aneurysms in neuro-
56. Schievink WI, Michels VV, Piepgras DG (1994) fibromatosis type 1. Am J Med Genet A 134:45–48
Neurovascular manifestations of heritable connective 74. You MW, Kim EJ, Choi WS (2011) Intracranial and
disorders: a review. Stroke 25:889–903 extracranial fusiform aneurysms in a patient with
57. Vasseur M, Carsin-Nichol B, Ebran JM, Willoteaux S, neurofibromatosis type 1: a case report.
Martin L, Leftheriotis G (2011) Carotid rete mirabile Neurointervention 6:34–37
and pseudoxanthoma elasticum: an accidental associ- 75. Rosser TL, Vezina G, Packer RJ (2005) Cerebrovas-
ation? Eur J Vasc Endovasc Surg 42:292–294 cular abnormalities in a population of children with
58. Dean JCS (2007) Marfan syndrome: clinical diagno- neurofibromatosis type 1. Neurology 64:553–555
sis and management. Eur J Hum Genet 15:724–733 76. Patro SN, Gupta AK, Aravinda HR, Jolapara MB,
59. Eberth JF, Taucer AI, Wilson E, Humphrey JD (2009) Saini J (2009) Combined transarterial and percutane-
Mechanics of carotid arteries in a mouse model of ous coiling of a spontaneous vertebrovertebral fistula
Marfan syndrome. Ann Biomed Eng 37:1093–1104 associated with neurofibromatosis Type 1.
60. Ferruzzi J, Collins MJ, Yeh AT, Humphrey JD (2011) J Neurosurg 111:37–40
Mechanical assessment of elastin integrity in fibrillin- 77. Deans WR, Bloch S, Leibrock L, Berman BM,
1-deficient carotid arteries: implications for Marfan Skultety FM (1982) Arteriovenous fistula in patients
syndrome. Cardiovasc Res 92:287–295 with neurofibromatosis. Radiology 144:103–107
61. Loeys BL, Dietz HC, Braverman AC et al (2010) 78. Gao P, Chen Y, Zhang H, Ling F (2013) Vertebral
Towards a revised Ghent nosology for the Marfan arteriovenous fistulae (AVF) in Neurofibromatosis
syndrome. J Med Genet 47:476–485 type 1: a report of two cases. Turk Neurosurg
62. Schievink WI, Parisi JE, Piepgras DG (1997) Intra- 23:289–293
cranial aneurysms in Marfan’s syndrome: an autopsy 79. Baccin CE, Krings T, Alvarez H, Ozanne A,
study. Neurosurgery 41:866–871 Lasjaunias PL (2007) A report of two cases with
63. Yuan SM, Jing H (2011) Cystic medial necrosis: dolichosegmental intracranial arteries as a new fea-
pathological findings and clinical implications. Rev ture of PHACES syndrome. Childs Nerv Syst
Bras Cir Cardiovasc 26:107–115 23:559–567
64. Bouziane Z, Bensaid B, Idrissi RE, Mesnaoui AE, 80. Heyer GL, Dowling MM, Licht DJ, Tay SKH,
Bensaid Y, Ratbi I (2013) Marfan syndrome and cer- Morel K, Garzon MC, Meyers P (2008) The cerebral
vical internal carotid artery aneurysm. Vasa vasculopathy of PHACES syndrome. Stroke
42:457–460 39:308–316
65. Sztajzel R, Hefft S, Girardet C (2001) Marfan’s syn- 81. Hess CP, Fullerton HJ, Metry DW, Drolet BA, Siegel
drome and multiple extracranial aneurysms. DH, Auguste KI, Gupta N, Haggstrom AN, Dowd CF,
Cerebrovasc Dis 11:346–349 Frieden IJ, Barkovich AJ (2010) Cervical and intra-
66. Matsuda M, Matsuda I, Handa H, Okamoto K (1979) cranial arterial anomalies in 70 patients with PHACE
Intracavernous giant aneurysm associated with syndrome. AJNR Am J Neuroradiol 31:1980–1986
Marfan’s syndrome. Surg Neurol 12:119–121 82. Siegel DH, Tefft KA, Kelly T, Johnson C, Mery D,
67. Riccardi VM (2000) The vasculopathy of NF1 and Burrows P, Pope E, Cordisco M, Holland KE,
histogenesis control genes. Clin Genet 58:345–347 Maheswari M, Keith P, Garzon M, Hess C, Frieden
68. Oderich GS, Sullivan TM, Bower TC, Glovickzi P, IJ, Fullerton HJ, Drolet BA (2012) Stroke in children
Miller DV, Babovic-Vuksanovic D, Macedo TA, with posterior fossa brain malformations, hemangi-
Stanson A (2007) Vascular abnormalities in patients omas, arterial anomalies, coarctation of the aorta and
with neurofibromatosis syndrome type I: clinical cardiac defects, and eye abnormalities (PHACE) syn-
spectrum, management, and results. J Vasc Surg drome: a systematic review of the literature. Stroke
46:475–484 43:1672–1674
9 Nonatherosclerotic Nondissection Diseases of Carotid Artery 161

83. Kirkwood ML, Chung J, Timaran CH, Valentine RJ 98. Langner S, Fleck S, Kirsch M, Petrik M, Hosten N
(2013) Extracranial carotid artery aneurysms in two of (2008) Whole-body ct trauma imaging with adapted
three monozygotic triplets with tuberous sclerosis and optimized CT angiography of the craniocervical
complex. J Vasc Surg 57:1120–1122 vessels: do we need an extra screening examination?
84. Jurkiewicz E, Jozwiak S (2006) Giant intracranial AJNR Am J Neuroradiol 29:1902–1907
aneurysm in a 9-year-old boy with tuberous sclerosis. 99. Fleck SK, Langner S, Baldauf J, Kirsch M,
Pediatr Radiol 36:463 Rosenstengel C, Schroeder HW (2010) Blunt
85. Hung PO, Wang HS, Chou ML, Wong AMC (2008) craniocervical artery injury in cervical spine lesions:
Tuberous sclerosis complex with multiple intracranial the value of CT angiography. Acta Neurochir
aneurysms in an infant. Pediatr Neurol 39:365–367 152:1679–1686
86. Sayed A, Elwan H, Fouad F, Taha A, Elhindawi K, 100. Ouffiah C, Hall E (2012) Imaging assessment of pen-
Khairy H, Gad A, Eldin HK (2010) Behcet extracra- etrating injury of the neck and face. Insights Imaging
nial carotid aneurysms: is there still a role for ligation. 3:419–431
Eur J Vasc Endovasc Surg 39:17–22 101. Chang FC, Lirng JF, Luo CB, Wang SJ, Wu HM, Guo
87. Albeyoglu S, Cinar B, Eren T, Filizcan U, WY, Teng MMH, Chnag CY (2008) Patients with
Bayserke O, Aslan C (2010) Extracranial carotid head and neck cancers and associated postirradiated
artery aneurysm due to Behcet’s disease. Asian carotid blowout syndrome: endovascular therapeutic
Cardiovasc Thorac Ann 18:574–576 methods and outcomes. J Vasc Surg 47:936–945
88. Bouarhroum A, Sedki N, Bouziane Z, Mahi OE, 102. Chang FC, Luo CB, Lirng JF, Lin CJ, Wu HM, Hung
Idrissi RE, Lahlou Z, Lekehel B, Sefiani Y, Mesnaoui SC, Guo WY, Teng MMH, Chang CY (2013) Evalu-
AE, Benjelloun A, Ammar F, Bensaïd Y (2006) ation of the outcomes of endovascular management
Extracranial carotid aneurysm in Behçet disease: for patients with head and neck cancers and associated
report of two new cases. J Vasc Surg 43:627–630 carotid blowout syndrome of the external carotid
89. Ozveren MF, Matsumoto Y, Kondo R, Takahashi A artery. Clin Radiol 68:e561–e569
(2009) Coil embolization of an unruptured intracra- 103. Chaloupka JC, Putman CM, Citardi MJ et al (1996)
nial aneurysm associated with Behcet’s disease. Endovascular therapy for the carotid blowout syn-
Neurol Med Chir (Tokyo) 49:471–473 drome in head and neck surgical patients: diagnostic
90. Kaku M, Hamada JI, Kuroda JI, Kai Y, Morioka M, and managerial considerations. AJNR Am J
Kuratsu JI (2007) Multiple peripheral middle cerebral Neuroradiol 17:843–852
aneurysms associated with Behcet’s disease. Acta 104. Lee CW, Yang CY, Chen YF, Huang A, Wang YH,
Neurochir (Wien) 149:823–827 Liu HM (2014) CT angiography findings in carotid
91. Garg K, Rockman CB, Lee V, Maldonaldo TS, blowout syndrome and its role as a predictor of 1-year
Jacobowitz GR, Adelman MA, Mussa FF (2012) survival. AJNR Am J Neuroradiol 35:562–567
Presentation and management of carotid artery aneu- 105. Luo CB, Teng MMH, Chang FC, Chang CY, Guo
rysms and pseudoaneurysms. J Vasc Surg WY (2008) Radiation carotid blowout syndrome in
55:1618–1622 nasopharyngeal carcinoma: angiographic features and
92. Longo GM, Kibbe MR (2005) Aneurysms of the endovascular management. Otolaryngol Head Neck
carotid artery. Semin Vasc Surg 18:178–183 Surg 138:86–91
93. Attigah N, Ku¨lkens S, Zausig N, Hansmann J, 106. Yanik B, Keyik B, Conkbayir I, Teber MA (2011)
Ringleb P, Hakimi M, Eckstein HH, Allenberg JR, Carotid blowout syndrome with oronasal hemor-
Bo¨ckler D (2009) Surgical therapy of extracranial rhage: magnetic resonance imaging findings. Jpn J
carotid artery aneurysms: long-term results over a Radiol 29:72–75
24-year period. Eur J Vasc Endovasc Surg 107. Headache Classification Committee of the Interna-
37:127–133 tional Headache Society (1988) Classification and
94. Nadig S, Barnwell S, Wax MK (2009) diagnostic criteria for headache disorders, cranial neu-
Pseudoaneurysm of the external carotid artery – ralgias and facial pain. Cephalalgia 8(Suppl 7):1–96
review of literature. Head Neck 31:136–139 108. Headache Classification Subcommittee of the Inter-
95. Cox MW, Whittaker DR, Martinez C, Fox CJ, national Headache Society (2004) The International
Feuerstein IM, Gillespie DL (2007) Traumatic Classification of Headache Disorders: 2nd edn.
pseudoaneurysms of the head and neck: early Cephalalgia 24(Suppl 1):9–160
endovascular intervention. J Vasc Surg 46:1227–1233 109. Kosaka N, Sagoh T, Uematsu H, Kimura H,
96. Maras D, Lioupis C, Magoufis G, Tsamopoulos N, Miyayama S, Noguchi M, Itoh H (2007) Imaging by
Moulakakis K, Andrikopoulos V (2006) Covered multiple modalities of patients with a carotidynia
stent-graft treatment of traumatic internal carotid syndrome. Eur Radiol 17:2430–2433
artery pseudoaneurysms: a review. Cardiovasc 110. Barker R, Massouh H (2009) Carotidynia – a valid
Intervent Radiol 29:958–968 cause of unilateral neck pain. Eur J Radiol Extra 71:
97. Welleweerd JC, Moll FL, de Borst GJ (2012) Techni- e97–e99
cal options for the treatment of extracranial carotid 111. Schaumberg J, Eckert B, Michels P (2011)
aneurysms. Expert Rev Cardiovasc Ther 10:925–931 Carotidynia. Magnetic resonance imaging and
162 S.K. Kannath and T.R. Kapilamoorthy

ultrasonographic imaging of a self-limiting disease. 120. Piagkou M, Anagnostopoulou S, Kouladouros K,

Clin Neuroradiol 21:91–94 Piagkos G (2009) Eagle’s syndrome: a review of the
112. Upton PD, Smith JRG, Charnock DR (2003) Histo- literature. Clin Anat 22:545–558
logic confirmation of carotidynia. Otolaryngol Head 121. Rodriguez-Vazquez JF, Merida-Velasco JR, Verdugo-
Neck Surg 129:443–444 Lopez S, Sanchez-Montesinos I, Merida-Velasco JA
113. Burton BS, Syms MJ, Petermann GW, Burgess LPA (2006) Morphogenesis of the second pharyngeal arch
(2000) MR imaging of patients with carotidynia. cartilage (Reichert’s cartilage) in human embryos.
AJNR Am J Neuroradiol 21:766–769 J Anat 208:179–189
114. Woo JKH, Jhamb A, Heran MKS, Hurley M, Graeb D 122. Lorman JG, Biggs JR (1983) The Eagle syndrome.
(2008) Resolution of existing intimal plaque in a patient AJR Am J Roentgenol 140:881–882
with carotidynia. AJNR Am J Neuroradiol 29:732–733 123. Ghosh LM, Dubey SP (1999) The syndrome of elon-
115. Commacho F, Bottin R, Brescia G, Tsilikas K, Volo T, gated styloid process. Auris Nasus Larynx
Tregnaghi A, Martini A (2012) Carotidynia: new 26:169–175
aspects of a controversial entity. Acta otorhino- 124. Dao A, Karnezis S, Lane JS III, Fujitani RM, Saremi
laryngol Ital 32:266–269 F (2011) Eagle syndrome presenting with external
116. Taniguchi Y, Horino T, Hashimoto K (2008) Is carotid artery pseudoaneurysm. Emerg Radiol
carotidynia syndrome a subset of vasculitis? 18:263–265
J Rheumatol 35:1901–1902 125. Muthusami P, Kesavadas C, Sylaja PN, Thomas B,
117. Amaravadi RR, Behr SC, Kousoubris PD, Raja S Harsha KJ, Kapilamoorthy TR (2013) Implicating the
(2008) [18F] Fluorodeoxyglucose positron-emission long styloid process in cervical carotid artery dissec-
tomography-CT imaging of carotidynia. AJNR Am J tion. Neuroradiology 55(7):861–867
Neuroradiol 29:1197–1199 126. Murtagh RD, Caracciolo RD, Fernandez G (2001) CT
118. Colby CC, Del Gaudio JM (2011) Stylohyoid com- findings associated with Eagle syndrome. AJNR Am J
plex syndrome-a new diagnostic classification. Arch Neuroradiol 22:1401–1402
Otolaryngol Head Neck Surg 137:248–252 127. Nakagawa D, Ota T, Iijima A, Saito N (2011) Diag-
119. Fusco DJ, Asteraki S, Spetzler RF (2012) Eagle’s nosis of eagle syndrome with 3-dimensional angiog-
syndrome: embryology, anatomy, and clinical man- raphy and near-infrared spectroscopy: case report.
agement. Acta Neurochir 154:1119–1126 Neurosurgery 68:E847–E849
Imaging of the Pathology of the
Vertebral Arteries 10
David Chiao and Max Wintermark

Contents Subclavian Steal Syndrome . . . . . . . . . . . . . . . . . . . . . . . . 180

Epidemiology and Clinical Findings . . . . . . . . . . . . . . . . 180
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 Pathophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Embryology and Anatomy . . . . . . . . . . . . . . . . . . . . . . . . . 164 Imaging Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Embryology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 Work-Up and Management . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Anatomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 Bow Hunter’s Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Vascular Territories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 Epidemiology and Clinical Findings . . . . . . . . . . . . . . . . 184
Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 Pathophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
Atherosclerotic Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 Imaging Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Epidemiology and Clinical Findings . . . . . . . . . . . . . . . . 172 Work-Up and Management . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Pathophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 Miscellaneous Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Imaging Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 Fibromuscular Dysplasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Work-Up and Management . . . . . . . . . . . . . . . . . . . . . . . . . . 174 Vasculitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Vertebral Artery Dissection . . . . . . . . . . . . . . . . . . . . . . . . 175 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Epidemiology and Clinical Findings . . . . . . . . . . . . . . . . 175
Pathophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
Imaging Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Work-Up and Management . . . . . . . . . . . . . . . . . . . . . . . . . . 178

D. Chiao • M. Wintermark (*)

Department of Radiology, School of Medicine, University
of Virginia, Charlottesville, VA, USA

# Springer Science+Business Media New York 2016 163

L. Saba, E. Raz (eds.), Neurovascular Imaging,
DOI 10.1007/978-1-4614-9029-6_17
164 D. Chiao and M. Wintermark

encountered vertebral artery diseases, vertebral

artery atherosclerosis and vertebral artery dissec-
The vertebral arteries are among the most clin-
tion. This is then followed by discussion of two
ically important arteries in the body, providing
functional disorders, subclavian steal syndrome
blood flow to the posterior circulation of the
and bow hunter’s syndrome. Finally, a brief over-
brain. Both radiologists and clinicians need to
view of miscellaneous conditions, such as
understand the basic anatomy and the pathol-
fibromuscular dysplasia and vasculitis, is covered
ogy that can affect the vertebral arteries in the
in the last section.
acute and non-acute settings. Radiologists, in
particular, need to be cognizant of the imaging
findings of vertebral artery pathology as
misdiagnosis of certain common conditions
Embryology and Anatomy
could be potentially disastrous. In this chapter,
the embryology and anatomy of the vertebral
arteries is reviewed, with a focus on normal
variants and vascular territories. Vertebral
Vertebral artery embryology forms a foundation
artery diseases are then discussed, including
for understanding normal and variant anatomy,
atherosclerotic disease, vertebral artery dissec-
which is useful in discriminating normal anatomic
tion, subclavian steal syndrome, and a variety
variation from pathologic processes. Recognition
of other conditions. Each section is organized
of normal variants is also important prior to sur-
by epidemiology and clinical findings, patho-
gical or endovascular intervention.
physiology, imaging findings, and work-up
The vertebral arteries develop from the dorsal
and management.
longitudinal neural arteries, which arise from lon-
gitudinal plexiform anastomoses of the first six
cervical intersegmental arteries (Fig. 1) [1]. The
Vertebral artery • Carotid-vertebrobasilar vari-
dorsal longitudinal arteries run parallel to the
ants • Vertebral artery fenestration • Basilar
primitive internal carotid arteries and appear by
artery fenestration • Posterior circulation • Ath-
day 30 of embryonic life. The cranial portions of
erosclerosis • Stroke • Vertebral artery stenosis
the dorsal longitudinal neural arteries gradually
• Vertebral artery dissection • Subclavian steal
appose each other, fusing in the midline around
syndrome • Bow hunter’s syndrome
day 35 to form the basilar artery [1]. As the basilar
artery matures, it begins supplying the posterior
circulation, which is previously supplied by the
Introduction primitive internal carotid arteries via the fetal pos-
terior communicating arteries.
Radiologists are frequently asked to evaluate the Prior to the development of the mature
course, caliber, and integrity of the vertebral arter- vertebrobasilar system, temporary anastomoses
ies. Successful evaluation of the vertebral arteries between the cranial portions of the dorsal longitu-
requires a thorough knowledge of vertebral artery dinal neural arteries and the primitive internal
anatomy and pathology. carotid arteries exist [1]. These carotid-
This chapter begins by reviewing the embryol- vertebrobasilar anastomoses, as well as the cervi-
ogy and anatomy of the normal vertebral arteries. cal intersegmental arteries, eventually regress
Normal variants are discussed in this section, as (with the exception of the seventh cervical
they are commonly seen in clinical practice. This intersegmental arteries which develop into the
is followed by a general discussion of the most subclavian arteries). If these anastomoses fail to
frequently employed imaging modalities used in regress, communication between the
evaluating the vertebral arteries. This is followed vertebrobasilar system and the internal carotid
by discussion of the two most commonly artery persists into adulthood. Four major
10 Imaging of the Pathology of the Vertebral Arteries 165

Fig. 1 Embryology of the Right internal carotid artery

vertebral arteries. The
vertebral arteries develop Right vertebral artery Right external carotid artery
from longitudinal plexiform
anastomoses of the first six Aorta
Right thyrocervical artery
cervical intersegmental
arteries. The seventh
cervical intersegmental
arteries develop into the
subclavian arteries

Cervical intersegmental arteries


carotid-vertebrobasilar variants have been neuralgia if there is significant mass effect on the
described in the literature and are named trigeminal nerve.
according to the cranial nerves that they parallel The persistent hypoglossal artery is the second
(Fig. 2): the persistent trigeminal artery, the per- most common carotid-vertebrobasilar variant,
sistent otic artery, the persistent hypoglossal present in 0.1–0.25 % of cerebral angiograms
artery, and the proatlantal intersegmental artery and autopsies [1, 3]. It arises from the distal cer-
[1–3]. In general, these variants are considered vical internal carotid artery (usually between the
incidental findings; however, an association with C1 and C3 vertebral bodies), passes through an
an increased incidence of aneurysms has been enlarged hypoglossal canal, and joins the proxi-
described [3]. mal basilar artery (Figs. 7 and 8). Patients are
The persistent trigeminal artery is the most typically asymptomatic but can uncommonly pre-
common carotid-vertebrobasilar variant, present sent with glossopharyngeal neuralgia and hypo-
in 0.1–1.0 % of cerebral angiograms and autopsies glossal paralysis if there is significant mass effect
[1, 3]. It arises from the cavernous internal carotid on the hypoglossal nerve.
artery and runs across the dorsum sellae to con- The persistent otic artery is exceedingly rare,
nect with the distal basilar artery (Fig. 3); the having only been described in a handful of case
basilar artery proximal to the aberrant connection reports [1, 3]. It arises from the petrous internal
is often hypoplastic. On sagittal imaging, the per- carotid artery, exits via the internal acoustic mea-
sistent trigeminal artery forms a characteristic tus, and joins the proximal basilar artery.
“tau” or “trident” sign (Fig. 4) [5]. The persistent A fourth variant, the proatlantal intersegmental
trigeminal artery can occasionally be mistaken for artery, is similar to the persistent hypoglossal
an aneurysm on cross-sectional imaging; 3D CTA artery in that it arises from the distal cervical
or DSA can be helpful in these difficult cases internal carotid artery and joins the proximal bas-
(Figs. 5 and 6). Patients are typically asymptom- ilar artery; however, instead of traveling through
atic but can uncommonly present with trigeminal the hypoglossal canal, the proatlantal
166 D. Chiao and M. Wintermark

Trigeminal artery

Otic artery


artery Fig. 3 MRA in a 45-year-old female demonstrating a
persistent trigeminal artery (arrow) which connects the
cavernous internal carotid artery with the distal basilar
Fig. 2 Carotid-vertebrobasilar variants. The persistent tri-
geminal artery is the most common variant, connecting the
cavernous internal carotid artery with the distal basilar common than vertebral artery fenestration, occur-
artery. The persistent otic artery connects the petrous inter-
nal carotid artery with the proximal basilar artery via the
ring in up to 5 % of cerebral angiograms and
internal acoustic meatus. The persistent hypoglossal artery autopsies, and also predisposes to aneurysms
connects the distal cervical internal carotid artery with the (Fig. 10) [1].
proximal basilar artery via the hypoglossal canal. The Duplication of the vertebral artery is very rare.
proatlantal intersegmental artery connects the distal cervi-
cal internal carotid artery with the proximal basilar artery
It is thought to be due to failure of fusion of the
via the foramen magnum cervical intersegmental arteries during
development [1].

intersegmental artery travels through the foramen

magnum [1, 5]. Anatomy
Vertebral artery fenestrations occur in 0.2–2.0 %
of cerebral angiograms and autopsies [1]. Extracra- The vertebral arteries usually arise from the prox-
nial fenestrations are thought to arise due to the imal subclavian arteries. However, in approxi-
persistence of the cervical intersegmental arteries mately 5 % of patients, the left vertebral artery
while intracranial fenestrations are thought to arise arises directly from the aortic arch (Fig. 11)
due to the persistence of basivertebral anastomoses. [5]. The vertebral arteries are typically asymmet-
On imaging, vertebral artery fenestrations are char- ric in size, most commonly with the left vertebral
acterized by a “gap” in the middle of a vertebral artery being dominant. When the vertebral arteries
artery; this can be confusing on axial imaging, with are approximately equal in size, they are consid-
visualization of two apparent lumens, but is nicely ered to be codominant. The normal luminal diam-
demonstrated on reformats (Fig. 9). Vertebral artery eter of the extracranial vertebral artery is
fenestrations are associated with other vascular approximately 3–5 mm.
anomalies and predispose to vertebral artery aneu- After arising from the subclavian arteries, the
rysms [1–3]. Basilar artery fenestration is more vertebral arteries ascend in the neck and are
10 Imaging of the Pathology of the Vertebral Arteries 167

Fig. 4 “Tau” sign (arrow)

on MRA sagittal reformat in
a patient with a persistent
trigeminal artery. The
persistent trigeminal artery
is the most common

foramen, and then exits superiorly through the

C2 transverse foramen to become the V3 segment.
The V2 segment resides inside the vertebral col-
umn (Fig. 13) and is vulnerable to vertebral frac-
tures, especially fractures which violate the
transverse foramina.

V3 Segment
The V3 segment courses posteriorly and medially
around the atlantooccipital joint (forming a
groove along the posterior ring of the atlas) and
then sharply turns anteriorly and superiorly to
pierce the dura to become the V4 segment. The
V3 segment resides outside the vertebral column
and is vulnerable to rotational forces which can
injure the artery at its dural insertion.
Fig. 5 3D carotid angiography nicely demonstrates a per-
sistent trigeminal artery (arrow) which was mistaken for an V4 Segment
aneurysm on prior MRA The V4 segment courses superiorly, anteriorly,
and medially through the foramen magnum and
categorized into four segments, abbreviated V1, then superomedially behind the clivus. The V4
V2, V3, and V4 (Fig. 12). segments unite to form the basilar artery near the
pontomedullary junction.
V1 Segment Along its course, the vertebral artery gives rise
The V1 segment is the first segment of the verte- to numerous arterial branches: cervical branches,
bral artery. The V1 segment typically arises from meningeal branches, and intracranial branches.
the subclavian artery, courses posteriorly and These arteries are described below.
superiorly in the superior mediastinum, and then
enters the C6 transverse foramen to become the Cervical Branches
V2 segment. The cervical branches originate from the extracra-
nial vertebral artery as it courses up the neck. The
V2 Segment cervical branches are categorized into the spinal
The V2 segment courses superiorly through the branches (which supply the spinal cord and verte-
C3–C6 transverse foramina, turns superolaterally bral bodies) and the muscular branches (which
through the “inverted L-shaped” C2 transverse supply the deep cervical musculature). The
168 D. Chiao and M. Wintermark

Fig. 6 Right carotid DSA

in the same patient as in
Fig. 5 with frontal (a) and
lateral (b) projections, again
demonstrating a persistent
trigeminal artery (arrow)

Intracranial Branches
The intracranial branches originate from the intra-
cranial vertebral artery (i.e., the V4 segment). The
intracranial branches are categorized into the ante-
rior spinal arteries, the posterior spinal arteries,
the posterior inferior cerebellar arteries, and the
perforating arteries. The anterior spinal arteries
unite in 50 % of cases and course inferiorly in
the anteromedian sulcus of the spinal cord to
supply the anterior spinal cord. The posterior
spinal arteries course inferiorly along the
posterolateral spinal cord to supply the posterior
spinal cord. The posterior inferior cerebellar
arteries course around the medullary tonsils to
supply portions of the medulla, inferior
cerebellum, and choroid plexus of the fourth ven-
tricle. The perforating arteries arise off the intra-
cranial vertebral artery to supply portions of the
medulla, inferior cerebellar peduncles, and
Fig. 7 CTA demonstrating a persistent hypoglossal artery olivary bodies.
(arrow) traversing through the hypoglossal canal
Variant anatomy of the posterior inferior cere-
muscular branches provide an important collateral bellar artery is common. Anomalous origin of the
pathway in the vertebrobasilar circulation, com- posterior inferior cerebellar arteries from V3
municating with the occipital branch of the exter- occurs in up to 10 % of cases. Duplication of the
nal carotid artery. posterior inferior cerebellar arteries is less com-
mon, occurring in approximately 2 % of cases
Meningeal Branches [4]. Occasionally, the posterior inferior cerebellar
The meningeal branches originate from the distal arteries arise from a single common trunk. Occa-
extracranial vertebral artery. The meningeal sionally, the vertebral artery terminates in the pos-
branches are categorized into the anterior menin- terior inferior cerebellar artery; in this situation,
geal artery and the posterior meningeal artery. The the contralateral vertebral artery supplies the
anterior meningeal artery arises from V2 and sup- majority of the posterior circulation blood flow
plies the dura around the foramen magnum. The [4]. Familiarity with these common variants helps
posterior meningeal artery arises from V3 and avoid confusion when interpreting the intracranial
supplies the falx cerebellum. branches of the vertebral arteries.
10 Imaging of the Pathology of the Vertebral Arteries 169

Fig. 8 Left carotid DSA in

the same patient as in Fig. 7
with frontal (a) and lateral
(b) projections
demonstrating a persistent
hypoglossal artery (arrow)

Fig. 9 (a) Axial time-of-

flight MRA demonstrates
two apparent lumens
(arrow) of the right
vertebral artery. (b) Coronal
time-of-flight MRA
projection reveals that this
represents a focal
fenestration (arrow) of the
V2 segment

Fig. 10 Right vertebral artery DSA (a) and 3D angiogram artery DSA after coil embolization demonstrates no signif-
(b) demonstrating a fenestrated proximal basilar artery icant filling of the aneurysm sac
with an associated saccular aneurysm. (c) Right vertebral

Vascular Territories insufficiency results in neurologic deficits that are

significantly different than those incurred by
The vertebrobasilar system supplies the posterior carotid artery insufficiency. Typical symptoms of
circulation of the brain. As such, vertebrobasilar vertebrobasilar insufficiency include bilateral
170 D. Chiao and M. Wintermark

motor and sensory deficits, nausea, vomiting, arteries, the posterior inferior cerebellar arteries,
vertigo, ataxia, diplopia, dysarthria, and dysmetria and the anterior inferior cerebellar arteries. The
[4, 5]. The precise deficits can be predicted based superior cerebellar arteries supply the superior
on the vascular territory of the arterial insult. surface of the cerebellum and the upper vermis.
The posterior inferior cerebellar arteries supply
Cerebellum the posterior and inferior surfaces of the cerebel-
The cerebellum is predominantly supplied by lum and inferior vermis. The anterior inferior cer-
three paired arteries: the superior cerebellar ebellar arteries supply the petrosal surface of the
cerebellum. The central portions of the vermis and
cerebellar hemispheres represent a watershed area
with variable contributions from all three vessels.

The brainstem is predominantly supplied by the
basilar artery and its perforating branches. How-
ever, the superior cerebellar arteries supply por-
tions of the superior pons while the distal vertebral
arteries and the posterior inferior cerebellar arter-
ies supply portions of the medulla. Occlusion of
one of the posterior inferior cerebellar arteries can
cause infarction of the lateral medulla, resulting in
Wallenberg syndrome. This syndrome is clini-
cally characterized by contralateral sensory defi-
cits affecting the body and extremities, as well as
ipsilateral sensory deficits affecting the face.


Fig. 11 Oblique sagittal reformat NECT demonstrating The vertebral arteries can be evaluated with a
the left vertebral artery arising directly from the aorta variety of imaging modalities, including ultra-
(arrow) sound, MRA, CTA, and angiography. Ultrasound

Fig. 12 3D vessel analysis

in a normal patient with
frontal (a) and oblique (b)
projections demonstrating
the four standard vertebral
artery segments, V1–V4
10 Imaging of the Pathology of the Vertebral Arteries 171

overcome many of these limitations but is still

prone to poor spatial resolution, patient motion,
and metal artifacts.
An alternative to MRA is CTA, which provides
excellent visualization of the vertebral arteries
with superior spatial resolution and shorter imag-
ing times. In addition, CTA is generally offered
day and night at most facilities. The scanning
range typically includes the aortic arch to the
circle of Willis in order to visualize the whole
length of the vertebral arteries. Automatic trigger-
ing or test bolus techniques can be used to time the
contrast injection. An adjunct non-contrast and
post-contrast head CT is often performed with
the CTA. One weakness of CTA is that it tends
to overestimate arterial stenosis in the setting of
heavy vessel calcification. In addition, the direc-
tion of arterial flow cannot be determined with
conventional CT imaging.
Angiography has historically been considered
the “gold standard” in vertebral artery imaging but
is now rarely performed as the initial imaging
examination. In the modern era, angiography is
Fig. 13 3D vessel analysis in a normal patient demon- typically reserved for resolving diagnostic
strating the relationship of V2 with the bony cervical
vertebral column
uncertainties, planning interventions, or
performing therapeutic procedures. Vertebral
artery angiography is typically performed with
with pulsed Doppler using a linear 5.0–7.5 MHz rapid sequences during subclavian artery injec-
probe is a relatively inexpensive examination but tions. Selective imaging can also be obtained,
is significantly limited by osseous structures, cal- usually in the AP, lateral views, and oblique
cified atherosclerosis, patient body habitus, and views. Angiography carries a small inherent risk
operator experience. Thus, cross-sectional imag- of vascular injury or distal embolization. The risk
ing is typically performed for full evaluation of of TIA and stroke is approximately 3 % and <1 %
the vertebral arteries. respectively with a cerebral angiogram
Two-dimensional time-of-flight MRA typi- examination [4].
cally provides satisfactory visualization of the The ideal imaging modality for the evaluation
vertebral arteries. A superior saturation band is of the vertebral arteries depends on the clinical
commonly applied to eliminate jugular venous context. In cases of suspected vertebral artery
flow. However, it should be noted that a superior occlusion or dissection, CTA of the head and
saturation band can potentially hide reversed ver- neck is a reasonable initial examination, providing
tebral artery flow. Thus, in cases of suspected rapid visualization of the vertebral arteries with
subclavian steal syndrome or bow hunter’s syn- excellent spatial resolution, thus potentially min-
drome, a phase contrast sequence can be imizing the time to revascularization. However, in
performed to unmask reversal of flow. The general cases of subclavian steal syndrome or bow
limitations of time-of-flight MRA include hunter’s syndrome, dynamic imaging with
overestimation of stenosis in the setting of vessel contrast-enhanced MRA may be preferred.
calcification, vessel tortuosity, or slow flow. Finally, in patients with tenuous renal function,
Dynamic contrast-enhanced MRA is able to time-of-flight MRA may be selected in order to
172 D. Chiao and M. Wintermark

avoid contrast agents and the risk of nephrogenic Pathophysiology

systemic fibrosis.
Atherosclerosis is a complex pathologic process
with endothelial injury playing a central role in its
Atherosclerotic Disease pathogenesis [8, 9]. Endothelial injury results in
increased permeability of the arterial wall to
Epidemiology and Clinical Findings lipids, which deposit in the intima resulting in
the fibrofatty plaque. Microscopically, the
The most common nontraumatic condition affect- fibrofatty plaque is composed of lipids, foam
ing the extracranial vertebral arteries is atheroscle- cells, and connective tissue. A fibrotic cap, com-
rotic disease [6]. Atherosclerosis is a systemic posed of smooth muscle and collagen, forms over
arterial disease, prevalent in industrialized the surface of the fibrofatty plaque. Disruption of
nations. It is responsible for the majority of ische- the fibrotic cap results in exposure of the
mic strokes, resulting in healthcare-associated thrombogenic fibrofatty plaque to the blood-
costs over $30 billion each year [7]. There are stream, which can result in acute thrombosis
multiple well-known risk factors for atherosclero- and/or thromboembolus [9]. In addition, as the
sis, both modifiable and nonmodifiable. Common fibrofatty plaque enlarges, it progressively nar-
nonmodifiable risk factors include family history, rows the arterial lumen. In general, a 50 %
age, and male gender. Common modifiable risk decrease in luminal diameter is required before
factors include age, hypertension, hyperlipidemia, hemodynamic consequences occur, as compensa-
diabetes, smoking, and renal failure (Table 1) [8]. tory physiologic vasodilation mitigates the effects
There is a strong association between vertebral of mild atherosclerotic disease [9]. It should be
artery atherosclerosis and carotid artery athero- noted that tandem stenoses can be additive in their
sclerosis; however, the vertebral arteries tend to hemodynamic effects, much like serial resistances
be less severely and less diffusely affected than in a circuit [8]. Ultimately, clinical symptoms
the carotid arteries [6]. occur when the arterial demand outstrips the arte-
Most patients with vertebral artery atheroscle- rial supply, resulting in a demand-supply
rosis are asymptomatic; however, if the degree of mismatch and end-organ ischemia. If the
vertebral artery stenosis is significant, patients can mismatch is transient and infarction has not
present with vertebrobasilar insufficiency which occurred, then the clinical manifestations are tran-
can manifest as vertigo, ataxia, imbalance, syn- sient and the episode is referred to as a transient
cope, and cranial nerve deficits [8]. Patients can ischemic attack. However, if the mismatch is
also present with a myriad of non-neurologic prolonged and infarction has occurred, then the
symptoms related to systemic atherosclerosis, clinical manifestations are prolonged and the epi-
including angina from coronary artery disease sode is referred to as a cerebrovascular accident or
and claudication from peripheral vascular disease. stroke. Patients with prior transient ischemic
attacks have a ten times increased risk of
stroke [9].
Table 1 Risk factors for atherosclerotic disease
Nonmodifiable Family history
Imaging Findings
Male gender
Vertebral artery stenosis severity is categorized
Modifiable Hypertension
per the North American Symptomatic Carotid
Endarterectomy Trial (NASCET) method
Smoking [10]. The NASCET method defines the percent
Renal failure stenosis as the ratio of the difference between the
normal and diseased luminal diameter to the
10 Imaging of the Pathology of the Vertebral Arteries 173

Table 2 Grading of atherosclerotic disease

Mild <50 %
Moderate 50–69 %
Severe 70 %

normal luminal diameter (i.e., percent stenosis =

(normal lumen minimal diseased lumen) / (nor-
mal lumen)  100%). Flow-limiting atheroscle-
rotic disease is graded by the percent stenosis,
with mild narrowing <50 %, moderate narrowing
50–70 %, and severe narrowing >70 % (Table 2).
Irregular atheromatous plaques are at an increased
risk of thromboembolism and are considered an
independent risk factor for stroke [4].
Ultrasound with pulsed Doppler has a very
limited role in evaluating vertebral artery athero-
Fig. 14 CTA demonstrating moderate stenosis of the left
sclerosis, largely due to its inability to visualize
vertebral artery origin (arrow) due to ostial calcified ath-
the vertebral artery in its entirety. As with other erosclerotic disease
blood vessels, arterial stenosis is characterized on
ultrasound by increased velocities, turbulence,
high-resistance waveforms, and elevated resistive
indices [6, 8, 11]. The post-stenotic segment is
characterized by decreased velocities, loss of
pulsatility, “parvus-tardus” waveforms, and
decreased resistive indices. Severe stenosis in
one vertebral artery can result in overestimation
in stenosis of the contralateral vertebral artery due
to compensatory increased blood flow [9].
CTA or MRA is often the initial diagnostic
examination for vertebral artery atherosclerosis,
with >90 % sensitivity and >90 % specificity
[5, 8, 9]. Both imaging modalities offer complete
visualization of the vertebral arteries and allow for
routine NASCET characterization of atheroscle-
rotic disease. In general, atherosclerosis is charac-
terized by smooth or irregular plaques resulting in
luminal narrowing, often with concomitant arte-
rial wall calcification [8]. Atherosclerotic disease
is typically most severe at the ostia of the vertebral
arteries (Fig. 14). Multiplanar reformats and/or Fig. 15 CTA demonstrating severe stenosis of the left
extracranial vertebral artery (arrow) due to atherosclerotic
3D vessel analysis can be useful in accurately disease
quantifying the degree of arterial stenosis, espe-
cially in vessel segments which do not run orthog- artery (Figs. 15 and 16 respectively). However,
onal to the axial plane. if the clinical suspicion for significant atheroscle-
Advanced atherosclerotic disease can manifest rosis is low (e.g., young patient without athero-
as near-occlusion or occlusion of the vertebral sclerotic risk factors), other causes of vertebral
174 D. Chiao and M. Wintermark

artery stenosis should be considered, such as ver- generally reserved for planning for endovascular
tebral artery dissection, radiation arteritis, interventions (Fig. 18).
fibromuscular dysplasia, or vasculitis [8]. In
these cases, MRA may provide additional infor-
mation by directly visualizing fibrofatty athero- Work-Up and Management
sclerotic plaque and excluding other causes of
luminal narrowing (Fig. 17). Lastly, the role of The management of vertebral artery atheroscle-
DSA in evaluating atherosclerotic disease is rotic disease has historically been with medical
therapy targeting atherosclerotic risk factors such
as hypertension or hypercholesterolemia. Sur-
gery, while frequently performed for carotid
artery atherosclerosis, is rarely performed for
vertebral artery atherosclerosis [12]. With the
advent of minimally invasive endovascular tech-
niques, angioplasty and stenting now have a role
in patients who fail maximum medical therapy.
Endovascular treatment involves aggressive pro-
cedural anticoagulation and deployment of a self-
expanding stent across the stenotic lesion. The
angiographic success rate of endovascular tech-
niques has been reported to be up to 95 %, with a
stroke rate of <2 % [9]. However, endovascular
stenting suffers from restenosis and neointimal
hyperplasia which often occurs at the margin of
previously placed stents (Fig. 19) [5]. Drug-
Fig. 16 CTA demonstrating complete occlusion of the eluting stents and other methods of reducing
right vertebral artery (arrow)

Fig. 17 Axial (a) and

coronal (b) MRA of the
neck demonstrating a
fibrofatty atherosclerotic
plaque (arrows) of the left
vertebral artery
10 Imaging of the Pathology of the Vertebral Arteries 175

Fig. 18 (a) Right vertebral

artery DSA demonstrates a
high-grade stenosis of V4.
(b) Right vertebral artery
DSA after angioplasty and
placement of a 2  8 mm
chromium cobalt stent
demonstrates an excellent
angiographic result

Fig. 19 (a) Right vertebral

artery DSA 2 years later in
the same patient as in
Fig. 18 demonstrates a >90
% stenosis at the distal
aspect of the previously
placed stent. (b) Right
vertebral artery DSA after
angioplasty and placement
of a second 2  8 mm
chromium cobalt stent
demonstrates an excellent
angiographic result

restenosis and neointimal hyperplasia are cur- fossa ischemic strokes [4, 13]. The mean age at
rently being developed and studied. presentation is approximately 45 years; there is a
slight male predominance [4, 14, 15]. Patients can
present with dizziness/vertigo, headache, neck
Vertebral Artery Dissection pain, ataxia, visual symptoms, nausea/vomiting,
nystagmus, Horner syndrome, sensory deficits,
Epidemiology and Clinical Findings cranial nerve palsies, dysphagia, and tinnitus
(Table 3) [13, 16]. In a meta-analysis of nearly
Vertebral artery dissection has an annual inci- 2,000 patients, Gottesman et al. found that the
dence of 1 per 100,000 patients and is one of the most common symptom was dizziness/vertigo
most common causes of stroke in young adult (58 %), followed by headache (51 %) and neck
patients [4]. Vertebral artery dissection constitutes pain (48 %) [16]. However, it should be noted that
approximately 27 % of all craniocervical arterial nearly one in four patients with proven vertebral
dissections and causes up to 40 % of all posterior artery dissection have no headache or neck pain at
176 D. Chiao and M. Wintermark

Table 3 Signs and symptoms of vertebral artery dissec- Table 5 Risk factors for vertebral artery dissection [16]
tion, sorted by frequency [16]
Dizziness/vertigo 58 % Hypertension
Headache 51 % Connective tissue disorders
Neck pain 46 % Fibromuscular dysplasia
Ataxia 38 % Vasculitis
Visual symptoms 36 % Autosomal dominant polycystic kidney disease
Nausea/vomiting 35 % Hyperhomocysteinemia
Nystagmus 29 % Osteogenesis imperfecta
Horner syndrome 22 %
Sensory deficits 21 %
Cranial nerve palsies 21 %
Major trauma (e.g., motor vehicle accident or
Dysphagia 13 %
Tinnitus 7%
penetrating injury) is the single most important
risk factor for vertebral artery dissection. How-
ever, even minor trauma (e.g., abrupt head turn-
ing, sports injury, and chiropractic injury) can
Table 4 Complications of vertebral artery dissection, result in vertebral artery dissection, being reported
sorted by location [16] in 15–70 % of cases [16]. Other risk factors
Stroke Any 60 % include hypertension, connective tissue disorders,
Extracranial 68 % fibromuscular dysplasia, and vasculitis (Table 5)
Intracranial 32 % [17]. When dissection occurs in the absence of
TIA Any 14 % trauma, it is referred to as spontaneous vertebral
Extracranial 21 % artery dissection. While rare, spontaneous verte-
Intracranial 11 %
bral artery dissection is a potentially devastating
SAH Any 35 %
event because delayed diagnosis can result in poor
Extracranial 0%
Intracranial 57 %


the time of diagnosis; thus, the absence of these Vertebral artery dissection results from a primary
symptoms do not exclude the diagnosis. tear of the intima, which allows blood to travel
The major complication of vertebral artery dis- from the lumen to the tunica media. The dissec-
section is cerebral infarction due to vertebral tion flap extends cranially (i.e., with the direction
artery occlusion; this occurs in approximately of blood flow) and forms a false lumen which can
60 % of patients [16]. The risk of stroke is highest compress and obstruct the true lumen resulting in
in the first few weeks after dissection but can cerebrovascular ischemia. The dissection flap can
occur up to a month after dissection [16]. Ischemic also extend into the adventitia resulting in
events are particularly common in extracranial pseudoaneurysm formation. In addition, vertebral
vertebral artery dissection with stroke occurring artery dissection can result in vascular stasis
in 68 % of patients with extracranial vertebral and become a nidus for distal thromboembolism
artery dissection compared to 32 % in intracranial [4, 18].
vertebral artery dissection [16]. In contrast, sub- The extracranial vertebral segments are more
arachnoid hemorrhage occurs almost exclusively prone to dissection than the intracranial segments
in intracranial vertebral artery dissection with an due to their greater mobility and proximity to
incidence of approximately 60 % in patients with adjacent bony structures (Table 6) [16, 19]. How-
intracranial vertebral artery dissection ever, the intracranial segments are more prone to
(Table 4) [16]. rupture due to their thinner adventitia [4].
10 Imaging of the Pathology of the Vertebral Arteries 177

Table 6 Vertebral artery dissection by vertebral artery Table 8 Outcomes after VAD by MRS [16]
segment [16]
MRS 0–1 67 %
V1 28 % MRS 2–4 18 %
V2 34 % MRS 5–6 10 %
V3 36 % MRS modified Rankin scale, MRS 0 no symptoms, MRS 1
V4 34 % no significant disability, MRS 2 slight disability, MRS 3
moderate disability, MRS 4 moderately severe disability,
MRS 5 severe disability, MRS 6 death

Table 7 Direct and indirect signs of vertebral artery dis-

section [19] Two patterns of vertebral artery dissection
Direct Double lumen 22 % have been described in the literature: the steno-
Intimal flap 21 % occlusive pattern and the dissecting aneurysmal
Indirect Arterial stenosis 51 % pattern [4, 20]. The steno-occlusive pattern is due
String of pearls 48 % to subintimal dissection resulting in narrowing of
Arterial dilatation 37 % the true lumen and is characterized by a tapered
Arterial occlusion 36 % stenosis or occlusion; long-segment stenosis has
Pseudoaneurysm 22 % been referred to as the “string” sign. This pattern is
typically seen in extracranial vertebral artery dis-
section and is associated with thromboembolic
phenomenon (Fig. 20) [20]. The dissecting aneu-
Most cases of vertebral artery dissection heal rysmal pattern is due to subadventitial extension
spontaneously with 90 % of patients having sig- of the dissection flap and is characterized by focal
nificant improvement in the first 2–3 months fusiform aneurysmal dilatation with or without
[4]. Vertebral artery dissection has an overall stenosis; alternating stenosis and dilatation has
recurrence rate of 15 %, with 50 % of cases been referred to as the “string-of-pearls” sign.
recurring in the first month [4]. This pattern is typically seen in intracranial verte-
bral artery dissection and is associated with sub-
arachnoid hemorrhage (Fig. 21) [4]. A very specific
Imaging Findings appearance of vertebral artery dissection is a nar-
row, eccentric lumen surrounded by crescent-
Given the variable and often nonspecific clinical shaped mural thickening [22, 23, 24, 27]. In rare
presentation of vertebral artery dissection, imag- cases, the lumen is preserved but is surrounded by
ing plays a primary role in diagnosis. Imaging hyperdense intramural hematoma, referred to as the
findings of vertebral artery dissection can be clas- “rind” sign [4, 21].
sified as direct or indirect (Table 7). Direct find- Conventional angiography has historically
ings include visualization of a double lumen or an been considered the “gold standard” in the diag-
intimal flap. Indirect findings include arterial ste- nosis of vertebral artery dissection, though recent
nosis, dilatation, alternating stenosis and dilata- studies have demonstrated that CTA has a near
tion (“string of pearls”), occlusion, or 100 % sensitivity and specificity when compared
pseudoaneurysm formation. Overall, the most to either conventional angiography or with the
common finding in vertebral artery dissection is final clinical diagnosis [21]. Today, CTA is typi-
vertebral artery stenosis, which is seen in approx- cally the initial diagnostic examination for the
imately half of patients [19]. While direct signs of evaluation of acute vertebral artery dissection.
vertebral artery dissection are more intuitive than MRA is comparable to CTA but takes longer to
indirect signs, they are seen in less than one in four perform and is generally indicated for the
patients with proven vertebral artery dissection non-emergent evaluation of vertebral artery dis-
[19] (Table 8). section [19]. One advantage of MRA over CTA is
178 D. Chiao and M. Wintermark

Fig. 20 Forty-three-year-old female with vertebral artery projections demonstrating abrupt narrowing and irregular-
dissection status post-motor vehicle collision. CTA of the ity of the right extracranial vertebral artery (arrows). Note
neck with axial (a) and coronal (b) maximum-intensity the adjacent transverse foramina fractures (open arrows)

Fig. 21 Forty-eight-year-
old female with
spontaneous intracranial
vertebral artery dissection
after awakening with severe
headache. (a) NECT
demonstrating diffuse
subarachnoid hemorrhage
and intraventricular blood.
(b) CTA demonstrating
occlusion of the right V4
segment (arrow); the left
vertebral artery is patent
(open arrow)

its superior ability of visualizing an intramural occlusion is seen in 20 % of cases, and an isolated
hematoma, which is characterized as an eccentric aneurysm is seen in 7 % of cases [4, 27].
region of high T1 signal intensity, due to methe-
moglobin (Fig. 22) [20]. Another advantage of
MRA is its greater sensitivity for acute infarcts, Work-Up and Management
the major complication of vertebral artery dissec-
tion (Fig. 23). Because patients with vertebral artery dissection
The imaging findings of vertebral artery dis- can present with nonspecific symptoms and often
section on angiography are similar to CTA/MRA. do not have significant risk factors (e.g., major
The characteristic finding is an abrupt caliber trauma or connective tissue disorders), a high
change (Fig. 24): an isolated stenosis is seen in level of clinical suspicion is needed for diagnosis.
46 % of cases, a stenosis with a concomitant One possible algorithm for the work-up of vertebral
aneurysm is seen in 27 % of cases, an isolated artery dissection is given below (Fig. 25) [19].
10 Imaging of the Pathology of the Vertebral Arteries 179

Fig. 22 Forty-two-year-
old male with spontaneous
vertebral artery dissection
after “thunderclap”
headache. (a) T1-WI with
fat saturation demonstrating
hyperintense intramural
hematoma (arrow). (b) 3D
contrast-enhanced MRA
demonstrating long-
segment stenosis (“string
sign”) of the left vertebral

Fig. 23 MRI/MRA of the

head in the same patient. (a)
Time-of-flight MRA
projection demonstrating an
abrupt cutoff of the left
P1-P2 junction (arrow).
DWI and ADC mapping (b,
c) demonstrating a left
posterior cerebral artery
territory infarct. Vertebral
artery dissection serves as a
nidus for thromboembolism

In this algorithm, patients with normal renal func- therapeutic intervention in extracranial dissec-
tion and no contrast allergy are evaluated with a tions [18]. In intracranial dissections, the risks of
CTA; patients with abnormal renal function or a anticoagulation are higher because of the
documented contrast allergy are evaluated with a increased incidence of subarachnoid hemorrhage.
time-of-flight MRA. Angiography is reserved for Endovascular or surgical intervention can be con-
troubleshooting in equivocal cases. sidered in selected patients.
Treatment of vertebral artery dissection is con- Follow-up imaging is often performed to eval-
troversial because randomized controlled studies uate for stability and/or resolution of vertebral
are lacking. Anticoagulation is the most frequent artery dissection (Fig. 26). On occasion, vertebral
180 D. Chiao and M. Wintermark

Fig. 24 Angiogram pre- (a) and post- (b) coil emboliza- representing thrombus formation. After coil embolization
tion of a right intracranial vertebral artery dissection (open of the right vertebral artery, the posterior circulation is
arrow) which involves the posterior inferior cerebellar supplied entirely by the left vertebral artery
artery. Note the intraluminal filling defect (arrow),

mainly in arteries not previously involved by the

dissection [18].
The clinical outcome of vertebral artery dissec-
tion is a function of the degree of cerebral ische-
mia caused by the dissection. Given that the
morbidity and mortality associated with vertebral
artery dissection is typically due to ischemic
infarcts, the modified Rankin scale (MRS) has
been used to categorize patient outcomes
(Table 6). Overall, patients have relatively good
outcomes after vertebral artery dissection, with
two-thirds of patients without significant perma-
nent disability [16].

Subclavian Steal Syndrome

Epidemiology and Clinical Findings

Fig. 25 An imaging work-up algorithm for vertebral
artery dissection The subclavian steal syndrome refers to the devel-
opment of neurologic symptoms secondary to
artery dissection can progress and/or convert into retrograde flow in a vertebral artery; it was first
pseudoaneurysms, necessitating endovascular or recognized by Contorni in 1960 and Reivich in
surgical intervention. In addition, follow-up imag- 1961 [29]. Retrograde flow without neurologic
ing is important for evaluating for recurrent ver- symptoms has been termed subclavian steal phe-
tebral artery dissection or delayed infarcts. Most nomenon. The Joint Study of Extracranial Arterial
recurrences occur within the first month and Occlusion reported that the incidence of subcla-
involve multiple cervical arteries [28]. After the vian steal phenomenon is approximately 2.5 %,
first month, the risk of recurrence is only about with only 5 % of patients with subclavian steal
1 % per year, and recurrent dissection occurs phenomenon presenting with neurologic
10 Imaging of the Pathology of the Vertebral Arteries 181

Fig. 26 CTA of the neck at

initial presentation (a) and
at 5-month follow-up (b)
after conservative
management of extracranial
vertebral artery dissection.
There is interval resolution
of the vertebral artery
stenosis (arrows)

Percentage of patients with neurologic symptoms

Fig. 27 Prevalence of 40%

neurologic symptoms in
patients with subclavian 35%
steal syndrome by brachial
blood pressure differential 30%






20-30 30-40 40-50 >50
Brachial blood pressure differential (mmHg)

symptoms [30]. The average age at presentation is cerebellum, and posterior cerebral hemispheres.
60 with a slight male predominance [4]. Because there is usually a proximal subclavian
Vertebrobasilar symptoms arise when there is artery stenosis resulting in the subclavian steal
inadequate blood flow to the posterior circulation phenomenon, patients can also present with ipsilat-
due to significant retrograde flow. Patients may eral ischemic arm symptoms such as arm claudica-
experience vertigo, ataxia, imbalance, and syn- tion, weakness, and paresthesias [4]. In fact, an
cope; in addition, cranial nerve deficits can occur important clinical marker for subclavian steal syn-
resulting in dysarthria, dysphagia, tinnitus, and drome is asymmetric brachial pressures [33, 34]. A
visual deficits [31–33]. Subclavian steal syndrome brachial blood pressure differential of greater than
rarely results in acute infarction because of the 20 mmHg is considered clinically significant, with
collateral blood supply from the contralateral ver- 55–85 % of cases demonstrating subclavian steal
tebral artery [4]. However, if the contralateral ver- phenomenon [33]. Increasing arm blood pressure
tebral artery is significantly diseased, patients are at differentials correlates with increasing incidence of
increased risk of infarction of the brainstem, subclavian steal syndrome (Fig. 27) [33].
182 D. Chiao and M. Wintermark

Pathophysiology Table 9 Grading of subclavian steal syndrome [33]

Grade I Reduced antegrade flow
As discussed above, the subclavian steal phenom- Grade II Alternating antegrade and retrograde flow
enon is typically due to a proximal, high-grade Grade III Permanent retrograde flow
subclavian artery stenosis or occlusion, resulting
in decreased distal subclavian artery pressures.
Vigorous exercise of the arm accentuates this Provocative maneuvers, such as ipsilateral arm
pressure gradient and provokes retrograde flow exercise or the hyperemia-ischemia cuff test,
from the ipsilateral vertebral artery, effectively may be employed to demonstrate blood flow
“stealing” blood flow from the posterior reversal.
circulation [4]. On Doppler ultrasound, the normal vertebral
The etiology of the inciting subclavian artery arteries demonstrate a low-resistance pattern with
stenosis is most commonly atherosclerosis cephalad flow throughout the cardiac cycle. In
[30]. In a study of over 7,000 neck ultrasounds, patients with subclavian steal syndrome, there is
Labropoulos et al. demonstrated that more than blunted antegrade or frank retrograde flow on the
80 % of cases of subclavian steal phenomenon affected side, often with compensatory increased
were due to atherosclerosis of the left subclavian flow (peak systolic velocity >60 cm/s) on the
artery [35]. This is thought to be due to the acute contralateral side [4, 36, 38, 39–42]. A grading
angle that the left subclavian artery takes as it system based on waveform characteristics has
arises off the aortic arch, predisposing to flow been developed, with grade I (mild steal) charac-
turbulence and atherosclerotic disease. Rarely, terized by reduced antegrade flow, grade II (mod-
subclavian steal syndrome can result from vascu- erate steal) characterized by alternating antegrade
litis, radiation angiopathy, aortic dissection, and retrograde flow, and grade III (severe steal)
extrinsic compression, or congenital stenosis [4]. characterized by permanent retrograde flow
A unique scenario can occur in patients with a (Table 9). The hyperemia-ischemia cuff test is
history of CABG with LIMA coronary bypass. In often used to provoke subclavian steal. The
these patients, the proximal subclavian artery ste- hyperemia-ischemia cuff test involves inflating a
nosis results in vascular steal not only from the blood pressure cuff at least 20 mmHg above the
ipsilateral vertebral artery but also from the LIMA systolic blood pressure for several minutes and
coronary bypass which can result in coronary then rapidly deflating the cuff with real-time
ischemia [36]. These patients may present with Doppler ultrasound of the ipsilateral vertebral
angina, especially after vigorous exercise of the artery. The rapid cuff deflation leads to transient
affected arm. increased blood flow to the arm, which can pro-
voke occult subclavian steal [36]. Infrequently,
the proximal subclavian artery stenosis can be
Imaging Findings visualized on Doppler ultrasound, characterized
by a high-resistance waveform, increased veloc-
Subclavian steal syndrome can be assessed with a ity, and turbulent flow.
variety of imaging modalities including ultra- While Doppler ultrasound is excellent at dem-
sound, CTA, MRA, and DSA. At some institu- onstrating the directional flow in the vertebral
tions, Doppler ultrasound is used as an initial arteries, it usually cannot visualize the proximal
screening test due to its availability and low cost subclavian artery stenosis. In these cases, cross-
[37]. CTA, MRA, and/or DSA are then performed sectional imaging with MRA or CTA can be
as confirmatory studies. The characteristic finding performed. The characteristic finding on cross-
is reversed flow in a vertebral artery with a prox- sectional imaging is a high-grade stenosis or
imal subclavian artery stenosis or occlusion. occlusion in the subclavian artery proximal to
10 Imaging of the Pathology of the Vertebral Arteries 183

Fig. 28 Forty-eight-year-
old female with history of
radiation therapy for
non-small cell lung cancer
who presents with
paroxysms of vertigo. (a)
CTA of the head and neck
demonstrating complete
occlusion of the proximal
subclavian artery (arrow).
(b) DSA with injection of
the left subclavian artery
showing the same finding

Fig. 29 Same patient as in

Fig. 28. (a) DSA with
injection of the right
vertebral artery. (b) Delayed
imaging demonstrates
retrograde filling of the left
vertebral artery (arrow)
with opacification of the left
subclavian artery (open
arrow), compatible with
subclavian steal

the takeoff of the vertebral artery (Figs. 28 and 29) “to-and-fro” flow in the affected vertebral artery,
[4]. Fibrofatty or calcified atherosclerotic plaque usually with a high-grade proximal subclavian
is a typical finding and may appear irregular artery stenosis (Fig. 32) [4].
and/or ulcerated. MRA is generally favored over
CTA because of its ability to visualize proximal
subclavian artery stenosis as well as its ability to Work-Up and Management
determine the direction of blood flow in the ver-
tebral arteries, especially with a phase contrast Only a small minority (approximately 18 %) of
sequence (Figs. 30 and 31). patients with subclavian steal phenomenon
As with most vascular pathology, angiography require therapeutic intervention [35]. Patients
is usually reserved for equivocal cases or prior to with mild symptoms are generally treated conser-
endovascular treatment. The characteristic finding vatively with medical treatment of the underlying
in subclavian steal syndrome is reversal of flow or atherosclerotic risk factors (such as
184 D. Chiao and M. Wintermark

Fig. 30 Fifty-five-year-old female with weakness and Fig. 31 Eighty-three-year-old male with lower extremity
paresthesias in her left hand after prolonged use. ataxia. 3D contrast-enhanced MRA demonstrates left sub-
Contrast-enhanced MRA maximum-intensity projection clavian artery occlusion (arrow)
demonstrates proximal left subclavian artery occlusion

Bow Hunter’s Syndrome

hyperlipidemia, hypertension, diabetes, or
smoking) [43]. Patients with severe symptoms Epidemiology and Clinical Findings
may require surgical or endovascular treatment
with carotid-subclavian bypass grafting or angio- Bow hunter’s syndrome, also known as positional
plasty/stenting. Surgical bypass has enjoyed wide or rotational occlusion of the vertebral artery, is a
success, with resolution of symptoms in the vast rare cause of vertebrobasilar insufficiency. The
majority of patients and patency rates of up to syndrome was originally described by Dr. Bruce
95 % at 10 years [44]. More recently, Sorensen in a case report in Neurosurgery in 1978
endovascular techniques have also been [48]. Sorensen described a 39-year-old man who
employed successfully, with patency rates of developed persistent right-sided weakness and
80 % at 5 years (Fig. 33) [45, 46]. Some authors left-sided numbness during archery practice.
argue that endovascular techniques are preferred Based on clinical findings, the patient was thought
over surgical techniques due to their similar effec- to have right medullary infarction resulting in
tiveness, lower complication rates, and shorter modified Wallenberg syndrome. However, the
hospitalizations; however, long-term patency is patient had no significant atherosclerotic risk fac-
less durable with endovascular therapy, with a tors. On angiography, the patient was found to
10-year target vessel revascularization rate of have significant right vertebral artery spasm.
47 % compared to 6 % for bypass [47]. Sorensen reasoned that the patient’s condition
10 Imaging of the Pathology of the Vertebral Arteries 185

Fig. 32 Sixty-year-old male with intermittent episodes of vertebral artery (arrow) with delayed opacification of the
diplopia and dizziness. (a) DSA with injection of the aortic right subclavian artery (open arrow). Right-sided subcla-
arch. The right subclavian artery is not seen. (b) Subse- vian steal syndrome is less common than left-sided subcla-
quent imaging demonstrates retrograde filling of the right vian steal syndrome

with their heads turned to the left (if right-

handed), the phenomenon of positional vertebral
artery occlusion is accentuated and could poten-
tially result in ischemia or infarction [48].
Since Sorensen’s original case report, at least
100 cases of bow hunter’s syndrome have been
reported in the literature [49]. The age at presen-
tation ranges widely, with most patients in their
40–60s [49, 50]. Patients typically present with
signs and symptoms of vertebrobasilar insuffi-
ciency, such as weakness, ataxia, numbness, tin-
gling, dysphagia, dysarthria, tinnitus, diplopia,
and nystagmus [4, 51]. Patients can also present
with nonspecific symptoms such as headache,
dizziness, nausea, and syncope [4, 52]. The key
finding is that clinical symptoms occur after a
sharp turn of the head.

Fig. 33 Same patient as in Fig. 32. DSA with injection of
the innominate artery after endovascular stent placement
(arrow) demonstrating antegrade flow in the right vertebral Bow hunter’s syndrome is thought to arise due to
(open arrow) and subclavian arteries positional occlusion of the contralateral vertebral
artery after sharply turning the head. In general,
stemmed from positional occlusion of the verte- transient positional vertebral artery occlusion is
bral artery during archery practice (i.e., turning considered physiologic and is asymptomatic due
the head sharply to the left will occlude the right to sufficient contralateral vertebral artery flow or
vertebral artery and vice versa). In archers and collateral flow via the posterior communicating
bow hunters, who stand perpendicular to a target arteries [50]. However, if the contralateral
186 D. Chiao and M. Wintermark

Fig. 34 DWI (a) and ADC

mapping (b) demonstrating
an area of restricted
diffusion in the right
thalamus, concerning for an
acute infarct

vertebral artery is hypoplastic or significantly dis- CTA can visualize areas of stenosis or occlusion.
eased, bow hunter’s syndrome is more likely to In addition, cross-sectional imaging has the
occur [50, 51]. Bow hunter’s syndrome is also advantage of evaluating extravascular pathology
more likely if the affected vertebral artery origi- which may explain or contribute to the patient’s
nates the dominant posterior inferior cerebellar symptoms. Complications of bow hunter’s syn-
artery [52]. drome, such as acute stroke, can also be visualized
Most cases of bow hunter’s syndrome occur at (Figs. 34 and 35).
the atlantoaxial level, which is especially prone to The “gold standard” for diagnosis remains
dynamic injury [4]. During neck rotation, the con- dynamic angiography with head turning
tralateral atlantoaxial joint rotates asymmetrically [52]. Angiography has the advantage of real-
forward and downward, potentially stretching the time visualization of stenosis or occlusion and
vertebral artery between the C1 and C2 foramina can demonstrate direction of flow in collateral
or under the atlantooccipital ligament [50]. Cervi- vessels (Fig. 36). Angiography is also frequently
cal spondylosis, herniated disks, and atlantoaxial performed to demonstrate angiographic improve-
instability can predispose to vertebral artery com- ment after an intervention is performed [53].
pression [50, 52]. In addition, cases of vertebral
artery compression by the longus colli and scalene
muscles have been reported [49]. Work-Up and Management

While bow hunter’s syndrome is rare, it can result

Imaging Findings in repeated vertebral artery ischemia and/or per-
manent neurologic deficits if not diagnosed and
Evaluation of bow hunter’s syndrome relies on treated appropriately. Work-up for bow hunter’s
dynamic imaging. Imaging is typically performed syndrome begins with a complete history and
in the neutral, left, and right positions. The key physical examination. In particular, a thorough
imaging finding is occlusion or stenosis of the description of the patient’s symptoms and exacer-
vertebral artery with the head turned to the left bating factors should be elicited, especially in
or right. Particular attention should be made to the patients with evidence of vertebrobasilar insuffi-
V3 segment of the vertebral artery (i.e., at the ciency. In particular, reproducibility of symptoms
atlantoaxial level), which is the most commonly with head turning is highly suggestive of bow
affected segment. hunter’s syndrome [49]. Occasionally, brainstem
Imaging is typically with MRA, CTA, or angi- auditory-evoked response testing is performed to
ography. Cross-sectional imaging with MRA or demonstrate increased latencies [51]. Dynamic
10 Imaging of the Pathology of the Vertebral Arteries 187

Fig. 35 CTA in the same

patient as in Fig. 34 with the
head turned to the left (a)
and with the head turned to
the right (b) demonstrating
positional narrowing of the
right vertebral artery
(arrow). Cervical
spondylosis, herniated
disks, and atlantoaxial
instability can predispose to
vertebral artery

Fig. 36 DSA of the right

vertebral artery in the same
patient as in Fig. 34 in the
neutral position (a) and with
head turned to the left (b)
demonstrating positional
occlusion of the vertebral
artery (arrow). The V3
segment is the most
commonly affected
segment in bow hunter’s

imaging studies can then be obtained to confirm patients, and surgical management, while effec-
the diagnosis. tive, can result in permanent restriction of neck
Patients with bow hunter’s syndrome are ini- rotation and can be complicated by a number of
tially managed with a cervical brace to restrict operative risks including infection, paralysis, and
neck rotation. Historically, medical treatment death [49]. Recently, endovascular therapy with
with anticoagulation was performed in mild self-expanding stents has been found to be an
cases and surgical decompression and/or cervical effective, minimally invasive treatment modality
spinal fusion was performed in severe cases. [52]. The downside to endovascular stenting is the
However, conservative management is effective potential for restenosis; data on long-term patency
in preventing stroke in only half of affected is unfortunately lacking. Nevertheless,
188 D. Chiao and M. Wintermark

endovascular therapy remains a treatment option Summary

and may supplant surgical management in
selected patients. The vertebral arteries are clinically important
blood vessels. Given the relatively high incidence
of vertebral artery pathology, thorough interroga-
tion of the vertebral arteries with diagnostic imag-
Miscellaneous Conditions ing is essential, with accurate interpretation
guiding management decisions. Collaboration
Fibromuscular Dysplasia among the radiology, neurology, and neurosur-
gery teams is of utmost importance. Clinical con-
Fibromuscular dysplasia is a noninflammatory text greatly aids in the selection of the most
condition, typically affecting middle-aged appropriate imaging modality. New and improved
women. It most commonly affects the renal, anatomic imaging is resulting in improved sensi-
carotid, or iliac arteries but can also affect the tivity and specificity in diagnosing vertebral artery
vertebral arteries in 10–40 % of cervical vascular diseases. In addition, dynamic imaging with or
cases. It is bilateral in two-thirds of patients. without provocative maneuvers is able to demon-
While there are several types of fibromuscular strate abnormal blood flow in functional vertebral
dysplasia, the most common type is artery disorders. Finally, interventional neuroradi-
medial fibroplasia characterized by a “string-of- ology is becoming increasingly prevalent in ther-
pearls” appearance. Fibromuscular dysplasia apeutic management, with a variety of minimally
predisposes to ischemia and infarct, likely due to invasive endovascular techniques being used in
the increased incidence of thromboembolism in the treatment of a number of vertebral artery
the affected arteries. It also predisposes to diseases.
spontaneous vertebral artery dissection. The
most sensitive examination for fibromuscular dys-
plasia is angiography, which is able to detect References
subtle cases. In symptomatic cases, treatment
with balloon angioplasty with or without stenting 1. Okahara M, Kiyosue H, Mori H, Tanoue S, Sainou M,
is warranted. Only rarely is surgical management Nagatomi H (2002) Anatomic variations of the cerebral
arteries and their embryology: a pictorial review. Eur
Radiol 12(10):2548–2561. doi:10.1007/s00330-001-
2. Dimmick SJ, Faulder KC (2009) Normal variants of
Vasculitis the cerebral circulation at multidetector CT angiogra-
phy. Radiographics 29(4):1027–1043. doi:10.1148/
Idiopathic vasculitis, including Takayasu’s arteri- 3. Luh GY, Dean BL, Tomsick TA, Wallace RC (1999)
tis and giant cell arteritis, can occasionally affect The persistent fetal carotid-vertebrobasilar anastomo-
the vertebral arteries. Vasculitis is characterized ses. AJR Am J Roentgenol 172(5):1427–1432.
by thickening of the arterial wall with long,
4. Osborn AG, Salzman KL, Barkovich AJ (2009) Diag-
smooth stenosis and/or “sausage-link” lesions. nostic imaging. Brain, 2nd edn. Amirsys, Salt Lake
Treatment is with steroids to reduce inflammatory City
injury. 5. Osborn AG, Jacobs JM, Osborn AG, Ralph Erskine
Conrad Memorial Fund (1999) Diagnostic cerebral
Radiation vasculitis can also affect the verte-
angiography, 2nd edn. Lippincott Willims & Wilkins,
bral arteries and can occur months to years Philadelphia, p 462
after radiation therapy. It is characterized by irreg- 6. Tay KY, U-King-Im JM, Trivedi RA et al (2005) Imag-
ular, calcified arteriopathy confined to the radia- ing the vertebral artery. Eur Radiol 15(7):1329–1343.
tion portal. Treatment is challenging, as irradiated