Escourolle and Poirier's Manual of

Basic Neuropathology 6th Edition

Francoise Gray Md (Editor)
Visit to download the full and correct content document:
https://ebookmass.com/product/escourolle-and-poiriers-manual-of-basic-neuropatholo
gy-6th-edition-francoise-gray-md-editor/
i

Escourolle & Poirier’s

Manual of Basic Neuropathology
iii

Escourolle & Poirier’s

MANUAL OF BASIC
NEUROPATHOLOGY
S I X T H E D I T IO N

EDI T E D B Y
F R A NÇ O I S E G RAY, MD , P H D
PROFES SOR OF PAT HOL OGY ( RE T.) , FACULT É DE M É D E C I N E PA R I S D I D E R O T, U N I V E R S I T É PA R I S ,
D IDER OT, S OR BONNE PARI S CI T É , PARI S, F RANCE
NEUROPATHOLOGI ST, DE PART M E NT OF PAT HOL OGY, LA R I B O I S I È R E H O S P I TA L ( R E T. ) , A P H P,
PARIS , FRANCE

C H A R L E S D U YC K A ERTS , MD , P H D
PROFES SOR OF PAT HOL OGY, UNI VE RSI T Y PI E RRE E T M A R I E C U R I E , PA R I S , FR A N C E
D IR EC TOR, NEUR OPAT HOL OGY L ABORAT ORY, PI T I É / S A LP Ê T R I È R E H O S P I TA L, PA R I S , FR A N C E

U M B E R T O D E G IRO LA MI, MD
PROFES SOR OF PAT HOL OGY, HARVARD M E DI CAL SC H O O L, B O S T O N , M A , U S A
NEUROPATHOLOGI ST, BRI GHAM AND W OM E N’S HO S P I TA L, B O S T O N , M A , U S A

1
1
Oxford University Press is a department of the University of Oxford. It furthers
the University’s objective of excellence in research, scholarship, and education
by publishing worldwide. Oxford is a registered trade mark of Oxford University
Press in the UK and certain other countries.

Published in the United States of America by Oxford University Press

198 Madison Avenue, New York, NY 10016, United States of America.

© Françoise Gray, Charles Duyckaerts, Umberto De Girolami 2019

All rights reserved. No part of this publication may be reproduced, stored in

a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by license, or under terms agreed with the appropriate reproduction
rights organization. Inquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above.

You must not circulate this work in any other form

and you must impose this same condition on any acquirer.

Library of Congress Cataloging-​in-​Publication Data

Names: Gray, Françoise, editor. | Duyckaerts, Charles, editor. | De Girolami, Umberto, editor.
Title: Escourolle & Poirier’s manual of basic neuropathology / edited by Françoise Gray,
Charles Duyckaerts, Umberto De Girolami.
Other titles: Manual of basic neuropathology
Description: Sixth edition. | New York, NY : Oxford University Press, [2019] |
Preceded by Escourolle & Poirier’s manual of basic neuropathology / [edited by] Françoise Gray,
Charles Duyckaerts, Umberto De Girolami. c2014. |
Includes bibliographical references and index.
Identifiers: LCCN 2018011890 | ISBN 9780190675011 (alk. paper)
Subjects: | MESH: Central Nervous System Diseases—pathology
Classification: LCC RC347 | NLM WL 301 | DDC 616.8/047—dc23
LC record available at https://lccn.loc.gov/2018011890

This material is not intended to be, and should not be considered, a substitute for medical or other professional advice. Treatment for the
conditions described in this material is highly dependent on the individual circumstances. And, while this material is designed to offer
accurate information with respect to the subject matter covered and to be current as of the time it was written, research and knowledge about
medical and health issues is constantly evolving and dose schedules for medications are being revised continually, with new side effects
recognized and accounted for regularly. Readers must therefore always check the product information and clinical procedures with the
most up-to-date published product information and data sheets provided by the manufacturers and the most recent codes of conduct and
safety regulation. The publisher and the authors make no representations or warranties to readers, express or implied, as to the accuracy or
completeness of this material. Without limiting the foregoing, the publisher and the authors make no representations or warranties as to the
accuracy or efficacy of the drug dosages mentioned in the material. The authors and the publisher do not accept, and expressly disclaim,
any responsibility for any liability, loss or risk that may be claimed or incurred as a consequence of the use and/or application of any of the
contents of this material.

9 8 7 6 5 4 3 2 1

Printed by Sheridan Books, Inc., United States of America

v

Contents

Foreword vii 4. Vascular Pathology 82

Martin A. Samuels
Preface to the Sixth Edition ix
Contributors xi
1. Basic Pathology of the Central Nervous
System 1
Pedro de Sá Cavalcante Ciarlini, Danielle Seilhean,
Umberto De Girolami, and Françoise Gray
2. Tumors of the Central Nervous
System 21
Keith L. Ligon, Sandro Santagata, and Franck Bielle
3. Central Nervous System Trauma 65
Colin Smith and R. Ross Reichard
Jean-​Jacques Hauw, Steven K. Feske,
Pierre Amarenco, and Umberto De Girolami
5. Infections of the Central Nervous
System 122
Fabrice Chrétien, Gregory Jouvion,
Kum Thong Wong, and Leroy R. Sharer
6. Human Prion Diseases 159
James W. Ironside
7. Multiple Sclerosis and Related
Inflammatory Demyelinating
Diseases 172
Hans Lassmann and Danielle Seilhean

• v
 8. Pathology of Degenerative Diseases of 12. Pathology of Skeletal Muscle 299
the Nervous System 186 Romain Gherardi, Anthony A. Amato,
Charles Duyckaerts Hart G. Lidov, and Umberto De Girolami

9. Acquired Diseases of the Nervous 13. Pathology of Peripheral Nerves 341

System: Metabolic, Toxic, and Related Jean-​Michel Vallat, Douglas C. Anthony, and
to Systemic Disease 219 Umberto De Girolami
Leila Chimelli and Françoise Gray
14. Diseases of the Pituitary Gland 373
10. Hereditary Metabolic Diseases 244 Vânia Nosé, Sandro Santagata,
Douglas C. Anthony and Hans H. Goebel and Edward R. Laws

11. Congenital Malformations and Appendix: Brief Survey of Neuropathological

Perinatal Diseases 275 Techniques 395
Homa Adle-​Biassette, Brian Harding, and David Meredith
Jeffrey A. Golden
Index 417

vi • C ontents
vi
Foreword
Most textbooks of medicine, even the most
venerated, undergo a new edition every 5 years
or even longer. This interval reflects the relatively
slow progress in most fields. After all, change in
the practice of medicine usually reflects the ar-
duous process of preclinical basic science followed
by the lengthy testing of drugs in animals and then
humans. As most clinical trials are negative, only
a small number of real changes occur in less than
a decade. However, this viscous process does not
hold in neuroscience. Driven mainly by advances
in molecular biology, cell biology, and genetics,
the clinical neurosciences (neurology, neurosur-
gery, and psychiatry) are experiencing dramatic
change. Together with cancer, the neurosciences are
undergoing a transformation that is analogous to
cardiology in the last century. Patients are now re-
ceiving the benefits of this momentum in the form
of treatments for virtually every category of nervous
system disorder. The modern neurology department
is now differentiated into 15–​20 divisions. This pro-
cess may be bemoaned by the older generation, but
is inexorable, reflecting as it does the increasingly
complex biology, pathophysiology, and therapeutics
of the various fields. This process, which occurred
in internal medicine in the twentieth century, is
now in full bloom in the twenty-​first century. For
this reason, a new edition of Escourolle and Poirier’s
Manual of Basic Neuropathology has been hatched in
only 5 years.
In generations past, the basic science of neuro-
science was neuropathology. Over time, other fields,
such as genetics, imaging, neuropharmacology, and
molecular biology have become vital to the training
and practice of the clinical neurosciences. But, con-
trary to becoming extinct, neuropathology has in-
corporated these newer disciplines and has, in effect,
reclaimed its leadership role in the understanding
and ultimately treatment of disorders of the nervous
system. In the sixth edition of the now classic
textbook, Escourolle and Poirier’s Manual of Basic
Neuropathology, Professors Umberto De Girolami,
Françoise Gray, and Charles Duyckaerts have
embodied this trend by creating what is, in effect,
• vii
a concise approach to the understanding of all the
major categories of nervous system disease. In a very
creative and truly unique manner, they have added
distinguished clinicians to many of the chapters to
ensure relevance of the basic information to the
actual practice of medicine. This reflects the fact
that many of the major questions in diseases of the
nervous system, such as regarding paralysis, pain,
disorders of consciousness, and cognitive decline,
do not fit neatly into any one of the old categories
that were artificially created in the past, as they no
longer reflect the relevant pathobiology.
Interdisciplinary neuroscience is the wave
of the future if we are to unlock the mysteries
of neurodegeneration, neuroinflammation, and
aging. In our own center, this is reflected in the re-
cent formation of a Program in Interdisciplinary
Neuroscience, which is meant to facilitate re-
search, clinical care, and education in areas that
fall between the classic disciplines, allowing for a
multiperspective approach to learning the causes and
creating preventive and curative strategies for the fu-
ture. As one salient example, the science underlying
neurodegeneration and repair has vital relevance to
a vast array of diseases, such as amyotrophic lateral
sclerosis, Alzheimer disease, and Parkinson disease.
In the past, these would have been considered en-
tirely separate illnesses, each cared for in different
divisions, such as neuromuscular diseases, cognitive
and behavioral neurology, and movement disorders.
It is now clear that the basic science of each of these
disorders is shared. Immunology has also taken a
major role in understanding diseases as apparently
disparate as stroke, dementia, and epilepsy.
Escourolle and Poirier’s Manual of Basic
Neuropathology is organized in its successful and
viii • F oreword
familiar manner, starting with a chapter on the basic
principles of the pathology of the nervous system.
This is followed by individual chapters on tumors,
trauma, vascular diseases, infections, prion diseases,
demyelinating processes, degenerative diseases, ac-
quired and inherited metabolic diseases, congenital
malformations, muscles, nerves, and diseases of the
pituitary gland. The book ends with an appendix
on commonly used neuropathology techniques. As
examples of the new approach, the distinguished
myologist Anthony Amato cowrote the chapter on
disease of skeletal muscle, as did the eminent neu-
rosurgeon Edward Laws for the pituitary chapter
and the venerated Steven Feske for the vascular
diseases chapter. As Gilbert and Sullivan’s Pooh-​Bah
famously said, this adds “artistic verisimilitude to an
otherwise bald and unconvincing narrative.”
Indeed, this new edition of Escourolle and Poirier’s
Manual of Basic Neuropathology is much more than
that. It is, in fact, a concise textbook of the modern
neurosciences, one that will be read and referred
to by the next generation of all those students and
physicians interested in diseases of the nervous
system.
Martin A. Samuels, MD, FANA, FAAN, FRCP,
MACP, DSc (hon)
Director, Program for Interdisciplinary
Neuroscience
Chair, Department of Neurology
Brigham and Women’s Hospital
Miriam Sydney Joseph Professor of Neurology
Harvard Medical School
Boston, Massachusetts
USA
ix
Preface to the Sixth Edition
The first two French editions of the Manuel de
Neuropathologie, published in 1971 and 1977,
were conceived, written, and edited by Raymond
Escourolle and Jacques Poirier. After the death
of Raymond Escourolle in 1984, Françoise Gray
joined Jacques Poirier for the third edition; in ad-
dition, Jean-​Jacques Hauw and Romain Gherardi
contributed to selected chapters. The first three
editions reached the English-​speaking public thanks
to the friendship and translating ability of the now-​
deceased Lucien Rubinstein. For the fourth edition,
Umberto De Girolami joined as co-​editor, and the
scope of the monograph was expanded with the col-
laborative efforts of multiple experts throughout the
world to write the English language text. Jacques
Poirier retired, and Charles Duyckaerts then agreed
to join the editorial team for the fifth and now the
sixth edition. For this sixth edition, there have again
been changes in authorship of several chapters in
response to the changing status of senior authors
and the need to recruit active investigators to
replace them.
This sixth edition of the manual attempts
to maintain the general intention of Professors
Escourolle and Poirier for the first and subsequent
editions of the monograph, that is, to provide, for
the reader resident-​trainee and staff member, a basic
description of the lesions underlying the diseases of
the nervous system and to limit pathophysiological
considerations to essential principles. Historical,
clinical neurologic and radiologic imaging data,
once again, have been deliberately excluded, as well
as bibliographic reference listings. This important
body of information essential for the practice of
neuropathology can be obtained in the many com-
prehensive treatises on the subject now readily avail-
able. We also have made the assumption that the
reader has some familiarity with general concepts of
neuroanatomy, neurohistology, and the principles of
anatomic pathology as well as clinical neurology.
With these guidelines in mind, our aim has been
to produce a text that mainly presents those labo-
ratory aspects of neuropathology that are morpho-
logical (and molecular, where appropriate) and to
• ix
demonstrate these with accurate descriptions and
good illustrations, all within the scope of a concise
and inexpensive “manual.”
For several specific reasons, we think that the
time is now right for a new edition since the last
one in 2014. Over the past several years, specialty
training in neurology, neurosurgery, and pathology
has very much changed throughout much of the
world, such that in these disciplines, less time is
being devoted to neuropathology. This has been due
in large part to the tremendous expansion of know-
ledge in allied subspecialty areas, requiring that
more time be devoted to them. Thus, the trainee in
these disciplines is now very much in need of a con-
cise introductory text.
In addition, several other important changes in
medicine and society have had an impact on the field
of neuropathology and demand being addressed in
this text.
• For a variety of social and scientific reasons,
autopsy studies are currently being performed
much less frequently than in years past. This
change has been brought about in part because
the progress in radiologic imaging, both
structural and functional, has decreased the
need to draw on clinico-​anatomic correlations
derived from postmortem data to guide clinical
practice. Oddly enough, conversely, autopsy-​
derived knowledge of the anatomic distribution
and the neuropathologic basis of lesions
continues to be a valuable body of information
for the interpretation of imaging data. To
support this aim, we have amply made use of
macroscopic illustrations and whole-​brain
celloidin/​paraffin-​embedded sections from our
archives.
• Progress in molecular biology and genetics has
revolutionized the laboratory diagnosis of many
groups of neurological diseases. Neuropathology
stands at the vanguard of the development and
implementation of these diagnostic studies. In the
last decade, progress in immunohistochemistry
methods for in situ identification of abnormal
proteins and the enormous advances in molecular
biology to uncover specific gene mutations have
led to greater understanding of many hereditary
neurologic diseases, including degenerative and
x • preface
metabolic disorders, developmental disorders,
and neuromuscular diseases. Morphologic
neuropathologic data, obtained at biopsy or
at postmortem examination, have therefore
needed to be integrated with this new knowledge
for the reinterpretation and reclassification of
many diseases. For example, neuropathologic
information obtained at biopsy, combined
with molecular biology and genetic data, is
now required for the diagnosis, prognosis, and
guidance of the choice of treatment modalities in
brain tumors.
• Last, an urgent responsibility to present an
updated synopsis of neuropathology is that this
knowledge is important to allied disciplines, as
there is constant need for surveillance of newly
emerging diseases, including infectious and
iatrogenic ones.
We need to thank first of all Susan Pioli, who al-
though now retired from the publishing business,
was instrumental in prior editions and led us to Craig
Panner and his colleagues with Oxford University
Press, who has given fundamental support. Second,
we thank present and past contributing authors and
their staff for the text and illustrations provided in
this new edition.
In the Introduction to the first edition, Professors
Escourolle and Poirier offered an apology to the
reader that is still valid some half a century later:
The compilation of a basic work designed to
familiarize physicians-​in-​training with such a
highly specialized discipline as Neuropathology
entails two opposing risks: in attempting to
compress the maximum amount of information
within the minimum space, the text is liable to
become unintelligible to beginners; if on the con-
trary, one tries to maintain too elementary a
level, the danger is that only the obvious will be
stated. In presenting to the non-​initiated reader
neuropathological information that some may
find too simple, we have preferred the hazard of
the second pitfall.
Françoise Gray
Charles Duyckaerts
Umberto De Girolami
xi
Contributors
Homa Adle-​Biassette, MD, PhD
Professor of Pathology, Faculté de médecine Paris
Diderot, Université Paris, Diderot, Sorbonne
Paris Cité, Paris, France
Pathologist, Neuropathologist, Fetal (neuro)
pathologist, Department of Pathology,
Lariboisière Hospital, APHP, Paris, France
Pierre Amarenco, MD
Professor and Chairman, University of Paris
Diderot, Sorbonne Paris Cité, Paris, France
Department of Neurology and Stroke Centre
Bichat Hospital, Paris, France
Anthony A. Amato, MD
Professor of Neurology, Harvard Medical School,
Boston, MA, USA
Vice-​Chairman, Department of Neurology and
Chief, Neuromuscular Division, Brigham and
Women’s Hospital, Boston, MA, USA
Douglas C. Anthony, MD, PhD
Professor of Pathology and Laboratory Medicine,
and Professor of Neurology, Alpert Medical
School of Brown University, Providence,
RI, USA
Pathologist-​in-​Chief, Lifespan Academic Medical
Center, Providence, RI, USA
Franck Bielle, MD, PhD
Assistant Professor of Pathology, University Pierre
et Marie Curie, Paris, France
Neuropathologist, Pitié/​Salpêtrière Hospital,
Paris, France
Leila Chimelli, MD, PhD
Professor of Pathology, Federal University of Rio de
Janeiro, Rio de Janeiro, Brazil
Neuropathologist, Laboratory of Neuropathology,
State Institute of Brain, Rio de Janeiro,
Brazil
• xi
Fabrice Chrétien, MD, PhD
Professor of Histology, Université Paris Descartes,
Paris, France
Neuropathologist, Laboratoire de Neuropathologie,
Centre Hospitalier Sainte Anne, Paris, France
Unité de Neuropathologie Expérimentale,
Institut Pasteur, Paris, France
Pedro de Sá Cavalcante Ciarlini, MD
Auxiliary Professor of Pathology, Federal University
of Ceará, Sobral, CE, Brazil
Pathologist/​Neuropathologist, Laboratório
Clementino Fraga, Fortaleza, CE, Brazil
Umberto De Girolami, MD
Professor of Pathology, Harvard Medical School,
Boston, MA, USA
Neuropathologist, Brigham and Women’s Hospital,
Boston, MA, USA
Charles Duyckaerts, MD, PhD
Professor of Pathology, University Pierre et Marie
Curie, Paris, France
Director, Neuropathology Laboratory,
Pitié/​Salpêtrière Hospital, Paris, France
Steven K. Feske, MD
Associate Professor of Neurology, Harvard Medical
School, Boston, MA, USA
Chief, Cerebrovascular Division, Department of
Neurology, Brigham and Women’s Hospital,
Boston, MA, USA
Romain Gherardi, MD
Professor of Histology, Paris-​Est University,
Créteil, France
Neuropathologist, Henri Mondor University
Hospital, Créteil, France
Hans H. Goebel, MD
Professor of Neuropathology (ret.),
Johannes Gutenberg University,
Mainz, Germany
Consultant Neuropathologist, Institute
of Neuropathology, Charité, Berlin,
Germany
Jeffrey A. Golden, MD
Ramzi S. Cotran Professor of Pathology, Harvard
Medical School, Boston, MA, USA
Chair of Pathology and Neuropathologist, Brigham
and Women’s Hospital, Boston, MA, USA
xii • C ontributors
Françoise Gray, MD, PhD
Professor of Pathology (ret.), Faculté de médecine
Paris Diderot, Université Paris, Diderot,
Sorbonne Paris Cité, Paris, France
Neuropathologist, Department of Pathology,
Lariboisière Hospital (ret.), APHP, Paris, France
Brian Harding, MA, DPhil, BMBCh, FRCPath
Professor of Pathology and Laboratory Medicine,
University of Pennsylvania, Philadelphia, PA, USA
Neuropathologist, Children’s Hospital of
Philadelphia, Philadelphia, PA, USA
Jean-​Jacques Hauw, MD
Professor of Pathology (ret.), Pierre et Marie Curie
University, Paris, France
Neuropathologist (ret.), Pitié-​Salpêtrière Hospital,
Paris, France
Académie Nationale de Médecine, Paris, France
James W. Ironside, MD
Professor of Clinical Neuropathology (ret.), School
of Clinical Sciences, University of Edinburgh,
Edinburgh, UK
Gregory Jouvion, PhD
Unité de Neuropathologie Expérimentale, Institut
Pasteur, Paris, France
Hans Lassmann, MD, PhD
Professor of Neuroimmunology, Medical
University of Vienna, Vienna, Austria
Chairman, Department of Neuroimmunology,
Center for Brain Research, Vienna, Austria
Edward R. Laws, Jr., MD, FACS
Professor of Neurosurgery, Harvard Medical
School, Boston, MA, USA
Neurosurgeon, Brigham and Women’s Hospital,
Boston, MA, USA
Hart G. Lidov, MD, PhD
Associate Professor of Pathology, Harvard Medical
School, Boston, MA, USA
Neuropathologist, Boston Children’s Hospital,
Boston, MA, USA
Keith L. Ligon, MD, PhD
Associate Professor of Pathology, Harvard Medical
School, Boston, MA, USA
Neuropathologist, Dana-​Farber Cancer Institute/​
Brigham and Women’s Hospital, Boston,
MA, USA
xi

David Meredith, MD, PhD Leroy R. Sharer, MD

Instructor of Pathology, Harvard Medical School, Professor of Pathology, Rutgers New Jersey
Boston, MA, USA Medical School, Newark,
Neuropathologist, Brigham and Women’s Hospital, NJ, USA
Boston, MA, USA Neuropathologist, University Hospital,
Newark, NJ, USA
Vânia Nosé, MD, PhD
Professor of Pathology, Harvard Medical School, Colin Smith, MD, FRCPath
Boston, MA, USA Professor of Neuropathology, University of
Associate Chief of Pathology for Anatomic and Edinburgh, Edinburgh, UK
Molecular Pathology, Massachusetts General Consultant Neuropathologist, Western General
Hospital, Boston, MA, USA Hospital, Edinburgh, UK
R. Ross Reichard, MD Jean-​Michel Vallat, MD
Associate Professor of Pathology, The Mayo Clinic Professor of Neurology, Faculté de
School of Medicine, Rochester, MN, USA Médecine, Limoges University,
Neuropathologist, Mayo Clinic, Rochester, Limoges, France
MN, USA Neurologist/​Neuropathologist, University
Hospital, Limoges, France
Sandro Santagata, MD, PhD
Assistant Professor of Pathology, Harvard Medical Kum Thong Wong, MBBS, MPath,
School, Boston, MA, USA FRCPath, MD
Neuropathologist, Brigham and Women’s Hospital, Professor, University of Malaya, Kuala Lumpur,
Boston, MA, USA Malaysia
Senior Consultant, University of Malaya
Danielle Seilhean, MD, PhD
Medical Centre, Kuala Lumpur, Malaysia
Professor of Pathology, University Pierre et Marie
Curie, Paris, France
Neuropathologist, Pitié/​Salpêtrière Hospital,
Paris, France

C ontributors • xiii
1
1
Basic Pathology of the Central
Nervous System
PE D RO D E SÁ CAVA LC A N T E C IA R L IN I, D A NIEL L E SEIL HEAN , UMBERTO DE GIROL AMI,
AN D F RA N ÇO I SE G R AY
AUTOPSY DIAGNOSIS in neuropathology
is based on the macroscopic and microscopic study
of the brain, brainstem, cerebellum, and spinal cord.
Increasingly, the ability to reach greater diagnostic pre-
cision is buttressed by the new laboratory techniques of
molecular biology and genetics. Three consecutive steps
are involved in reaching a diagnosis, and these are, in fact,
closely interrelated: (1) a morphological/​ laboratory
analysis of the lesions; (2) a topographical analysis of
the lesions; and (3) a critical integration of these findings
and their subsequent correlation with the clinical data
and the general autopsy findings, thus permitting an eti-
ological diagnosis to be made in most instances.
1 . 1 M OR P HOL OGICAL
AN ALY S IS OF CE N T RAL
N ER V OUS SYST E M L E SION S
A number of disorders of the central nervous system
(CNS) are characterized morphologically by the
co-​expression of multiple reactions to injury that may
not necessarily be diagnostic in themselves. These
reactions affect the cellular elements of the nervous
system (neurons, astrocytes, oligodendrocytes, and
microglia) and/or the supporting structures (me-
ninges, connective tissue, or blood vessels). Basic
cellular reactions are demonstrable on microscopic
examination, whereas tissue lesions associated with
more extensive destructive or atrophic changes are
recognized to the naked eye or with the help of a
magnifying lens.
Although, for didactic purposes, the reactions to
injury seen in the neurons, glia, connective tissue,
and vascular structures are described separately in
this chapter, we would emphasize that there is a close
functional interdependence of the various cellular
elements of the nervous system. This is particularly
important in the case of nerve cell alterations where,
except for very acute injury, the possibility of artifac-
tual change should be entertained whenever the reac-
tion is not accompanied by a glial cell response.
• 1
1 .1.1 Basi c c e l l u l a r
r e acti ons to C NS i n j u ry
1.1 . 1 . 1 . N E UR ON A L L ESI O NS
Neuronal injury may be sufficiently severe to result in
irreversible damage (cell death) or may be transient,
or minimal, and cause reversible functional damage.
Destruction of neurons may be focal or extend dif-
fusely, involving many populations of neurons
throughout the nervous system. In acute neuronal
injury, when the tissue is examined stained with he-
matoxylin and eosin (H&E) at a relatively short time
after a lethal insult to the cell (12–24 hours or some-
what longer), one observes eosinophilia of the cyto-
plasm, shrinkage and hyperchromasia of the nucleus,
and disappearance of the nucleolus; subsequent to
the disintegration of the cell, neuronophagia by scav-
enger cells is evident at a later time interval (days
later). In chronic diseases, evidence of cell death is
recognized morphologically as neuronal “cell loss”
or alternately (as “atrophy”) on macroscopic exam-
ination when the irreversible injury has occurred
relatively slowly (months or years) and has progres-
sively involved increasing numbers of cells. In some
degenerative diseases of the nervous system where
there is progressive loss or damage of neurons over
variable time periods, the affected cells have distinc-
tive morphologic hallmarks (e.g., neurofibrillary de-
generation, neuronal storage of metabolic products,
disorders associated with intracellular inclusions).
Nerve cell loss (i.e., reduction in the number of
cell bodies in a particular brain region as compared
to normal) when it involves less than 30% of the
normal cell population, may be difficult to ascertain
in the absence of rigorous morphometric analysis.
Furthermore, precise assessment depends on con-
sideration of the thickness of the section and on the
normal cytoarchitectonics of the region examined.
1.1.1.1.1. Nerve cell atrophy. Neuronal at-
rophy is the descriptive term that is given to a wide
range of irreversible neuronal injuries that give rise
to a relatively slowly evolving death of the cell.
Neuronal atrophy is characterized morphologically
by retraction of the cell body with diffuse basophilia
of the cytoplasm and pyknosis and hyperchromasia
of the nucleus of the neuron, in the absence of an
inflammatory reaction. Neuronal atrophy is thought
to occur in many degenerative disorders that in-
volve several interconnected neuronal systems (i.e.,
multiple-​system atrophy, Friedreich ataxia, and
2 • E scou r o l l e and P oi r ie r ’ s M anua l o f B asic N eu r opat h o l ogy
FIGURE 1.1 Two neurons undergoing apoptosis
are positively stained by in situ end labeling to demon-
strate internucleosomal DNA fragmentation. In one
neuron, on the left, only the nucleus is stained, whereas
in the other, which is at a later stage of the programmed
cell death process, the entire cell body is stained.
Compared to a normal neuron, on the right, both ap-
optotic neurons have similar morphological features
and show a pyknotic nucleus and shrunken cytoplasm.
amyotrophic lateral sclerosis). It is also seen in anter-
ograde and retrograde trans-​synaptic degeneration,
as may occur in the lateral geniculate body following
a lesion of the optic nerve.
Programmed cell death (apoptosis) is an active,
genetically controlled, energy-​consuming process
frequent in neurodegeneration and initially involves
the nucleus of the cell. Neurons undergoing simple
neuronal atrophy or apoptosis have similar mor-
phologic features and may show positive in situ end
labeling of internucleosomal DNA fragmentation
(Fig. 1.1) or be demonstrable by activated caspase
3 immunostaining.
Nerve cell atrophy should not be mistaken with
what is referred to as “dark neurons.” This phenom-
enon is now recognized to be an artifactual change
of the neuron cell body, seen particularly in brain
biopsies fixed in formalin by immersion, and is
characterized by shrinkage of the neuronal cyto-
plasm and a deeply stained, irregularly-shaped nu-
cleus, without other cellular alterations.
1.1.1.1.2. Acute neuronal necrosis (anoxic/​
ischemic neuronal change). Acute neuronal ne-
crosis (anoxic/​ischemic neuronal change) cell death
occurs in not only a wide range of acute injuries, in-
cluding anoxia and ischemia, but also many other
acute pathological processes (e.g., hypoglycemia
or exposure to excessive amounts of excitotoxic
3
FIGURE 1.2 Acute ischemic nerve cell change
(H&E). Eosinophilic, shrunken cytoplasm and
hyperchromatic nucleus.
neurotransmitters). Unlike apoptosis, the predom-
inant cellular changes in acute neuronal necrosis
involve the cytoplasmic organelles and the cell
membrane, which ruptures, leading to cell death.
In experimental animal studies and in carefully
studied human tissue at postmortem, the following
sequence of changes is noted by light and electron
microscopy over the course of 12 to 24 hours after
the insult: (a) cytoplasmic microvacuolization due
to swelling of mitochondria and endoplasmic retic-
ulum; (b) shrinkage of the cell body with retraction
of the cellular outlines and disappearance of Nissl
bodies with eosinophilic condensation of the cyto-
plasm (“red neuron”); (c) condensation of nuclear
chromatin and nuclear pyknosis (Fig. 1.2); (d) late
disappearance of the nuclear chromatin, resulting
in increased acidophilia of the nucleus, which
appears to merge into the surrounding cytoplasm
(karyorrhexis).
Occasionally, dead neurons, especially those
adjacent to old, mostly hemorrhagic, infarcts or to
traumatic scars, become encrusted with basophilic
mineral deposits, chiefly iron and calcium salts.
This condition is referred to as mineralization or
ferruginization of neurons (Fig. 1.3).
1.1.1.1.3. Central chromatolysis. Central
chromatolysis is characterized morphologically by
swelling of the cell body, disappearance of Nissl
bodies beginning centrally and extending outward,
and flattening and eccentric displacement of the nu-
cleus to the periphery (Fig. 1.4). It is seen usually in
lower motor neurons (anterior horns of the spinal
cord, cranial nerve nuclei), where it represents a re-
parative reaction of the cell body to a lesion of the
Chapter 1
FIGURE 1.3 Ferruginization (mineralization)
of the neurons at the edge of an old hemorrhagic
infarct (H&E).
axon (retrograde degeneration or axonal reaction).
Subsequent recovery of normal cell morphology or,
conversely, further progression to nerve cell degen-
eration depends on the reversibility of the axonal
lesion (Fig. 1.5). Central chromatolysis may also be
seen in upper motor neurons, but the phenomenon
is rare and difficult to interpret correctly. Axonal
lesions of neurons whose axons do not leave the
confines of the CNS apparently either do not pro-
duce changes in perikaryal cell-body morphology or
result in a “simple” type of atrophy. Oddly enough,
some metabolic disorders that do not a priori affect
axons (e.g., Wernicke’s encephalopathy, pellagra en-
cephalopathy, and porphyria) may be accompanied
by central chromatolysis in cortical neurons.
A confident diagnosis of central chromatolysis
requires comparison with the normal morphology
FIGURE 1.4 Central chromatolysis (Nissl stain).
Note the cellular swelling, the eccentric displace-
ment of the nucleus, and the margination of the Nissl
bodies.
Basic Pathology of the Central Nervous System • 3
 Complete
central
chromatolysis
Normal
neuron
Recovery
Cell death
Stages of
hyperchromasia
FIGURE 1.5 Nerve cell changes resulting from cen-
tral chromatolysis.
of the affected gray matter structure, because the
nerve cell body in some nuclei (e.g., the mesence-
phalic nucleus of the fifth cranial nerve, Clarke’s
column) normally contains rounded neurons with
marginated Nissl bodies.
1.1.1.1.4. Vacuolated neurons and neu­
ropil. Vacuolated neurons and/or vacuolated neu-
ropil is observed typically in Creutzfeldt-Jakob
disease (Fig. 1.6). In rare instances, swelling with
vacuolization of the nerve cell may result from
transsynaptic degeneration, such as occurs in the
neurons of the inferior olive in olivary hypertrophy
secondary to a lesion of the ipsilateral central teg-
mental tract or of the contralateral dentate nucleus;
this phenomenon is also designated “fenestrated
neurons” (Fig. 1.7).
FIGURE 1.6 Vacuolated neuron in a case of
Creutzfeldt–​Jakob disease (H&E).
4 • E scou r o l l e and P oi r ie r ’ s M anua l o f B asic N eu r opat h o l ogy
FIGURE 1.7 Fenestrated neuron in a case of olivary
hypertrophy (Nissl stain).
1.1.1.1.5. Binucleated neurons. Binucleated
neurons are seen rather infrequently, sometimes
under normal circumstances, or otherwise at the
edge of old focal destructive lesions, as a dysplastic/​
malformation phenomenon (e.g., tuberous sclerosis),
or in certain neoplasms (e.g., ganglion cell tumors).
1.1.1.1.6. Neuronal storage. In some heredi-
tary metabolic diseases related to enzymatic defects
involving synthetic or degradative pathways for
lipids or carbohydrates, interruption of the pathway
leads to cytoplasmic accumulation of interme-
diate substrates or their by-​products, resulting in
swelling and distention of the cell body of nerve
cells, with eccentric displacement of the nucleus
(Fig. 1.8). In several neuronal storage disorders, the
stored material has distinctive histochemical and
ultrastructural features that may help characterize
FIGURE 1.8 Distended nerve cell bodies in a case
of neurolipidosis (combined Luxol fast blue and
Bodian silver impregnation).
5
FIGURE 1.9 Lipofuscin in neuronal cell
body (H&E).
clinically suspected cases of a given storage disease.
Biochemical tests on blood, leukocytes, and other
body fluids are now particularly useful to more pre-
cisely diagnose many of these disorders.
Lipofuscin accumulation within the perikaryon of
neurons and other cells in the nervous system is a
characteristic aging change. Lipofuscin accumulates
in neurons diffusely throughout the brain in
ceroid lipofuscinosis, a neuronal storage disorder.
Lipofuscin is identified on H&E preparations as
refractile yellow-​ brown pigment aggregates (Fig.
1.9). It is autofluorescent and rich in acid phospha-
tase. The pigment is periodic acid–​Schiff (PAS)-
positive and can be stained by Luxol fast blue. It has
distinctive ultrastructural features (see Chapter 10).
1.1.1.1.7. Alzheimer neurofibrillary degenera­
tion and granulovacuolar degeneration. Alzheimer
neurofibrillary degeneration is characteristically seen
in the brains of aged individuals and in patients with
senile dementia of the Alzheimer type, but it has also
been described in a variety of other cerebral disorders.
This degenerative change is manifest by the forma-
tion of neurofibrillary tangles (NFTs), structures
that are well demonstrated by silver impregnation
and by immunohistochemical techniques; they
consist of thickened and tortuous skeins within the
neuronal perinuclear cytoplasm. The configuration
of the tangle may vary according to the anatomic
site, the type of neuron affected, and the stage of
formation (Fig. 1.10). A band-​ shaped perikaryal
NFT can be seen in both large and small pyramidal
cells and is perhaps an early stage of NFT develop-
ment (Fig. 1.10A). A triangular flame-​shaped per-
ikaryal NFT is seen mainly in large pyramidal cells
(Fig. 1.10B, C). Small compact globose perikaryal
Chapter 1
NFTs are typically seen in small cortical neurons
(Fig. 1.10D). Large globose NFTs reminiscent of a
ball of string are more common in the nucleus basalis
of Meynert and in the brainstem (Fig. 1.10E). In the
final stages of the disease, the cell outline disappears
and only the distorted fibrils remain as “ghost NFT”
(Fig. 1.10F).
The predominant biochemical component of
NFTs is the microtubule-​associated protein tau, which
accumulates in an abnormally highly phosphorylated
form. Tangles are particularly well demonstrated by
tau immunocytochemistry, which is now used rou-
tinely in diagnostic work. NFTs may on occasion also
be immunoreactive for ubiquitin and p62. By elec-
tron microscopy, most NFTs consist of paired helical
filaments measuring around 20 nm across with a reg-
ular constriction to 10 nm occurring every 80 nm. In
Alzheimer disease, they may also be associated with
straight filaments. In progressive supranuclear palsy
(PSP), tangles have been found to consist mainly of
straight filaments measuring 15 nm in diameter.
Granulovacuolar degeneration is a neuronal alter-
ation found in pyramidal cells of Ammon’s horn; it
is seen in normal aging as well as in Alzheimer di-
sease and Pick disease. The abnormality consists of
an accumulation of small, clear vacuoles measuring
4 to 5 μm in diameter and containing an argyrophilic
granule that is well stained by hematoxylin (Fig.
1.11). Some of the granules are immunoreactive
for phosphorylated neurofilament, tubulin, tau,
and ubiquitin, suggesting that the vacuoles are
autophagic lysosomal structures in which cytoskel-
etal components are being degraded.
1.1.1.1.8. Intraneuronal inclusion bodies.
Intracytoplasmic or intranuclear inclusion bodies
are important indicators of neuronal injury. They
occur in degenerative, metabolic, and viral diseases
and often have diagnostic immunocytochemical
and ultrastructural features.
Pick bodies are round homogeneous intracyto­
plasmic neuronal inclusions (Fig. 1.12) charac-
teristic of Pick disease, where they may be seen in
pyramidal neurons and dentate granule cells of the
hippocampus, as well as in affected regions of the ne-
ocortex. The structure is intensely argyrophilic and
immunoreactive for tau and the 3R-tau isoform. Pick
bodies are also reactive for ubiquitin and tubulin. The
ultrastructure of the body is characterized by poorly
circumscribed masses of intermediate filaments,
15-​nm straight filaments, and some paired helical
filaments, as well as entrapped vesicular structures.
Basic Pathology of the Central Nervous System • 5
 A B

C D

E F

FIGURE 1.10 Different types of neurofibrillary tangles (Bodian silver impregnation combined with Luxol fast
blue): band-​shaped perikaryal NFT (A); triangular flame-​shaped perikaryal NFT (B, C); small, compact, glo-
bose perikaryal NFT (D); large globose NFT (E); “ghost NFT” (F).

Lewy bodies are neuronal cytoplasmic inclusions;
their appearance varies according to whether they
are found in the perikaryon or in the nerve cell
processes, cortex, brainstem, or sympathetic gan-
glia (Fig. 1.13). Typical (brainstem) Lewy bodies
are roughly spherical with an eosinophilic core
surrounded by a paler “halo.” One or more of these
structures may be present in the cytoplasm of a
single neuron (Fig. 1.13A, B). They may also be
oval or elongated structures, especially when they
occur in axonal processes or in sympathetic gan-
glia (Fig. 1.13C, D). Cortical Lewy bodies are less
clearly outlined and consist of a homogeneous zone
of hypereosinophilia that usually lacks the char-
acteristic surrounding halo (Fig. 1.13E, F). Lewy
bodies are immunoreactive for α synuclein as well as

6 • E scou r o l l e and P oi r ie r ’ s M anua l o f B asic N eu r opat h o l ogy

7
FIGURE 1.11 Granulovacuolar degeneration
(Bodian silver impregnation).
for ubiquitin and αB crystallin. By electron micros-
copy, they consist in an amorphous electron-​dense
core surrounded by a halo of radiating filaments.
Their presence defines several conditions termed
Lewy body disorders; the most common disorder in
this group is Parkinson disease.
Hirano bodies are brightly eosinophilic rod-​shaped
or elliptical cytoplasmic inclusions that appear to
overlap the cell border of a neuron cell body (Fig.
1.14). They are mostly found in the CA1 field of
the hippocampus and are particularly numerous in
Alzheimer disease and Pick disease and in patients with
the Guam parkinsonism–​ dementia complex. They
are immunoreactive for actin and actin-​ associated
proteins. Ultrastructurally, they are highly character-
istic and consist of parallel filaments 60–​100 nm in
length that alternate with a longer sheet-​like material.
Bunina bodies are eosinophilic, nonviral
intracytoplasmic inclusions found in motor neurons
in cases of familial or sporadic amyotrophic lateral
FIGURE 1.12 Neuronal argyrophilic inclusion in
Pick disease (Bodian silver impregnation).
Chapter 1
sclerosis (Fig. 1.15A, B). They are immunoreactive
for cystatin C. Ultrastructurally, they appear as
electron-​dense, membrane-​bound bodies.
Skein –​like inclusions are abnormal ubiquitinated,
p-​62 positive, structures occurring in anterior horn
cells in motor neuron diseases. They are linear,
thread-​like structures; in some instances, they occur
as a single inclusion, and in others, form networks of
threads. Occasionally, the threads are aggregated to
form larger and dense inclusions (Fig. 1.16). They
contain TDP-​43, ordinarily a nuclear protein, and
accumulate within the cytoplasm of motor neurons
in amyotrophic lateral sclerosis. Ultrastructurally,
they consist of bundles of filaments 15–​25 nm in di-
ameter, with a tubular profile on cross section.
Marinesco bodies are small eosinophilic
intranuclear inclusions located chiefly in melanin-​
containing brainstem neurons (Fig. 1.17A). They
are strongly ubiquitin positive.
When ubiquitinated, intranuclear inclusions
may occur in other regions of the brain and sug-
gest other disorders. Small, round eosinophilic
inclusions (about the same size of the nucleolus),
immunoreactive for IC2 antipolyglutamine expan-
sion antibody, are found in neurons of CAG re-
peat diseases (including spinal cerebellar ataxias,
Huntington disease, or dentatorubral-​pallidoluysian
atrophy) (Fig. 1.17B). Larger eosinophilic,
ubiquitinated, inclusions are found in CGG repeats
(X fragile) and neuronal intranuclear inclusion disease
(NIID). Similar large intranuclear inclusions are
found in intranuclear inclusion body disease (INIBD).
Lafora bodies are rounded structures composed of
polyglucosan (polymers of sulfated polysaccharides)
and are similar to corpora amylacea (see discussion
further in the chapter) in composition and staining
characteristics. They are found in large numbers
in myoclonic epilepsy both in the CNS (chiefly in
the dentate nucleus) and in certain tissues outside
the CNS, such as sweat glands, liver, and skeletal
muscle. They usually have a dense, intensely PAS-​
positive core surrounded by filamentous, fainter
PAS-​positive structures (Fig. 1.18).
Viral inclusions are eosinophilic intranuclear
inclusions that occupy a greater or a lesser pro-
portion of the nucleus and may be surrounded by
a clear halo; they are associated with some viral
infections of the CNS (cf. Chapter 5). They are
seen in herpes virus infections, particularly in
necrotizing encephalitis caused by herpes sim-
plex virus (Fig. 5.37C), and in subacute sclerosing
panencephalitis (Fig. 5.31A). In rabies, the viral
Basic Pathology of the Central Nervous System • 7
 A B

C D

E F

FIGURE 1.13 Lewy bodies (H&E). Single (A) and multiple (B) Lewy body(ies) in the perikaryon of
pigmented neurons of the substantia nigra in a case of Parkinson disease. Lewy bodies in axonal processes (C,
D), in the dorsal nucleus of cranial nerve X in a case of Parkinson s disease. Cortical Lewy bodies (E, F) in the
perikaryon of cortical neurons in a case of Lewy body disease.

inclusions are intracytoplasmic and are referred to
as Negri bodies (Fig. 5.30). In rare instances (e.g.,
cytomegalovirus infection; (Fig. 5.38)), both
intranuclear and intracytoplasmic inclusions may
be seen. Viral inclusion bodies are immunoreactive
with appropriate antivirus antibodies, allowing for
a specific diagnosis. Electron microscopy may also
be used to identify virions; however, it is now used
less often in diagnostic work.
1.1.1.1.9. Axonal alterations. Following ax-
onal lesions that disrupt the integrity and con-
tinuity of the cell process, the distal part of the
axon undergoes Wallerian degeneration, which

8 • E scou r o l l e and P oi r ie r ’ s M anua l o f B asic N eu r opat h o l ogy

9

FIGURE 1.14 Hirano bodies are brightly eo-

sinophilic and can be rod shaped, as illustrated, or
elliptical (H&E).
is described further, elsewhere in this book (see
basic lesions of the peripheral nervous system;
Chapter 13).
In conditions associated with nerve cell “at-
rophy” as described previously, the destruction of
the cell body of the neuron results in degeneration
of all of its processes, including the dendrites and
the axon, which becomes swollen, then fragmented,
eventually undergoing atrophy. This phenomenon,
if widespread, as occurs in system degenerations,
results in thinning out and rarefaction of the white
matter, demonstrable with myelin and axon stains.
In these diseases, the phenomenon seems to start at
the most distal portions of the longest axons.
Axonal swellings or spheroids are localized eo-
sinophilic enlargements of the axon. At these
sites along the axon, there is condensation of the
neurofilaments, organelles, and other molecules that
are normally conveyed along the axon by an antero-
grade transport system but accumulate focally when
FIGURE 1.16 Skein-​like inclusion in an anterior
horn cell in a case of motor neuron disease (immuno-
cytochemistry for ubiquitin).
the transport system is interrupted. Spheroid forma-
tion characterizes axonal damage by diverse extrinsic
insults and occurs in trauma and ischemia. Spheroids
are well demonstrated by either silver impregna-
tion (Fig. 1.19A) or immunostaining with ubi-
quitin (Fig. 1.19B), neurofilaments, and precursor
of the beta amyloid protein (β-​APP) (Fig. 1.19C).
The latter is transported by the axonal flow and
accumulates when it is disrupted. The term torpedo is
applied to Purkinje cell axonal swellings and occurs
in a wide range of metabolic and degenerative cere-
bellar diseases. They are well demonstrated by silver
impregnations and by the immunohistochemical
methods mentioned. They are most notable in the
initial portion of the axon, before the origin of the
collateral branches (Fig. 1.19C).
The axonal swellings that develop when ax-
onal transport is disrupted by neuronal metabolic

A B

FIGURE 1.15 Bunina bodies in anterior horn cells of the spinal cord in a case of motor neuron disease (H&E)
(A); immunocytochemistry for cystatin C (B).

Chapter 1 Basic Pathology of the Central Nervous System • 9

 A B

FIGURE 1.17 Intranuclear inclusions: (A) Marinesco bodies: small intranuclear inclusion in a pigmented
neuron of the substantia nigra (H&E); (B) ubiquitin-​positive intranuclear inclusion in a case of spinocerebellar
degeneration with CAG repeat expansion. (Courtesy of Professor Francesco Scaravilli.)

dysfunction are usually termed dystrophic. This
occurs in some acquired (e.g., vitamin E deficiency)
or inherited metabolic diseases. Extensive forma-
tion of axonal swellings is an important patholog-
ical manifestation of neuroaxonal dystrophy and of
some leukodystrophies.
The term dystrophic neurite is used to describe
a neuronal process within the gray matter that
is distended by tau protein or other abnormal
ubiquitinated material. These occur in several neu-
rodegenerative diseases.
1.1 . 1 . 2 . AS T R OC YTI C L ESI O NS
1.1.1.2.1. Gliosis (astrogliosis). The presence
of gliosis (alternate term astrogliosis) is the most cer-
tain indication that a microscopic abnormality is
of pathologic significance and not artifactual. This
FIGURE 1.18 Lafora body in a case of myoclonic
epilepsy (periodic acid–​Schiff).
reactive process accompanies almost any type of
subacute or chronic injury of the CNS. The process
of gliosis is in essence the response of astrocytes to
CNS tissue injury. The associated morphological
changes include an increase in the number of as-
trocyte nuclei per unit area, eosinophilia of the cy-
toplasm around the nucleus, and expansion and
distortion of the astrocytic cytoplasmic arboriza-
tion. For reasons that are not understood, mitotic
figures are only rarely identified in gliotic tissue, and
techniques that bring out dividing cells (Mib-​1/​Ki
67) also confirm the slow turnover.
The morphologic aspects of the process of gliosis
will vary depending on the location, stage of evo-
lution, and nature of the pathological process. The
early stages are characterized by hypertrophy of the
nucleus, which is often hyperchromatic and eccen-
trically placed in the perikaryon. As mentioned, the
cytoplasm around the nucleus and cell processes
become more extensive than normal and are found
to contain glycogen (Fig. 1.20A). Characteristically,
at this stage, in H&E preparations, the cytoplasm
is homogenized and eosinophilic; these reactive
astrocytes are referred to as gemistocytic astrocytes
(Fig. 1.20B, C).
Over time, in chronic disease states and slowly
evolving degenerative processes, astrocyte nuclei re-
turn to their resting size and shape, though their cy-
toplasmic network of cell processes is more extensive
and can best be appreciated with immunostaining
for glial fibrillary acidic protein (GFAP).
An older term, isomorphorphic fibrillary gliosis,
refers to the alignment of reactive astrocyte
processes conforming to a degenerating fiber tract.

10 • E scou r o l l e and P oi r ie r ’ s M anua l o f B asic N eu r opat h o l ogy

1

A B

FIGURE 1.19 Axonal swellings in the white matter identified on silver impregnation (A) (Bodian stain)
and on ubiquitin immunostain (B). Torpedo (axonal swelling) on a Purkinje cell axon identified by β-​APP
immunostaining (C).

1.1.1.2.2. Alzheimer type II glia. Alzheimer
type II glia is seen particularly in hyperammonemic
states such as occur in Wilson disease and in liver
failure from acquired or hereditary metabolic disease,
but can also be found in a variety of other systemic
metabolic disorders (e.g., renal failure). This reac-
tion of astrocytes is characterized by enlargement of
the nucleus, reaching 15 to 20 μm in diameter, which
appears irregular in shape and pale and empty looking
because of the disappearance of chromatin granules
(Fig. 1.21). One or two dense, rounded PAS-​positive
bodies resembling nucleoli are often seen next to the
nuclear membrane, which is always sharply defined.
The cell body is not usually visible on conventional
preparations and stains poorly with GFAP. Alzheimer
II glia are unrelated to Alzheimer disease; they tend to
occur in the gray matter, involving particularly deep
gray nuclei, especially the pallidum and the dentate
nuclei, and also the cerebral cortex. Alzheimer type
II glia are metabolically active cells engaged in the
catabolism of toxic substances such as ammonia; on
ultrastructural study, they are shown to contain nu-
merous mitochondria.
1.1.1.2.3. Rosenthal fibers. By light micros-
copy, Rosenthal fibers are rounded, oval, or elon-
gated beaded structures measuring 10 to 40 μm
that appear homogeneous and brightly eosinophilic
(Fig. 2.4B). On electron microscopy, they con-
sist of swollen astrocytic processes that are filled
with electron-​dense amorphous granular material
and glial filaments. With immunohistochemical
method peripheral labeling for GFAP, ubiquitin
and αB-​crystallin can be demonstrated. Rosenthal
fibers are seen in various pathological conditions
that have in common intense fibrillary gliosis of
long standing, as seen throughout the brain in
multiple sclerosis plaques, in the spinal cord in
syringomyelia, and in the hypothalamus around
craniopharyngiomas. They are also characteristic of
certain neoplasms (pilocytic astrocytoma, partic-
ularly of the cerebellum, but also elsewhere in the

Chapter 1 Basic Pathology of the Central Nervous System • 11

 A B

FIGURE 1.20 Gliosis. Fibrillary gliosis (A) hypertrophy of nucleus and of cytoplasm and processes that
are very visible on GFAP stain. Gemistocytic astrocytes with large homogenized and eosinophilic cytoplasm
(H&E) (B), (GFAP) (C).

brain) (cf. Chapter 2), and in Alexander disease (cf.
Chapter 10).
Eosinophilic granular bodies are rounded hya-
line droplet aggregates that occupy the cytoplasm
of astrocytes and are particularly seen in tumors,
including pilocytic astrocytoma, pleomorphic
xanthoastrocytoma, and ganglion cell tumors.
FIGURE 1.21 Alzheimer type II glial cells (H&E).
1.1.1.2.4. Inclusions and storage mate­
rial. Accumulation of lipofuscin occurs in astrocytes
as part of aging as it does in neurons. Similarly, in
lipid storage diseases, glial lipid storage may accom-
pany neuronal storage.
Tau protein, which is the main component of
NFTs, can also accumulate in astrocytes, particularly
in PSP and corticobasal degeneration (cf. Chapter 8).
Tufted astrocytes are considered to be highly char-
acteristic of PSP (Fig. 8.20A). The whole length
of their processes contains tau protein, and they
are often binucleated. They may be demonstrated
by Gallyas stain or tau immunocytochemistry. In
corticobasal degeneration, accumulation of tau pro-
tein in astroglia forms distinctive structures in gray
matter areas termed astrocytic plaques: Tau protein
accumulates at the end of the astrocytic processes,
while the center of the plaque is devoid of tau
immunoreactivity (Fig. 8.23).
Thorn-​shaped astrocytes and have an argyrophilic
cytoplasm with a few short processes (Fig. 8.20B)

12 • E scou r o l l e and P oi r ie r ’ s M anua l o f B asic N eu r opat h o l ogy

13

immunoreactive for ubiquitin, αB crystallin, and α

synuclein.
Accumulation of tau protein in oligodendrocytes,
known as “coiled body,” may be found in PSP,
corticobasal degeneration and argyrophilic grain
disease (cf. Chapter 8). These consist of fibrillary
structures “coiling” around the nucleus.
Similar coiled bodies, but tau negative and p62
and TDP-​43 positive, are seen in the brain and
spinal cord in TDP-​43–​related frontotemporal de-
generation and amyotrophic lateral sclerosis.
FIGURE 1.22 Numerous corpora amylacea in the
subependymal region (H&E).
1. 1. 1. 4. MIC R OGLIAL LES IONS
Microglial cells are of monocyte lineage and
and often a small excentric nucleus. They are com-
have important phagocytic functions. They can
monly associated with granular-​fuzzy astrocytes and
be demonstrated by silver impregnation, lectin-​
seen in aging-​related tau astrogliopathy (ARTAG).
binding techniques, and immunohistochemical
Viral inclusion bodies may also be found
techniques using antibodies that react with mono-
in astrocytes, such as in subacute sclerosing
cyte/​macrophages (e.g., CD68, CD163, IBA-​1, and
panencephalitis and CMV infection (cf. Chapter 5).
Tmem119) (Fig. 1.23B).
Corpora amylacea are spherical, basophilic, PAS-​
In normal brain, microglia have been subdivided
positive inclusions, 10 to 50 μm in diameter, that
into (a) resident microglia, with little turnover and
are predominantly found in astrocyte processes,
present throughout the CNS parenchyma, and
although they occasionally occur within axons.
(b) perivascular microglia, found within the perivas-
Ultrastructurally, they consist of densely packed 6-​
cular basal lamina and with characteristic turnover
to 7-​nm filaments that may be admixed with amor-
as with hematogenous monocytes. Microglial ac-
phous granular material and are not membrane
tivation occurs in inflammatory conditions of the
bound. Corpora amylacea increase in number with
CNS and involves (a) increased entry of hematoge-
aging, particularly in the subpial and subependymal
nous monocytes into the CNS; (b) proliferation of
regions (Fig. 1.22), around small blood vessels,
resident microglia; (c) expression or secretion of a
and in the posterior columns of the spinal cord.
wide range of proteins, most of which are concerned
Adult polyglucosan body disease (cf. Chapter 10)
with antigen presentation and inflammation.
is characterized by diffuse accumulation of corpora
Microglia reactive processes are typically
amylacea, involving the cortex and white matter and
associated with diffuse or focal myelin damage.
• Macrophage proliferation and phagocytosis (the
cells are also known as compound granular
1. 1 . 1 . 3 . L E S IONS O F
corpuscles, foam cells, lipid phagocytes, or
OL IGODE N DR OC YTES
gitter cells). This is a frequent finding in many
Like neurons and astrocytes, oligodendrocytes may brain lesions, particularly those associated with
be infected by virus and manifest intranuclear or demyelinating processes or with traumatic or
intracytoplasmic inclusions (e.g., progressive mul- ischemic tissue destruction. After a destructive
tifocal leukoencephalopathy) (cf. Chapter 5) or be or demyelinating insult, macrophages invade
affected by excess storage of lipid or glycogen in the damaged region within 48 hours of injury.
genetically determined enzymopathies (e.g., meta- These are rounded cells with distinct cytoplasmic
chromatic leukodystrophy). borders and measure 20 to 30 μm in diameter.
Cytoplasmic inclusions involving mainly They have small, darkly staining, and sometimes
oligodendrocytes have been shown to be a char- eccentric nuclei and a clear, granular cytoplasm
acteristic feature of certain degenerative diseases that can contain lipids, hemosiderin pigment, or
including multiple system atrophy (cf. Chapter 8). any other phagocytized material (Fig. 1.23A, B).
These inclusions are usually flame or sickle shaped, The number of these scavenger cells increases
can be demonstrated by silver impregnation, and are over a period of days and weeks and may be still

Chapter 1 Basic Pathology of the Central Nervous System • 13

 A B
FIGURE 1.23 Perivascular lipid-​laden macrophages (compound granular corpuscles, foam cells, or gitter
cells) in a demyelinating lesion (Luxol fast blue combined with Bodian silver impregnation) (A) and with CD68
immunostaining) (B).
be present in injured tissue months after the
injury. Most derive from blood monocytes.
•​ Rod cell proliferation (Figs. 1.24 and 5.27)
is a form of microglial response to subacute
parenchymal injury in which necrosis is
minimal or absent. Rod cells are elongated,
spindle-​shaped cells that can be recognized
on hematoxylin and eosin preparations by the
presence of a cigar-​shaped nucleus. The best
descriptions of this glial change are found in
reports of cases of general paresis in the older
literature (cf. Chapter 5). Rod cells are also seen
in cases of subacute encephalitis and evolving
ischemic lesions.
•​ Microglial nodules consist of discrete clusters
of microglial cells that are typically found in
subacute encephalitis in and around sites of
neuronal destruction: neuronophage nodules (cf.
Chapter 5).
FIGURE 1.24 Rod-​shaped microglia in a case of
general paresis of the insane (Nissl stain).
14 • E scou r o l l e and P oi r ie r ’ s M anua l o f B asic N eu r opat h o l ogy
1. 1. 1. 5. EP ENDY MAL C ELLS
Ependyma have a limited range of reactions to in-
jury. Along with neurons and other glial cells, epen-
dymal cells may be infected in viral diseases. In the
adult CNS, ependymal cells do not proliferate in
response to injury and cell loss. Their destruction
leaves bare stretches of the ventricular lining; this
is accompanied by proliferation of subependymal
astrocytes, which form small hillocks along the ven-
tricular surface: ependymal granulations. Occasionally,
surviving ependymal cells may be overgrown by the
astrocytic reaction and appear as clusters of tubules
buried within the ependymal granulations.
1 .1 .2 . G e n e r a l t is s u e
r e a c t io n s o f t h e C N S
t o in ju r y a n d h e r n ia t io n s
A set of general tissue reactions is known to occur in
the CNS that stand apart from the reactions to spe-
cific pathological processes (i.e., vascular, infectious,
inflammatory, demyelinating, metabolic, degener-
ative). As described in other chapters of this book,
these may accompany one or more of the specific
pathologic processes described previously that are
visible under the microscope or may result in more
extensive changes that can be visible to the naked eye.
1. 1. 2. 1. C ER EBR AL ATR OP HY
Cerebral atrophy is the end stage of a number of
neurological diseases. The brain is lighter than a
normal age-​matched control. Macroscopically, there
is narrowing of the gyri and widening of sulci. On
15
section, the cortical ribbon is thinned, and ventric-
ular dilation is often present (Fig. 8.9). The histolog-
ical substratum consists of a variable loss of neurons
often associated with gliosis, depending on the un-
derlying illness, and a variety of neuronal alterations,
which are discussed in subsequent chapters.
1. 1 . 2 . 2 . C E R E BRA L ED EMA
Cerebral edema is defined as an increase in the brain’s
volume due to an increase in its water and sodium
contents. Depending on its pathogenesis, brain
edema has been classified as vasogenic, cytotoxic, or
interstitial (hydrocephalic). A combination of these
prototypes of edema is frequent.
• Vasogenic edema, probably the most common type
of brain edema, complicates head injury, abscess,
tumors, and hemorrhages. Both vasogenic
edema and cytotoxic edema occur with ischemia.
Vasogenic edema results from blood–​brain
barrier injury leading to increased permeability
of the microcirculation to macromolecules,
particularly to proteins. By radiologic imaging,
sites of vasogenic edema are marked by contrast
enhancement because the injected contrast
medium leaks across the permeable vascular
lining. Biochemically, the edema fluid resembles
a plasma filtrate. It is located chiefly in the
extracellular spaces of the white matter.
• In cytotoxic edema, excessive amounts of
water enter one or more of the intracellular
compartments of the CNS (neurons, glia,
endothelial cells, or myelin sheaths) because the
cellular concentration of osmotically active solutes
is increased. This usually results from an injury
impairing the capacity of the cell to maintain
ionic homeostasis. It is also seen in association
with systemic disturbances in fluid and electrolyte
metabolism. Cytotoxic edema complicates
hypoxia and ischemia because of failure of the
adenosine triphosphate (ATP)-​dependent
sodium pump in the affected cells. It also occurs
in osmotic disequilibrium syndromes associated
with hemodialysis or diabetic ketoacidosis and
in acute plasma hypo-​osmolality states, such as
water intoxication and inappropriate secretion
of antidiuretic hormone. Because the so-​called
cerebrovascular macromolecular barrier remains
intact, disease processes that give rise to cytotoxic
edema are not associated with radiologic
enhancement after injection of contrast medium.
Chapter 1
• Interstitial or hydrocephalic edema is the
accumulation of cerebrospinal fluid (CSF) in the
extracellular spaces of the periventricular white
matter resulting from obstructive hydrocephalus.
As fluid collects within the obstructed ventricles,
pressure increases, and the CSF is forced
across the ependymal lining into the adjacent
extracellular spaces.
Macroscopically, the edematous areas of brain are
swollen and soft (Fig. 1.25). The swelling increases
the volume of the intracranial contents, with conse-
quent increased intracranial pressure (see Section
1.1.2.3). When the brain is cut, the slice surfaces
may be wet and shiny. If the edema is diffuse, the
ventricles are compressed; in severe cases, they are
reduced to slit-​like cavities.
Under light microscopy, myelin stains demon-
strate pallor of the white matter. The cerebral tissue
has a loose appearance and is split by vacuoles of
variable size. Glial cells are swollen; perivascular and
pericellular spaces are dilated.
These macroscopic and microscopic features cor-
respond to ultrastructural features that vary according
to the etiological and pathogenetic mechanism. They
include dilation of the perivascular and extracellular
spaces, swelling of astrocytic cell processes, and split-
ting of the myelin lamellae (Fig. 1.26).
1. 1. 2. 3. HY DR OC EP HALUS
Hydrocephalus is an abnormal increase in the in-
tracranial volume of CSF associated with dilation
of all or some portion of the ventricular system. It
FIGURE 1.25 Cerebral edema of the left cerebral
hemisphere with swelling of the parenchyma, which
appears paler; flattening of the gyri; and narrowing of
the sulci and left lateral ventricle.
Basic Pathology of the Central Nervous System • 15
 A B
Ast.
A A A
L

Ast. G
Ast. E.C. G
B.M. A
E.C.S
Ast. A
G G
N.

A
V
M
V V

M V

A A

FIGURE 1.26 Cerebral edema: principal ultrastructural forms. (A) Gray matter (traumatic and inflammatory
edemas). Swelling of astrocytic processes, especially near capillaries. Ast, astrocytic cell process; EC, capillary
endothelial cell; L, capillary lumen; N, neuropil; BM, basement membrane. (B) White matter (traumatic and
inflammatory edema). Expansion of extracellular space. A, myelinated axons; G, glial cell process; ECS, extracel-
lular space. (C) White matter (cytotoxic edema due to triethyltin poisoning). Splitting of myelin lamellae at the
intraperiod line. A, myelinated axon; M, myelin; V, vacuole.

is secondary to a dysequilibrium between CSF for-
mation and reabsorption. Rarely, it results from
increased production of CSF (e.g., choroid plexus
papilloma). More commonly, it is the consequence of
altered flow and absorption of the CSF as a result of
obstruction of CSF pathways within the ventricular
system (noncommunicating hydrocephalus) or in
the subarachnoid space (communicating hydroceph-
alus). Obstruction at “bottleneck” areas such as the
foramina of Monro, the aqueduct of Sylvius, and the
exit foramina of the fourth ventricle (lateral foramina
of Luschka and midline foramen of Magendie) can
occur when there is extension of blood or tumor into
the ventricular system. Subarachnoid pathways most
often become blocked over the cerebral convexities
and around the rostral brainstem (incisural block)
as a result of inflammation or hemorrhage. In the
acute phases, the blood clot or inflammatory exudate
forms a barrier to flow. Subsequently, organization of
the clot or exudate leads to fibrous obliteration of the
subarachnoid space.
Hydrocephalus is often associated with increased
intracranial pressure. In children, in the absence of
appropriate shunting procedures, the head can be-
come enlarged when hydrocephalus develops be-
fore the cranial sutures close. When the progressive
obstructive lesion causing the hydrocephalus is not
severe, the hydrocephalic process may stabilize and
the CSF pressure returns to normal limits (“normal-​
pressure hydrocephalus”).
Several alterations in the brain are common to
all forms of hydrocephalus. These include dilation
of the ventricular system; interstitial edema; reduc-
tion of the volume of the white matter; accentua-
tion of the primary, secondary, and tertiary cerebral
sulci (producing a prominent gyral pattern); and
perforation of the septum pellucidum. Disruption
and loss of the ependymal lining, with localized

16 • E scou r o l l e and P oi r ie r ’ s M anua l o f B asic N eu r opat h o l ogy

17

subependymal astrocytic proliferations protruding
into the ventricular cavities, termed ependymal
granulations, is frequent (see Section 1.1.1.5).
Proliferation of the subependymal glia may bring
about stenosis of the aqueduct, which is a cause of
obstructive hydrocephalus in childhood.
1. 1 . 2 . 4 . IN C R E A SED I NTRA CRA NI A L
P R E S S UR E AN D B RA I N H ERNI ATI O N
After closure of the sutures, the volume of the
cranial cavity is fixed by rigid bony walls and
compartmentalized by partitions of bone and dura.
The normal contents of the cranial cavity (blood,
brain, and CSF) are relatively incompressible. Under
these circumstances, an increase in the volume of
the cranial contents will result in increased intracra-
nial pressure.
The intracranial contents may expand because
of diffuse brain edema, increased cerebral blood
flow and blood volume, or the development of
space-​occupying lesions such as tumors, abscesses,
hematomas, or large, recent infarcts accompanied
by edema. The effects of space-​occupying lesions
on intracranial pressure are the result not only of
the mass of the lesion, but also of the accompanying
edema and obstruction of venous or CSF pathways.
In children with still open cranial sutures, increase in
volume of intracranial contents will lead to splaying
of the sutures, resulting in an increase in the size of
the skull and in digital convolutional markings. In
older children and in adults when bony skull can no
longer expand, intracranial hypertension leads to
compression of the brain surfaces against the inner
table of the skull, with consequent flattening of ce-
rebral gyri, narrowing of intervening sulci, and ac-
centuation of foraminal and tentorial markings on
the inferior cerebellar and medial temporal surfaces.
The expanding cerebral mass will also insinuate it-
self into the anatomic openings that can accommo-
date it. These compensatory displacements of brain
from one intracranial compartment to another,
caused by an increase in the volume of intracranial
contents, are referred to as cerebral herniations. They
differ depending on whether the space-​occupying
lesion is supratentorial or infratentorial (Fig. 1.27).
1.1.2.4.1. Cerebral Herniations in Supra­
tentorial Lesions.
• A unilateral lesion (Fig. 1.28) that increases the
hemispheric volume is likely to cause a herniation
of the cerebral hemisphere through openings
limited by the inferior border of the falx and
by the free edge of cerebellar tentorium on the
ipsilateral side of the lesion. Depending on the
size and the site of the expanding lesion within
the hemisphere, one of several forms of herniation
will occur, sometimes in combination:

A Skull B C 5
Falx 1

Tentorium 2
cerebelli
Tentorial 4
3
notch
Foramen
magnum
D E F

6 6

3 4 4
3
7 7
FIGURE 1.27 Localization of the principal types of cerebral herniation. (A) normal aspect of rigid structures
within the skull. (B) Unilateral hemispheric expanding lesions. (C) Herniation through bone flap. (D) Midline
hemispheric expanding lesion. (E) Bilateral hemispheric expanding lesions. (F) Expanding infratentorial lesion.
(1) Cingulate subfalcine herniation. (2) Lateral displacement of midline structures. (3) Central diencephalic
herniation. (4) Temporal herniation. (5) External herniation (through bone flap). (6) Superior cerebellar herni-
ation (through tentorial opening). (7) Cerebellar tonsillar herniation (through foramen magnum).

Chapter 1 Basic Pathology of the Central Nervous System • 17

 A B

FIGURE 1.28 Cerebral herniations: (A) Inferior aspect of the cerebral hemispheres; note the herniated rim of the
right hippocampal gyrus compressing the oculomotor nerve and displacing the brainstem. (B) Cerebral metastases
causing temporal herniation; note displacement of the midline structures and cingulate herniation. (C) Midbrain;
note hemorrhagic lesion in the crus of the peduncle contralateral to the temporal herniation (Kernohan’s notch).
(D, E) Midbrain and pontine hemorrhages involving mostly the tegmentum, secondary to temporal herniation.

• Herniation of the cingulate gyrus under the falx
(subfalcine herniation) with lateral displacement
of the anterior cerebral arteries.
• Lateral displacement of the midline structures (i.e.,
the third ventricle, pineal gland, vein of Galen).
• Downward herniation of the diencephalon
through the tentorial notch with downward
displacement of the floor of the hypothalamus
and of the mammillary bodies (central,
diencephalic herniation).
• Herniation of the hippocampal gyrus in the
tentorial notch between the brainstem and
the free edge of the tentorium cerebelli. The
herniated temporal lobe can compress and
stretch the third and sixth cranial nerves. When
the ipsilateral cerebral peduncle is compressed
directly, a hemiparesis contralateral to the lesion
may ensue; when the contralateral peduncle is
displaced and compressed against the free edge of
the tentorium (Kernohan’s notch), an ipsilateral
hemiparesis may follow; if the adjacent posterior
cerebral artery is compressed, there can be
secondary infarction anywhere along its territory
of distribution.
• Compression due to temporal herniation and
the downward thrust of central diencephalic
herniation may result in stretching of the
blood vessels, especially the veins that
supply the midbrain and pons, which may be
torn and cause potentially lethal brainstem
hemorrhages; these are called Duret
hemorrhages.

18 • E scou r o l l e and P oi r ie r ’ s M anua l o f B asic N eu r opat h o l ogy

19
• External cerebral herniation through surgical or
traumatic defects in the calvarium may also occur.
• Bilateral cerebral lesion or circumstances that result
in global increase of the volume of both hemispheres
will ordinarily result in central diencephalic
herniation or bilateral temporal lobe herniation.
• A midline expanding lesion will likely result in
central diencephalic herniation.
1.1.2.4.2. Cerebellar herniations in infra­
tentorial lesions. Two types of herniations exist:
• Upward herniation of the mesencephalon and
cerebellum through the tentorial notch. Direct
mesencephalic lesions may result from this
complication, as well as secondary lesions due to
vascular compression.
• Cerebellar tonsillar herniation through the
foramen magnum is the most frequent and most
dangerous complication of an infratentorial
expanding process, regardless of the nature of
the insult or, in case of a neoplasm, the degree of
malignancy. The result of increased intracranial
pressure in the posterior fossa is the herniation
of the cerebellar tonsils downward through the
foramen magnum (Fig. 1.29), culminating in
medullary compression with compromise of vital
cardiorespiratory centers.
1 . 2. TOP OG RAPHIC AN ALYSIS
OF C EN TR AL N E RVOUS
S Y S TEM LE SION S
Topographic analysis of the lesions observed is
just as important as the study of their morphologic
A B
FIGURE 1.29 Cerebellar tonsillar herniation: (A) posterior view; (B) anterior view.
Chapter 1
aspects. It constitutes a crucial step in the attempt
to arrive at an etiological diagnosis and necessitates
a rigorous and systematic examination of all the
neural structures. Systematic sampling of mul-
tiple anatomic levels is necessary, and wherever
possible, techniques that allow for whole-​brain
sections provide invaluable material that permits
the synchronous study of various areas of the CNS
under the dissecting and light microscopes.
1 .2 .1 . D if f u s e d is t r ib u t io n
Lesions that are diffusely distributed throughout
the brain may be seen in systemic diseases, such
as metabolic or circulatory disorders, or also can
be the result of blood-​borne infective processes.
Some of the degenerative diseases may likewise
cause diffuse lesions of the CNS. It is nevertheless
important to emphasize that, despite the diffuse
character of these changes, lesions often show re-
gional predominance.
1 .2 .2 . F o c a l d is t r ib u t io n
• Lesions may be localized to an anatomically well-​
defined area (lobe of the cerebral hemisphere,
basal ganglia, brainstem), and certain preferential
sites of involvement are linked to specific
etiological entities (e.g., some cerebral tumors
preferentially occur in certain locations of the
brain).
• Lesions may be localized to a vascular territory. By
definition, cerebral infarcts exemplify this type of
focal lesion.
Basic Pathology of the Central Nervous System • 19
1 .2.3. D i ssem i n a te d
d is tri bution
Disseminated distribution is seen essentially in mul-
tifocal processes, of which multiple sclerosis is the
most characteristic example.
1 .2.4. Systema ti ze d
d is tri bution
A number of nervous system disorders, especially
degenerative diseases, cause changes that involve cer-
tain functionally related morphological systems, such
as involvement of upper and lower motor neurons
in amyotrophic lateral sclerosis; spinocerebellar in-
volvement in Friedreich ataxia, and others.
1 .3. S Y NTHE T IC
I N TEGR ATION
The findings in the two aspects of neuropatholog-
ical examination, artificially set apart as morphologic
20 • E scou r o l l e and P oi r ie r ’ s M anua l o f B asic N eu r opat h o l ogy
and topographic analyses, need to be integrated.
Furthermore and most important, correlation of
these findings with the clinical data, laboratory and
radiologic data, general autopsy findings, and all
other available diagnostic data must occur to arrive
at an accurate etiologic diagnosis.
Thus, for example, an understanding of cere-
bral infarcts is possible only after careful and com-
plete postmortem examination of the vascular
tree, heart, and lungs and after comparing the ana-
tomical findings with information provided by the
clinical picture, the chronology of the functional
disturbances, and data from cerebral and vascular
imaging.
Likewise, the study of the lipidoses cannot
be based solely on neuropathological findings.
It necessitates detailed alignment with data
from the general postmortem examination and
neurochemical/​genetic analysis.
Finally, the analysis of hereditary disorders of the
CNS or peripheral nervous system and of diseases of
skeletal muscles now requires correlation with mo-
lecular and genetic data.
21
2
Tumors of the Central Nervous System
KE I T H L . L I G O N , S A ND R O S A NTA G ATA , A N D F RAN CK BIEL L E
2 . 1. C LAS S IF ICAT ION
TRADITIONALLY, THE classification of nervous
system tumors has relied on the histological ap-
pearance of a particular neoplasm supplemented
by immunohistochemical features; however, tumor
classification and patient management are increas-
ingly dependent on the evaluation and incorporation
of molecular and genetic information. The classifica-
tion scheme used in this book is based on the (2016)
World Health Organization (WHO) Classification
of Tumors of the Central Nervous System and the
International Society of Neuropathology—​Haarlem
Consensus Guidelines. These works support that
practitioners provide “integrated diagnoses,” which
include histologic diagnosis, histologic grade, and
molecular/​genetic information.
Tumors of the central nervous system
(CNS) can be broadly grouped into two major
categories: primary tumors and secondary tumors.
Primary tumors arise from cells that are intrinsic to
the CNS or its coverings, including the calvarium,
and include tumors of neuroepithelial origin and
nonneuroepithelial origin. Secondary tumors arise
from sites elsewhere in the body and involve brain
or spinal cord mainly by hematogenous dissem-
ination (metastases) or less often by contiguous
extension.
The CNS tumors can also be grouped according
to location and their corresponding incidence by age.
In adults, approximately 70% of all brain tumors occur
supratentorially (i.e., within the cerebral hemispheres
or coverings); these include, in order of frequency,
metastases followed by gliomas, meningiomas,
and schwannomas. In children, approximately 70%
of brain tumors occur infratentorially (i.e., within
the cerebellum or brainstem) and are neuroepi-
thelial in origin, the most common being pilocytic
astrocytoma, medulloblastoma, and ependymoma
in decreasing order of frequency. Spinal cord tumors
constitute about 15% of all primary CNS tumors and
include schwannomas, meningiomas, and gliomas
(ependymoma, astrocytoma) in decreasing order of
frequency.
• 21
2 .2. P R IM ARY N E OPL ASMS 2.2.1.1.1. Astrocytic diffuse gliomas. The
clinical characteristics of patients with the astro-
2 .2.1. Tum ors o f cytic diffuse glioma type of astrocytoma typically
include the following features: occurrence at any
n e uroepitheli a l ti s s u e site throughout the CNS of adult individuals, dif-
2.2 . 1 . 1 . DIF F US E G L I O MA S fuse infiltration of adjacent and often distant brain
structures, and tendency for progression to anaplasia
Diffuse gliomas are characterized by the single-​
over time.
cell infiltration of normal brain tissue by tumor
A number of grading schemes have been used
cells beyond the tumor core in distinction to other
to define diffusely infiltrating astrocytoma bases
types of neuroepithelial neoplasms that are more
on morphologic criteria; the current WHO classi-
circumscribed (e.g., medulloblastoma, pilocytic
fication grading system is adapted from the Sainte
astrocytoma). Complete surgical resection of dif-
Anne/​Mayo grading system. The grading criteria are
fuse gliomas is consequently not possible in most
based on the presence or absence of four histologic
circumstances. However, outcomes are highly vari-
features (nuclear pleomorphism, mitoses, microvas-
able depending on the tumor type and patient age.
cular proliferation, and necrosis) and allow for a three-​
Clinical relapse is almost always the case in patients
tier grading system (Table 2.2). The prognostic
with the majority of diffuse gliomas; resistance to
value of these grading systems has been well estab-
chemotherapy and radiotherapy develops with inev-
lished previously but is currently under reevaluation
itable disease progression. The time course of tumor
in the context of the molecular analysis of the case
progression however is variable, from a few months
and the integrated diagnosis.
to several tens of years, and depends on a number of
clinical parameters.
2.2.1.1.1.1. Diffuse astrocytoma (WHO grade II).
Once the diagnosis of diffuse glioma is established
The diffuse astrocytoma tumors constitute about
on microscopic examination, an attempt is made to
10%–​15% of all astrocytic neoplasms. All age groups
identify molecular pathways of oncogenesis to arrive
can be affected, although these are mainly tumors of
at an integrated diagnosis. These molecular pathways
adults, with a median age in the mid-​30s. The tumor
include the presence or absence of genetic alterations
most commonly localizes to one of the cerebral
(Table 2.1). Each may be associated with a character-
hemispheres (especially the frontal lobes), followed
istic tumor cell phenotype (i.e., oligodendroglial or as-
by the brainstem and spinal cord and, rarely, the
trocytic). Importantly, genetic alterations rather than
cerebellum. The clinical manifestations that bring
cell lineage (defined by morphology or gene expres-
the patient to medical attention reflect the area of
sion) are typically now the basis of the classification of
the brain involved by the tumor. Seizures are a fre-
diffuse gliomas, as the tumors within this group have a
quent presenting symptom. Imaging studies usually
high degree of shared lineage phenotype. Accordingly,
show a poorly defined, homogeneous, low-​density,
some previously described entities within the diffuse
non–​contrast-​enhancing lesion. The development
glioma category have been included under other
of focal contrast enhancement in the course of the
categories. For example, the entity “oligoastrocytoma,”
patient’s illness strongly suggests progression to-
which was defined on morphologic grounds as a
ward anaplasia and a higher histologic grade with
diffuse glioma with mixed oligodendroglial and as-
more aggressive clinical course.
trocytic lineage cells, has now been removed as a
The macroscopic distinguishing features of dif-
separate classification as such tumors are now readily
fuse astrocytoma are those of a mass lesion with
classified as astrocytomas or oligodendrogliomas
indistinct boundaries that distorts involved brain
by cytogenetic or molecular testing. If molecular
structures, resulting in blurring of normal anatomic
testing of a diffuse astrocytoma cannot be obtained,
landmarks (Fig. 2.1A). Cyst formation of varying
the 2016 WHO classification allows for a designation
size may give the tissue a spongy or gelatinous ap-
of tumor type based solely on microscopic criteria
pearance. Focal calcifications may be present.
(e.g., oligodendroglioma, astrocytoma, glioblastoma
The microscopic characteristics of diffuse
[GBM]) plus the designation of NOS (not otherwise
astrocytoma include low-​to-​moderate tumor cell
specified) if molecular testing results are not available
density with cytologic properties resembling those
or NEC (not elsewhere classified) if no definitive sub-
of well-​differentiated astrocytes (Fig. 2.1B). Some
class criteria are felt to be present.
nuclear atypia is usually present, which helps to

22 • E scou r o l l e and P oi r ie r ’ s M anua l o f B asic N eu r opat h o l ogy

 23

Table 2.1 Integrated Diagnosis of Diffuse Gliomas

I N T E G R AT E D OLIGODENDROGLIOMA, ASTROCYTOMA ASTROCYTOMA GLIOBLASTOMA, GLIOBLASTOMA, DIFFUSE MIDLINE

DIAGNOSIS I D H M U TA N T A N D 1 P /​ I D H M U TA N T IDH WILD TYPE I D H M U TA N T IDH WILD TYPE GLIOMA, H3 K27M
1 9 Q C O -​D E L E T E D M U TA N T

Present genetic IDH1/​2 and TERT muta- IDH1 and ATRX Sometimes EGFR, IDH1 mutation TERT mutation Histone H3.3 or H3.1
alterations tion with 1p/​19q whole mutation PDGFRA, BRAF, K27M mutation
arm co-​deletion FGFR, MYB/​
MYBL1
Absent genetic ATRX, TP53 Absence of 1p/​19q Absence of IDH1/​2 1p/​19q co-​deletion Absence of IDH1/​ IDH1/​2 mutation or
alterations co-​deletion mutation, 1p/​19q 2 or H3 K27M 1p/​19q co-​deletion
co-​deletion mutation
Grade II or III II, III II, III IV IV IV
Table 2.2 Grading of Diffuse Astrocytoma

WHO GRADE D E S I G N AT I O N HISTOLOGIC CRITERIA

II Diffuse astrocytoma Rare mitoses

III Anaplastic astrocytoma Frequent mitosesa
Microvascular proliferation AND/​OR necrosisb usually
IV Glioblastoma associated with frequent mitoses
a
The presence of a single mitosis is deemed sufficient to classify a diffuse astrocytoma as a WHO grade III tumor if the tissue sample is small (i.e., in
the case of stereotactic needle biopsy). In the case of a large tumor resection sample, the finding of rare mitoses is not considered sufficient justification to
designate the tumor as grade III. Nevertheless, no quantitative threshold has been defined.

b
Necrosis is not required for the diagnosis of glioblastoma as long as microvascular (endothelial) proliferation is present.

distinguish the neoplastic cells from nonneoplastic
astrocyte nuclei. Mitoses are rare or absent.
Microvascular proliferation or necrosis should not
be present. The background matrix may be loose,
vacuolated, or microcystic. The Ki-​67/​MIB-​1 la-
beling index (a measure of cellular proliferation)
is usually less than a few percent (typically < 5%).
Also, an important feature of diffuse astrocytoma is
the characteristic fibrillary nature of the tissue sur-
rounding the tumor cell nuclei, which have scant
perikaryal cytoplasm on conventional hematoxylin
and eosin (H&E) preparations within a loose but
consistently glial fibrillary acidic protein (GFAP)-​
positive fibrillary matrix. In some cases, the tumor

A B

C D

FIGURE 2.1 Diffuse astrocytoma: (A) thalamic astrocytoma (macroscopic). Microscopic features: (B)
low-​grade fibrillary astrocytoma (H&E); (C) gemistocytic astrocytoma (H&E); (D) anaplastic
astrocytoma (H&E).

24 • E scou r o l l e and P oi r ie r ’ s M anua l o f B asic N eu r opat h o l ogy

25
contains abundant gemistocytic cells that have
eosinophilic cytoplasm and peripherally displaced
nuclei and strongly express GFAP; these are des-
ignated as gemistocytic astrocytoma (Fig. 2.1C).
Perivascular inflammation is commonly present
in this variant. Astrocytomas often involve the
isocortex and in some cases show extensive mucoid
degeneration and a cobweb-​like appearance; this
morphologic variant, previously designated proto-
plasmic astrocytoma, is now no longer considered a
separate entity.
The characteristic genetic changes in grade II
astrocytomas include mutations of isocitrate de-
hydrogenase gene 1 (IDH1 mutations); mutations
of the TP53 tumor suppressor gene, which leads
to accumulation of P53 protein; and mutations of
the ATRX gene. While IDH2 mutations have been
described, these are rare and most commonly found
in oligodendroglioma. While sequencing assays can
be used to identify mutations in IDH1, TP53, and
ATRX, immunohistochemical surrogates to such
assays have been developed that provide useful and
rapid tools for inferring the status of these genes.
One such antibody recognizes the most common
IDH1 mutation R132H, which is present in over
90% of IDH1 mutant cases. Integrated diagnoses
for diffuse astrocytomas now include whether the
tumor is IDH mutant or IDH wild type. The loss of
ATRX expression is associated with ATRX mutation
and is a valuable tool for supporting the integrated
diagnosis of diffuse astrocytoma, IDH mutant.
Most patients with diffuse low-​grade astrocytoma
will develop progression to a higher grade tumor
(i.e., anaplastic astrocytoma WHO grade III or sec-
ondary GBMs IDH-​mutant WHO grade IV). The
median survival has been reported as between 6 and
8 years in the past, but reassessment of patient out-
come in light of IDH mutation status suggests that
patients with IDH-​ mutant diffuse astrocytomas
have a much longer median survival of approxi-
mately 11 years. IDH wild-​type diffuse astrocytoma
is a rarer neoplasm.
Patients with tumors morphologically similar
to diffuse astrocytomas but lacking IDH mutations
have a broader range of clinical outcomes and ge-
netic alterations. These alterations frequently
overlap with pediatric or young adult low-​grade
diffuse glioma subtypes that typically do not prog-
ress to higher grade lesions and have favorable
long-​term survival even compared to IDH-​mutant
astrocytoma. However, it is critical to also ensure
that an IDH wild-​type, lower grade astrocytomas do
Chapter 2
not represent undersampled GBM IDH wild type,
which often can be surmised based on advanced pa-
tient (>50 years old) and radiological features.
2.2.1.1.1.2. Anaplastic astrocytoma (WHO grade
III). The anaplastic tumor develops most often
in the setting of a preexisting low-​grade diffuse
astrocytoma but de novo presentation also occurs.
The age of onset on average is about 10 years
older than that of patients with low-​grade diffuse
astrocytoma. The tumor location, clinical expres-
sion of signs/​symptoms, macroscopic appearance,
and imaging features are otherwise similar to
grade II astrocytoma, and most are nonenhancing.
However, some irregular or mild contrast enhance-
ment may be present within portions of the tumor,
but not the typical obvious “ring” enhancement
seen in GBM.
Microscopically, the tumor has the appearance
of a diffusely infiltrating astrocytoma but shows
increased cellularity, nuclear atypia, and mitotic ac-
tivity in comparison to its low-​grade counterpart,
but microvascular proliferation and necrosis are
absent (Fig. 2.1D). Many, but not all, tumor cells
may express GFAP. Ki-​67/​MIB-​1 labeling indices
are generally increased (usually in the range of 5%–​
10%) but can overlap with both low-​grade diffuse
astrocytoma and GBM. The typical clinical course
of patients with this tumor is aggressive, with sur-
vival of only 2–​3 years from the time of diagnosis.
At the molecular level, anaplastic astrocytomas
share many of the genetic features of diffuse
astrocytoma grade II, including mutations in
IDH1, ATRX, and TP53. The median survival of
patients with IDH-​mutant anaplastic astrocytoma
is approximately 9 years. IDH wild-​type anaplastic
astrocytomas frequently have genetic lesions sim-
ilar to those in primary GBM IDH wild type, in-
cluding inactivation of cell cycle control pathway
genes CDKN2A/​p16/​ARF and RB, amplifica-
tion of CDK4/​ 6, losses on chromosome 10,
polysomy 7, and amplification of EGFR. The out-
come for patients with IDH wild-​type anaplastic
astrocytoma is similar to that of patients with
GBM. Some cases of IDH wild-​type anaplastic
astrocytoma may represent instances of GBM that
have been undersampled, which explains their sim-
ilarly short survival.
2.2.1.1.1.3. Glioblastoma (WHO grade IV).
Glioblastoma (previously known as “glioblastoma
multiforme” and still abbreviated as GBM) is a
Tumors of the Central Nervous System • 25
malignant, rapidly progressive, and fatal astrocytic
neoplasm. GBM is the most common primary
brain tumor, accounting for approximately 10%–​
15% of all intracranial tumors and 40%–​50% of all
glial tumors. The tumor most often arises de novo
in the absence of a preexisting lower grade astro-
cytic tumor (“primary GBM,” > 90% of tumors)
but may also develop from a less malignant diffuse
astrocytoma typically associated with IDH1, TP53,
and ATRX mutations (“secondary GBM”). GBM
occurs in all age groups but most often occurs in
adult individuals, with a peak incidence between the
ages of 55 and 85 years. IDH wild-​type GBM is un-
common in patients less than 40 years old.
The tumor may arise in any region of the CNS;
however, it is most commonly centered in the sub-
cortical white matter of the cerebral hemispheres,
particularly the temporal, parietal, frontal, and oc-
cipital lobes in descending order of incidence. In
contrast, IDH-​mutant GBM commonly localizes to
the frontal lobes. GBM can also occur in the thal-
amus or the basal ganglia, particularly in children;
it occurs rarely in the brainstem or the spinal cord.
In these midline regions, ruling out the presence of
Histone H3 K27M mutation is important.
The radiologic features of GBM typically in-
clude the presence of an irregular mass with a cen-
tral hypodense region of necrosis surrounded by a
contrast-​enhancing “ring” that represents the more
cellular and vascularized portions of the tumor.
A classic radiologic (and macroscopic) presenta-
tion is the “butterfly” pattern due to spread of the
tumor across the corpus callosum to the contra-
lateral hemisphere. In the tissue surrounding the
tumor, there may be considerable vasogenic edema
manifested on imaging studies as hyperintensity on
a T2-​weighted MRI scan.
Macroscopically, the tumor often appears as a
relatively well-​defined mass lesion, although on
microscopic examination there is almost always ex-
tensive infiltration of tumor into the surrounding
brain tissue. A “variegated” appearance with solid
gray-​pink tissue at the periphery and yellow zones
of central necrosis is typical (Fig. 2.2A). In some
instances, there is evidence of both old and recent
hemorrhage. GBM tumors have extensive tumor cell
infiltration into adjacent tissue and extension along
anatomic fiber tracts, sometimes at considerable dis-
tance from the main tumor mass. As such patients
may present with multiple foci seemingly distant
from main tumor mass, creating the impression of
a multifocal lesion on neuroimaging (see discussion
26 • E scou r o l l e and P oi r ie r ’ s M anua l o f B asic N eu r opat h o l ogy
of gliomatosis cerebri further in the chapter). While
true multifocal gliomas may occur, their exact fre-
quency appears to be much lower than their previ-
ously estimated range (2.4%–​7.5% of all gliomas),
and most such cases are a single tumor with regional
variation in tumor density. Multifocal tumors would
by definition be polyclonal and proven so only by
molecular analysis. In some cases, there is exten-
sion of the tumor into the subarachnoid space or
ventricles, with the potential for cerebrospinal fluid
(CSF) dissemination, although this complication is
infrequent. Extracranial extension and hematoge-
nous dissemination are very rare in general and only
occur in patients who have had prior surgery.
When the GBM is IDH wild type, the patient’s
clinical course is comparable to that of the most
highly malignant of human tumors, where the mean
survival ranges from less than 1 year to 18 months,
with less than 5% of patients surviving longer than
5 years. The typical microscopic features of GBM
include high cellular density, marked nuclear atypia,
mitoses, and generally both microvascular prolif-
eration and necrosis, although the presence of ei-
ther one of these is sufficient to render a diagnosis
of GBM. However, the microscopic appearance
of GBM can be highly variable with considerable
regional heterogeneity. In some cases, the tumor
cells may have considerable nuclear and cyto-
plasmic pleomorphism with multinucleated giant
cells (Fig. 2.2B), whereas in others the picture may
consist mainly of small “undifferentiated” cells with
scant cytoplasm and often poor GFAP expression
(see small-​cell GBM discussion). While in many
cases the tumor has zones with better differentiated
fibrillary and gemistocytic astrocytes, in all
examples of astrocytomas, including GBM, there is
sometimes considerable morphologic evidence of
oligodendroglioma-​like cell differentiations. Other
cell types that may be present in GBM include cells
with glandular or epithelioid features, granular cells
that are periodic acid–​Schiff (PAS) positive, heavily
lipidized cells, as well as interspersed densely packed
bundles of sarcomatous cells.
Proliferative activity is prominent in GBM, and
both typical and atypical mitoses are found. Ki-​67/​
MIB-​1 labeling indices are likewise high, commonly
averaging 15%–​20%. Necrosis is a characteristic
feature of GBM and can consist of either large con-
fluent foci of coagulative necrosis or small, band-​like
palisading necrosis (a characteristic and highly diag-
nostic pattern with a rim of densely packed tumor
cells around necrotic foci) (Fig. 2.2C). Microvascular
27

D
C

FIGURE 2.2 Glioblastoma: (A) glioblastoma (macroscopic). Microscopic features: (B) cellular anaplasia,
mitoses (H&E); (C) necrosis with pseudopalisading (H&E); (D) microvascular proliferation with glomeruloid
structures (H&E).

proliferation is defined as an abnormality of the vessel
wall of the microcirculation that becomes thickened
to two or more layers by mitotically active endothe-
lial cells; in extreme instances, these take the form
of balls or proliferating cells and are referred to as
glomeruloid structures (Fig. 2.2D). Microvascular
proliferation is frequently associated with throm-
bosis of the affected vessels.
The pathogenesis and molecular genetics of
GBM have been an area of intense investigation
in recent years. The origin of GBM remains con-
troversial; the “traditional” explanation is that the
tumor arises from differentiated adult astrocytes
or astrocyte precursors. However, based on recent
human and animal studies, it is now suggested that
the tumor may arise instead from neuroepithelial
progenitor cells, including oligodendroglial progen-
itor cells that are present throughout adult life.
Glioblastoma is one of the most well-​
characterized cancers at the molecular level and was
one of the first cancers to be studied using large-​scale
integrative genomic approaches. Collectively, these
molecular studies have consistently identified three
core signaling pathways that are disrupted in GBM
IDH wild type: increased activation of receptor
tyrosine kinases/​RAS/​Phosphatidylinositol 3-ki-
nase signaling, loss of function in TP53 signaling,
and reduced signaling of the retinoblastoma (RB)
pathway. The activation of RTK/​RAS/​PI3K signaling
is evident in 88% of GBMs and most characteristi-
cally occurs due to amplification of the EGFR gene,
along with rearrangements and overexpression of

Chapter 2 Tumors of the Central Nervous System • 27

mutant EGFRvIII and extracellular domain mutants.
Additional activation of these pathways can occur
through amplification of PDGFRA, MET, AKT, or
PIK3CA and aberrations that lead to loss of func-
tion for the PTEN tumor suppressor gene. Studies
of the TP53 gene have shown that rates of muta-
tion and inactivation of this pathway are higher than
once previously thought in adult GBM, and TP53 is
now known to be the most frequently mutated gene
in GBM, occurring in at least 42% of adult tumors.
The RB pathway is targeted through a number of
different means, including genomic losses and muta-
tion of RB, along with genomic losses targeting the
CDKN2 family of genes or amplification of the nega-
tive regulators of the RB pathway, such as CDK4.
Molecular studies have shown that in secondary
GBM there are different sets of genetic alterations
compared to those in primary GBM, predominantly
IDH1 mutations. Furthermore, secondary GBM
arises from lower grade astrocytomas by a sequen-
tial acquisition of high-​grade genetic alterations,
including concurrent tumor suppressor losses (e.g.,
CDKN2A/​B) and oncogenic drivers (e.g., PIK3CA,
PDGFRA).
In childhood, GBM mostly arises de novo, and
the tumor differs genetically from adult GBM, with
mutations in histone H3.1/​H3.3 genes being most
common. The K27M mutation in H3.1/​H3.3 or
G34R/​V mutation in H3.3 defines the majority of
these tumors. The H3 K27M mutation is generally
found in midline tumors of the pons and thalamus,
and the H3 G34R/​V mutation is typically present in
tumors arising in the cerebral hemispheres. Tumors
that have H3 mutations and are in the midline
structures (thalamus, brainstem, spine) are referred
to in the WHO 2016 classification as a separate
class, designated diffuse midline gliomas, H3 K27M
mutant. The cerebral tumors more commonly occur
in teenagers and young adults. A subset of cases of
diffuse intrinsic pontine gliomas has mutations in
ACVR1. In the pediatric age group, cases of GBM
arising in the hemispheres also frequently harbor
mutations in TP53 and ATRX as well as ampli-
fication or mutation of PDGFR and deletion of
CDKN2A. Unlike adult GBM, amplification of
EGFR is rare. GBM presenting in childhood rarely
arises from progression from a lower grade glioma.
The prognostic evaluation of patients with GBM
is a frequent clinical problem, and genetic testing has
emerged as a reliable means for predicting patient
outcomes. Studies have consistently identified that
increased methylation of the promoter of the MGMT
28 • E scou r o l l e and P oi r ie r ’ s M anua l o f B asic N eu r opat h o l ogy
gene is associated with increased progression-​free
survival in adult GBM. MGMT promoter methyl-
ation and decreased expression of MGMT protein
is detected in nearly 50% of GBM. MGMT encodes
a DNA repair protein that removes alkyl groups
from the O6 position of guanine, thereby mitigating
the therapeutic benefits of DNA alkylation by
temozolomide. Detection of IDH1 mutation by
IHC or sequencing as described previously is also
an effective means of identifying patients with a sig-
nificantly more favorable prognosis.
2.2.1.1.1.4. Glioblastoma variants
2.2.1.1.1.4.1. Giant cell GBM (WHO grade
IV) This is a rare variant of IDH wild-​type GBM
that accounts for less than 5% of all cases of GBM.
The tumor usually arises de novo without evi-
dence of a preexisting astrocytoma and is oth-
erwise similar in clinical presentation to typical
GBMs. Radiologically and macroscopically, the
tumor tends to be better circumscribed compared
to ordinary GBM, often arising subcortically in
the temporal or parietal lobes. The typical micro-
scopic characteristics include the presence of giant
and multinucleated cells, which may show vari-
able expression of GFAP. These features can raise a
problem of differential diagnosis as they are shared
with anaplastic pleomorphic xanthoastrocytoma
(PXA; reticulin deposits favor PXA). In giant cell
GBM, there is a high frequency of TP53 mutations
and also commonly PTEN mutations. Amplification
of EGFR and deletion of CDKN2A is uncommon.
Affected patients generally have a poor prognosis,
although some reports have suggested a somewhat
better clinical outcome (>10% 5-​ year survival),
which if true may possibly result from better surgical
access to a more circumscribed tumor mass.
2.2.1.1.1.4.2. Gliosarcoma (WHO grade IV)
The designation gliosarcoma applies to a neo-
plasm having an intermixed or biphasic pattern
containing both neoplastic glial and mesenchymal
tissue elements. This entity comprises approxi-
mately 2% of all cases of GBM. The tumor is usu-
ally located in the cerebral hemispheres, most often
involving the temporal lobes. The radiographic
appearance can be identical to typical GBM or, if
the sarcomatous component predominates, can
manifest as a hyperdense, circumscribed, uni-
formly contrast-​ enhancing mass mimicking a
meningioma. The diagnosis rests on establishing
the presence of unequivocally malignant glial
29

and mesenchymal elements. The sarcomatous re- 2.2.1.1.2. Oligodendroglial diffuse glioma
gions usually consist of malignant spindle cells ar- 2.2.1.1.2.1. Oligodendroglioma, IDH mutant,
ranged in a fascicular, herringbone, or sometimes and 1p/​19q co-​deleted (WHO grade II). The des-
a “storiform” pattern and may rarely show other ignation oligodendroglioma applies to a diffusely
types of mesenchymal differentiation, including infiltrating glioma composed of tumor cells mor-
cartilage, bone, skeletal, and smooth muscle. Rare phologically resembling mature oligodendrocytes.
examples have shown cytokeratin-​positive epithe- The tumor is now defined on a genetic basis by the
lial and adenoid or gland-​like structures. GFAP demonstration of an IDH mutation and co-​deletion
immunohistochemistry is very helpful in dis- of the chromosomal arms 1p and 19q. The tumor
tinguishing between the glial and mesenchymal accounts for approximately 5% of all intracranial
components; likewise, a reticulin stain will show gliomas and occurs mostly in adults, with a peak
abundant reticulin fibrils in the sarcomatous, but incidence between the ages of 30 and 60. It most
not glial, component. There appears to be a mon- often arises in the cerebral hemispheres, usually
oclonal origin of both the glial and mesenchymal the frontal lobes. The tumor must be distinguished
components of the tumor, with the mesenchymal from a separate category of diffuse leptomeningeal
component probably representing an advanced glioneuronal tumors that may also demonstrate
form of glial-​mesenchymal transition analogous to “oligodendroglioma” morphologic aspects together
epithelial-​mesenchymal transition in carcinomas. with the 1p/​19q co-​deletion. Because of the slow-​
The overall prognosis of gliosarcoma is essentially growing nature of the tumor, patients may have a
the same as ordinary GBM. long history of neurological symptoms, most often
a seizure disorder. Imaging studies usually show a
2.2.1.1.1.4.3. Epithelioid GBM (WHO grade well-​demarcated mass, often with calcification and
IV) This tumor is a variant of GBM composed occasionally intratumoral hemorrhage; peritumoral
predominantly of dense collections of GFAP-​ edema and contrast enhancement are uncommon.
positive malignant epithelioid cells; mitoses, mi- The macroscopic appearance of oligodend­
crovascular proliferation, and necrosis are seen. roglioma is characteristically that of a variably well-​
Scattered rhabdoid-like cells are present. The circumscribed, grayish pink neoplasm including
tumor histologically overlaps with anaplastic PXA, areas of mucoid change, which may result in a gelati-
suggesting a shared histogenesis. This point is fur- nous consistency as well as zones of cystic degenera-
ther supported by the fact that about half of the cases tion, focal hemorrhage, and calcification.
are BRAF V600E mutant, and the tumor occurs in The microscopic characteristics of oligodend­
children and young adults similar to PXAs. A very roglioma are very distinctive. In routine paraffin-​
poor prognosis equivalent to more classical GBM embedded sections stained with H&E, the tumor
has been reported in patients with this tumor. cells are closely packed and appear swollen,
consisting of a small, round nucleus (usually
2.2.1.1.1.4.4. Other GBM subtypes A GBM slightly larger than a normal oligodendrocyte)
with primitive neuronal component is a re- surrounded by a clear halo (Fig. 2.3A). This imparts
cently described variant of high-​ grade glioma a very characteristic “honeycomb,” “fried egg,” or
that contains regions of discrete nodules of “small “owl’s eye” appearance to the tumor. However,
blue cells” resembling medulloblastoma. By this pattern is not apparent in smear preparations
immunohistochemistry, these undifferentiated foci or frozen sections and is often absent in paraffin
express neuronal markers (e.g., synaptophysin), lack sections made from previously frozen material.
astrocytic markers (GFAP, vimentin), and harbor a The tumor stroma also characteristically contains a
very high Ki-​67/​MIB-​1 labeling index (greater than network of thin-​walled branching capillaries, often
30% and often above 80%). Homer Wright rosettes described as a “chicken-​wire” or “wishbone” vas-
or large cell/​anaplastic histology may be present in cular branching pattern (Fig. 2.3B). The presence
these areas. Patients with this variant may be more of occasional mitoses or nuclear atypia (which
likely to show dissemination of the tumor via the may be marked in some cases) is still compatible
CSF. A subset of these tumors has amplification of with the designation of the lesion as a WHO grade
MYCN or MYC. Others have mutations in IDH1 II neoplasm. Elevated mitotic activity, microvas-
and PDGFRA alterations, supporting that some of cular proliferation, or necrosis indicates anaplastic
these tumors are secondary GBMs. transformation (WHO grade III). In most WHO

Chapter 2 Tumors of the Central Nervous System • 29

 A B

C D

FIGURE 2.3 Microscopic appearance of oligodendroglioma: (A) oligodendroglioma with microcalcifications

(H&E); (B) “chicken-​wire” vascular pattern (H&E); (C) IDH1 R132 immunoreactivity in tumor
cells. Anaplastic oligodendroglioma: (D) microvascular proliferation (H&E); (E) nuclear atypia,
minigemistocytes (H&E).

grade II oligodendroglioma, the Ki-​67/​MIB-​1 la-
beling index is less than 10%. In some instances, the
tumor may contain nodules of increased cellularity;
careful attention to these areas may reveal other
anaplastic features within them. Small calcifications
(calcospherites) are a characteristic histologic
feature (Fig. 2.3A) but are only seen in 20% of
cases and are not specific for oligodendroglioma.
Another characteristic and diagnostically useful
feature is the presence of perineuronal, perivas-
cular, or subpial tumor aggregates when the tumor
infiltrates cortex. It is not uncommon to find cells
with well-​ differentiated GFAP-​positive astrocyte
morphology and with visible cytoplasm admixed

30 • E scou r o l l e and P oi r ie r ’ s M anua l o f B asic N eu r opat h o l ogy

Another random document with
no related content on Scribd:
at the Italian opera in Paris; but the only foundation for such a report
seems to be that it was not uncommon for violinist composers of that
period to enlist the aid of their friends in writing for the orchestra.
Viotti was a broadly educated musician, whose experience with
orchestras was wide.

Second in importance to the concertos are the duets for two violins
written during his stay in Hamburg. These are considered second in
musical charm only to Spohr’s pieces in the same manner. That
Viotti was somewhat low in spirit when he was at work on them,
exiled as he was from London and Paris, is shown by the few words
prefixed to one of the sets, ‘This work is fruit of the leisure which
misfortune has brought me. Some pieces came to me in grief, others
in hope.’

Viotti had a brilliant and unrestricted technique. He was among the

greatest of virtuosi. But little of this appears in his music. That is
distinguished by a dignity and a relative simplicity, well in keeping
with the noble traditions inherited from a country great in more ways
than one in the musical history of the eighteenth century. But as far
as form and style go he is modern. He undoubtedly owes something
to Haydn. Moreover, Wasielewski makes the point that there is no
trace in his music of the somewhat churchly dignity one feels in the
sonatas of Corelli and Tartini. Viotti’s is a thoroughly worldly style, in
melody and in the fiery but always musical passage work. He is at
once the last of the classic Italians and the first of the moderns,
standing between Corelli and Tartini on the one hand and Spohr,
David, and Vieuxtemps on the other.

The list of the men who came to him for instruction while he was in
Paris contains names that even today have an imposing ring. Most
prominent among them are Rode, Cartier, and Durand. And among
those who were not actually his pupils but who accepted him as their
ideal and modelled themselves after him were Rodolphe Kreutzer
and Pierre Baillot. These men are the very fountain head of most
violin music and playing of the nineteenth century. They set the
standard of excellence in style and technique by which Spohr and
later Vieuxtemps ruled themselves.

IV
Before considering their work, the development of violin music in
Germany during the eighteenth century must be noticed. The
influence of the Italians was not less strong here than in France.
Both Biber and Strungk had come under it in the late seventeenth
century, Strungk being, as we know, personally acquainted with
Corelli and at one time associating closely with him in Rome. The
German violinists of the eighteenth century either went to Italy to
study, or came under the influence of various Italians who passed
through the chief German cities on concert tours.

The most conspicuous of them are associated with courts or cities

here and there. For instance, early in the century there is Telemann
in Hamburg; a little later Pisendel in Dresden; J. G. Graun in Berlin;
Leopold Mozart in Salzburg; the gifted Stamitz and his associates
Richter, Cannabich and Fränzl in Mannheim; and the most amiable if
not the most gifted of all, Franz Benda, here and there in Bohemia,
Austria and Saxony. Though these and many more were widely
famous in their day as players, and Mozart was influential as a
teacher, little of their music has survived the centuries that have
passed since they wrote it. The eighteenth century was in violin
music and likewise in opera, the era of Italian supremacy; and in
violin music we meet with little except copies outside of Italy.

Georg Philipp Telemann, it is true, wrote that he followed the French

model in his music; but as Wasielewski says, this applies evidently
only to his vocal works and overtures, for his violin compositions are
very clearly imitations of Corelli’s. All his music, and he wrote
enormous quantities in various branches, is essentially
commonplace. Between 1708 and 1721 Telemann occupied a
position at the court of Eisenach. It was chiefly during these years
that he gave himself to the violin and violin music. Afterwards he
went to Hamburg and there worked until his death in 1767.

Johann Georg Pisendel is a far more distinguished figure. He was

born on the twenty-sixth of December, 1687, at Carlsburg in
Franconia, and died in Dresden, after many years’ service there, in
November, 1755. While still a boy the Marquis of Anspach attached
him to his chapel, on account of his beautiful voice. In the service of
the same prince at that time was Torelli, the great Italian composer
for the violin; and Pisendel was his pupil for a considerable period.
Later in life he was able to journey in Italy and France, and was
apparently at one time a pupil of Vivaldi’s in Venice. From 1728 to
the time of his death he was first violin in the royal opera house at
Dresden. His playing was distinguished by care in shading, and in
his conducting he was said to have laid great importance upon ‘loud
and soft.’ As a composer he is without significance, though some of
his works—concertos and sonatas—have been preserved. But his
influence served to educate violinists in that part of Germany, so that
little by little Germans came to supplant the Italians in that branch of
music, and to find occupation in connection with the opera house
orchestra, which had been up to that time almost entirely made up of
Italians.

Most conspicuous among those who were actually his pupils was
Johann Gottlieb Graun, brother of the still familiar Carl Heinrich. But
Graun was not content with instruction in Germany alone, and
betook himself to Tartini in Padua. After his return to his native land,
he eventually found his place at the court of Frederick the Great,
who was still crown prince. With him at this time were Quantz, the
flute player, and Franz Benda. After the accession of Frederick to the
throne of Prussia, Graun was made first violin and concert master in
the royal orchestra; and he held this place until his death in 1771.
His compositions, like all others for the violin at this period, are
hardly more than imitations of the Italian masterpieces. And like
Pisendel, his importance is in the improvement of the state of
instrumental music in Germany, and especially of the orchestra at
Berlin.
His successor in this royal orchestra was Franz Benda, who, not only
by reason of the romantic wanderings of his life, is one of the most
interesting figures in the history of music in Germany during the
eighteenth century. His father, Hans Georg, had been a sort of
wandering player, as well as a weaver; and his brothers, Johann,
Georg, and Joseph, were all musicians who won a high place in their
day. Georg was perhaps the most distinguished of the family, but in
the history of violin-music Franz occupies a more important place.

The Bendas were Bohemians, but most of them settled in Germany

and accepted German ideals and training. Franz Benda, after a
changing career as a boy singer in various places, finally came
under the influence of Graun and Quantz in the crown prince’s
orchestra, at Rheinsberg. The principal instruction he received upon
the violin came from Graun, who was himself a pupil of Tartini’s; so,
although Benda shows the marks of an independent and self-
sufficient development, not a little of Italian influence came close to
him. He remained in the service of the Prussian court from 1733,
when Quantz befriended him, until his death as an old man in 1786.

His playing was admired for its warm, singing quality, which showed
to such advantages in all slow movements that musicians would
come long distances to hear him play an adagio. Burney heard him
in 1772 and was impressed by the true feeling in his playing. Burney,
too, mentioned that in all Benda’s compositions for the violin there
were no passages which should not be played in a singing and
expressive manner. He went on to say that Benda’s playing was
distinguished in this quality from that of Tartini, Somis, and Veracini,
and that it was something all his own which he had acquired in his
early association with singers.

He had indeed been a great singer, and he gave up public singing

only because after singing he was subject to violent headache. He
trained his two daughters to be distinguished singers of the next
generation.

His works for the violin are numerous, but only a small part of them
was published, and this posthumously. In spite of the often lovely
melodies in the slow movements they have not been able to outlive
their own day. Wasielewski calls attention to the general use of
conventional arpeggio figures in the long movements, which,
characteristic of a great deal of contemporary music for the violin,
may have been written with the idea of offering good technical
exercise in the art of bowing.

Among Benda’s many pupils the two most significant are his own
son, Carl, and Friedrich Wilhelm Rust. The former seems to have
inherited a great part of his father’s skill and style. The sonatas of the
latter are among the best compositions written in Germany for the
violin in the second half of the eighteenth century. Rust died in
February, 1798. His name is remembered as much for his sonatas
for pianoforte as for his violin compositions. Another pupil, Carl
Haack, lived until September, 1819, and thus was able to carry the
Benda tradition over into the nineteenth century. On the whole Franz
Benda may be said to have founded a school of violin playing in
Berlin which has influenced the growth of music for that instrument in
Germany. Its chief characteristic was the care given to simplicity and
straightforwardness, especially in the playing of slow movements
and melodies, which stands out quite distinctly against the current of
more or less specious virtuosity running across the century.

Johann Peter Salomon (1745-1815) has been associated with the

Berlin group, though his youth was spent in and about Bonn, and his
greatest activity was displayed as an orchestral conductor in London.
It was he who engaged Haydn to come to London and to compose
symphonies specially for a London audience; and he occupies an
important place in the history of music in England as one of the
founders (1813) of the Philharmonic Society. He published but little
music, and that is without significance.

One of the outstanding figures in the history of violin music in

Germany is Leopold Mozart, the father of Wolfgang. He is hardly
important as a composer, though many of his works were fairly well
known in and about Salzburg where the greatest part of his life was
spent; but his instruction book on playing the violin marks the
beginning of a new epoch in his own country. This was first
published in Augsburg in 1756, was reprinted again in 1770, 1785,
and in Vienna in 1791 and 1804. It was for many years the only book
on the subject in Germany.

Much of it is now old-fashioned, but it still makes interesting reading,

partly because he was far-seeing enough to seize upon fundamental
principles that have remained unchanged in playing any instrument,
partly because the style is concise and the method clear, partly
because of the numerous examples it contains of both good and bad
music. Evidently his standard of excellence is Tartini, so that we still
find violin music in Germany strongly under the influence of the
Italians. But the great emphasis he lays upon simplicity and
expressiveness recalls Benda and his ideals, so that it would appear
that some wise men in Germany were at least shrewd enough to
choose only what was best in the Italian art. Among the many
interesting points he makes is that it takes a better-trained and a
more skillful violinist to play in an orchestra than to make a success
as a soloist. Evidently many of the German musicians distrusted the
virtuoso. Emanuel Bach, it will be remembered, cared nothing for
show music on the keyboard. C. F. D. Schubart, author of the words
of Schubert’s Die Forelle, said that an orchestra made up of virtuosi
was like a world of queens without a ruler. He had the orchestra at
Stuttgart in mind.

V
Meanwhile about the orchestra at Mannheim there was a band of
gifted young men whose importance in the development of the
symphony and other allied forms has been but recently recognized,
and now, it seems, can hardly be overestimated. The most
remarkable of these was J. C. Stamitz, a Bohemian born in 1719,
who died when less than forty years old. His great accomplishments
in the domains of orchestral music have been explained elsewhere
in this series. In the matter of violin music he can hardly be said to
show any unusual independence of the Italians, but in the meagre
accounts of his life there is enough to show that he was a great
violinist. He was the teacher of his two sons, Carl (1746-1801) and
Anton (b. 1753), the latter of whom apparently grew up in Paris,
where the father, by the way, had been well known at the house of
La Pouplinière. Anton, as we shall see, was the teacher of Rodolphe
Kreutzer, already mentioned as one of the great teachers at the
Paris Conservatory in the first of the nineteenth century.

Christian Cannabich, a disciple if not a pupil of Stamitz, was likewise

a famous violinist, but again like his master, was more influential in
what he accomplished with the famous orchestra at Mannheim than
in his playing or composing for the violin. He seems to have spent
some years in Naples to study with Jomelli, and the Italian influence
is evident in all he wrote for the violin. Wilhelm Cramer, the father of
the now more famous J. B. Cramer, was another violinist associated
with the Mannheim school, until in 1773 he went to London on the
advice of Christian Bach. Here he lost one place after another as
conductor, owing now to the arrival of Salomon, now to that of Viotti.
He died in 1799 in great poverty.

Others connected with the orchestra at Mannheim are Ignaz Fränzl,

whose pupil, F. W. Pixis, became the teacher of Kalliwoda and Laub,
and whose son Ferdinand (1770-1833) was a distinguished violinist
in the next century; and Johann Friedrich Eck (b. 1766) and his
brother Franz. Their father was, like Stamitz, a Bohemian. Indeed
Stamitz seems to have induced Eck the elder to leave Bohemia and
come to Mannheim. Franz Eck is most famous today as one of the
teachers of Ludwig Spohr.

In Vienna the Italian influence was supreme down to nearly the end
of the century. The first of the Viennese violinists to win an
international and a lasting renown was Karl Ditters von Dittersdorf (b.
1739), the friend of Haydn and Gluck. Though two of his teachers,
König and Ziegler, were Austrians, a third, who perfected him, was
an Italian, Trani. Through Trani Dittersdorf became familiar with the
works of Corelli, Tartini, and Ferrari, after which he formed his own
style. Practically the first German to draw a circle of pupils about him
was Anton Wranitzky (b. 1761). Among his pupils the most
distinguished was Ignaz Schuppanzigh, who, as the leader of the
Schuppanzigh quartet, won for himself an immortal fame, and really
set the model for most quartet playing throughout the nineteenth
century. He was the son of a professor at the Realschule in Vienna.
From boyhood he showed a zeal for music, at first making himself a
master of the viola. At the time Beethoven was studying counterpoint
with Albrechtsberger he was taking lessons on the viola with
Schuppanzigh. Later, however, Schuppanzigh gave up the viola for
the violin. His most distinguished work was as a quartet leader, but
he won fame as a solo player as well; and when the palace of Prince
Rasoumowsky was burned in 1815, he went off on a concert tour
through Germany, Poland and Russia which lasted many years. He
was a friend not only of Beethoven, but of Haydn, Mozart, and of
Schubert as well; and was the principal means of bringing the
quartet music of these masters to the knowledge of the Viennese
public. He died of paralysis, March 2, 1830. Among his pupils the
most famous was Mayseder, at one time a member of the quartet.

What is noteworthy about the German violinist-composers of the

eighteenth century is not so much the commonness with which they
submitted to the influence of the Italians, but the direction their art as
players took as soon as they began to show signs of a national
independence. Few were the match of the Italians or even the
French players in solo work. None was a phenomenal virtuoso. The
greatest were most successful as orchestral or quartet players; and
their most influential work was that done in connection with some
orchestra. This is most evident in the case of the Mannheim
composers. Both Stamitz and Cannabich were primarily conductors,
who had a special gift in organizing and developing the orchestra.
Their most significant compositions were their symphonies, in the
new style, in which they not only gave a strong impetus to the
development of symphonic forms, but brought about new effects in
the combination of wood-wind and brass instruments with the
strings. Leopold Mozart’s opinion that a man who could play well in
an orchestra was a better player and a better musician than he who
could make a success playing solos, is indicative of the purely
German idea of violin music during the century. And it cannot be
denied that great as Franz Benda and Johann Graun may have been
as players, they contributed little of lasting worth to the literature of
the violin, and made practically no advance in the art of playing it.
But both were great organizers and concert masters, and as such
left an indelible impression on the development of music, especially
orchestral music, in Germany.

VI
Before concluding this chapter and passing on to a discussion of the
development of violin music in the nineteenth century a few words
must be said of the compositions for the violin by those great
masters who were not first and foremost violinists. Among these,
four may claim our attention: Handel, Bach, Haydn, and Mozart.

Handel is not known to have given much time to the violin, but it is
said that when he chose to play on it, his tone was both strong and
beautiful. He wrote relatively little music for it. Twelve so-called solo
sonatas with figured bass (harpsichord or viol) were published in
1732 as opus 1. Of these only three are for the violin: the third, tenth,
and twelfth. The others are for flute. Apart from a few characteristic
violin figures, chiefly of the rocking variety, these solo sonatas might
very well do for clavier with equal effect. There is the sane, broad
mood in them all which one associates with Handel. In the edition of
Handel’s works by the German Handel Society, there are three
additional sonatas for violin—in D major, A major, and E major.
These seem to be of somewhat later origin than the others, but they
are in the same form, beginning with a slow movement, followed by
allegro, largo, and final allegro, as in most of the cyclical
compositions of that time. One cannot deny to these sonatas a
manly dignity and charm. They are in every way plausible as only
Handel knows how to be; yet they have neither the grace of Corelli,
nor the deep feeling of Bach. One may suspect them of being, like
the pieces for clavier, tossed off easily from his pen to make a little
money. What is remarkable is that sure as one might be of this, one
would yet pay to hear them.

There are besides these solo sonatas for violin or flute and figured
bass, nine sonatas for two violins, or violin and flute with figured
bass, and seven sonatas, opus 5, for two instruments, probably
intended for two violins.

Among the most remarkable of J. S. Bach’s compositions are the six

sonatas for violin without any accompaniment, written in Cöthen,
about 1720. These works remain, and probably always will remain,
unique in musical literature, not only because of their form, but
because of the profound beauty of the music in them. Just how much
of a violinist Bach himself was, no one knows. He was fond of
playing the viola in the court band at Cöthen. It can hardly be
pretended that these sonatas for violin alone are perfectly adapted to
the violin. They resemble in style the organ music which was truly
the whole foundation of Bach’s technique. In that same organ style,
he wrote for groups of instruments, for groups of voices, for clavier
and for all other combinations.

On the other hand no activity of Bach’s is more interesting, and

perhaps none is more significant, than his assiduous copying and
transcribing again and again of the violin works of Vivaldi, Torelli, and
Albinoni. Especially his study of Vivaldi is striking. He used themes
of the Italian violinists as themes for organ fugues; he transcribed the
concertos of Vivaldi into concertos for one, two, three, or four
harpsichords. And not only that, practically all his concertos for a
solo clavier are transcriptions of his own concertos for violin.

But the polyphonic style of the sonatas for violin alone is peculiarly a
German inheritance. Walter and Biber were conspicuous for the use
of double stops and an approach to polyphonic style. Most
remarkable of all was a pupil of the old Danish organist, Buxtehude,
Nikolaus Bruhns (1665-1697), who was able to play two parts on his
violin and at the same time add one or two more with his feet on the
organ pedals. Though Corelli touched gently upon the polyphonic
style in the movements of the first six of his solo sonatas, the
polyphonic style was maintained mostly by the Germans. As Bach
would write chorus, fugue, or concerto in this style, so did he write
for the violin alone.

Of the six works the first three are sonatas, in the sense of the
sonate da chiesa of Corelli, serious and not conspicuously
rhythmical. The last three are properly suites, for they consist of
dance movements. The most astonishing of all the pieces is the
Chaconne, at the end of the second suite. Here Bach has woven a
series of variations over a simple, yet beautiful, ground, which finds
an equal only in the great Passacaglia for the organ.

The three sonatas of this set can be found transcribed, at least in

part, by Bach into various other forms. The fugue from the first, in G
minor, was transposed into D minor and arranged for the organ. The
whole of the second sonata, in A minor, was rearranged for the
harpsichord. The fugue in the third sonata for violin alone exists also
as a fugue for the organ.

There are besides these sonatas for violin alone, six sonatas for
harpsichord and violin, which are among the most beautiful of his
compositions; and a sonata in E minor and a fugue in G minor for
violin with figured bass. It is interesting to note that the six sonatas
for harpsichord and violin differ from similar works by Corelli and by
Handel. Here there is no affair with the figured bass; but the part for
the harpsichord is elaborately constructed, and truly, from the point
of view of texture, more important than that for the violin.

Bach wrote at least five concertos for one or two violins during his
stay at Cöthen. One of these is included among the six concertos
dedicated to the Margrave of Brandenburg. All of these have been
rearranged for harpsichord, and apparently among the harpsichord
concertos there are three which were originally for violin but have not
survived in that shape. The concertos, even more than the sonatas,
are not essentially violin music, but are really organ music. The style
is constantly polyphonic and the violin solos hardly stand out
sufficiently to add a contrasting spot of color to the whole. Bach’s
great work for the violin was the set of six solo sonatas. These must
indeed be reckoned, wholly apart from the instrument, as among the
great masterpieces in the musical literature of the world.

Haydn’s compositions for violin, including concertos and sonatas,

are hardly of considerable importance. His associations with
violinists in the band at Esterhazy, and later in Vienna with amateurs
such as Tost and professionals like Schuppanzigh, gave him a
complete idea of the nature and the possibilities of the instrument.
But the knowledge so acquired shows to best advantage in his
treatment of the first and second violin parts in his string quartets, in
many of which the first violin is given almost the importance of a solo
instrument. Eight sonatas for harpsichord and violin have been
published, but of these only four were originally conceived in the
form.

The young Mozart was hardly less proficient on the violin than he
was on the harpsichord, a fact not surprising in view of his father’s
recognized skill as a teacher in this special branch of music. But he
seems to have treated his violin with indifference and after his
departure from Salzburg for Paris to have quite neglected his
practice, much to his father’s concern. The most important of his
compositions for the violin are the five concertos written in Salzburg
in 1775. They were probably written for his own use, but just how
closely in conjunction with the visit of the Archduke Maximilian to
Salzburg in April of that year cannot be stated positively. Several
serenades and the little opera, Il re pastore, were written for the fêtes
given in honor of the same young prince. The concertos belong to
the same period. In Köchel’s Index they are numbers 207, 211, 216,
218, and 219. A sixth, belonging to a somewhat later date, bears the
number 268. Of these the first in B-flat was completed on April 14,
1775, the second, in D, June 14, the third, in G, September 12, the
fourth, in D, in October, and the fifth, in A, quite at the end of the
year.

The sixth concerto, in E-flat, is considered both by Jahn and Köchel

to belong to the Salzburg period. It was not published, however, until
long after Mozart’s death; and recently the scholarly writers, Messrs.
de Wyzewa and de St. Foix, have thrown considerable doubt upon
the authenticity of large parts of it. According to their theory[50] the
opening tutti and the orchestral portion at the beginning of the
development section are undoubtedly the work of Mozart, but of the
mature Mozart of 1783 and 1784. Likewise the solo passages in all
the movements seem to bear the stamp of his genius. But apart from
these measures, the development of the solo ideas and the
orchestral accompaniment were completed either by André, who
published the work, or by Süssmayer, who was also said by Mozart’s
widow to be the composer of a mass in B-flat, published by C. F.
Peters as a composition of Mozart’s.

In addition mention should be made of the concertos introduced

between the first and second movements of various serenades,
according to the custom of the day. Most of these are of small
proportions; but one, in G major (K. 250), written in Salzburg some
time in July, 1776, has the plan of an independent composition.

It was the custom for a master like Schobert in Paris, or Mozart in

Vienna, to ‘accompany’ the young ladies who played pianoforte or
harpsichord sonatas of his composition and under his instruction with
music on the violin. There are many sonatas for harpsichord
published by Schobert, with a violin part ad libitum. This in the main
but reinforces the chief melodic lines of the part for harpsichord or
pianoforte; and works with such a violin part, ad libitum, are not at all
violin sonatas in the sense of the term accepted today, i.e., sonatas
in which violin and piano are woven inextricably together. They are
frankly pianoforte or harpsichord sonatas with the ‘accompaniment’
of a violin.

On the other hand, we have found the violin masters like Corelli and
Tartini writing sonatas for violin, with figured bass for harpsichord,
lute, or even viol. Such sonatas were often called solo sonatas, as in
the case of those of Handel, recently mentioned. The accompanying
instrument had no function but to add harmonies, and a touch of
imitation in the written bass part, here and there.
Between these two extremes lies the sonata with harpsichord
obbligato, that is to say, with a harpsichord part which was not an
accompaniment but an essential part of the whole. In these cases
the music was generally polyphonic in character. The violin might
carry one or two parts of the music, the harpsichord two or three.
Very frequently, if the instruments played together no more than
three parts, the composition was called a Trio. The sonatas by J. S.
Bach for harpsichord and violin are of this character. Though the
harpsichord carries on more of the music than the violin, both
instruments are necessary to the complete rendering of the music.

Mozart must have frequently added improvised parts for the violin to
many of his sonatas written expressly for the keyboard instrument.
Among his earliest works one finds sonatas for clavecin with a free
part for violin, for violin or flute, for violin or flute and 'cello. Oftenest
the added part does little more than duplicate the melody of the part
for clavecin, with here and there an imitation or a progression of
thirds or sixths. But among his later works are sonatas for pianoforte
with added accompaniment for violin in which the two instruments
contribute something like an equal share to the music, which are the
ancestors of the sonatas for violin and piano by Beethoven, Brahms,
and César Franck. Among the most important of these are six
published in November, 1781, as opus 2. In Köchel’s Index they bear
the numbers 376, 296, 377, 378, 379, and 380. The greatest of them
is that in C major, K. 296, with its serious and rich opening adagio, its
first allegro in Mozart’s favorite G minor, and the beautiful variations
forming the last movement. Four more sonatas, of equal musical
value, were published respectively in 1784, 1785, 1787, and 1788.

VII
Looking back over the eighteenth century one cannot but be
impressed by the independent growth of violin music. The Italians
contributed far more than all the other nationalities to this steady
growth, partly because of their native love for melody and for sheer,
simple beauty of sound. The intellectual broadening of forms, the
intensifying of emotional expressiveness by means of rich and
poignant harmonies, concerned them far less than the perfecting of a
suave and wholly beautiful style which might give to the most singing
of all instruments a chance to reveal its precious and almost unique
qualities. This accounts for the calm, classic beauty of their music,
which especially in the case of Corelli and Tartini does not suffer by
changes that have since come in style and the technique of
structure.

The success of the Italian violinists in every court of Europe, both as

performers and as composers, was second only to the success of
the great singers and the popular opera composers of the day. Their
progress in their art was so steadfast and secure that other nations
could hardly do more than follow their example. Hence in France and
Germany one finds with few exceptions an imitation of Italian styles
and forms, with a slight admixture of national characteristics, as in
the piquancy of Cartier’s, the warm sentiment of Benda’s music.
What one might call the pure art of violin playing and violin music,
abstract in a large measure from all other branches of music, was
developed to perfection by the Italian violinist-composers of the
eighteenth century. Its noble traditions were brought over into more
modern forms by Viotti, henceforth to blend and undergo change in a
more general course of development.

Perhaps only in the case of Chopin can one point to such a pure and
in a sense isolated ideal in the development of music for a single
instrument, unless the organ works of Bach offer another exception.
And already in the course of the eighteenth century one finds here
and there violin music that has more than a special significance. The
sonatas for unaccompanied violin by Bach must be regarded first as
music, then as music for the violin. The style in which they were
written is not a style which has grown out of the nature of the
instrument. They have not served and perhaps cannot serve as a
model for perfect adaptation of means to an end. Bach himself was
willing to regard the ideas in them as fit for expression through other
instruments. But the works of Corelli, Tartini, Nardini and Viotti are
works which no other instrument than that for which they were
written may pretend to present. And so beautiful is the line of melody
in them, so warm the tones which they call upon, that there is
scarcely need of even the harmonies of the figured bass to make
them complete.

In turning to the nineteenth century we shall find little or no more of

this sort of pure music. Apart from a few brilliant concert or salon
pieces which have little beyond brilliance or charm to recommend
them, the considerable literature for the violin consists of sonatas
and concertos in which the accompaniment is like the traditional half,
almost greater than the whole. In other words we have no longer to
do with music for which the violin is the supreme justification, but
with music which represents a combination of the violin with other
instruments. Glorious and unmatched as is its contribution in this
combination, it remains incomplete of itself.
 FOOTNOTES:
[48] See A. Schering: Geschichte des Instrumentalkonzerts.

[49] Die Violine und ihre Meister.

[50] See ‘W. A. Mozart,’ by T. de Wyzewa and G. de St. Foix, Paris, 1912.
Appendix II, Vol. II, p. 428.
 CHAPTER XIII
VIOLIN MUSIC IN THE NINETEENTH
CENTURY
The perfection of the bow and of the classical technique—The
French school: Kreutzer, Rode, and Baillot—Paganini: his
predecessors, his life and fame, his playing, and his compositions
—Ludwig Spohr: his style and his compositions; his pupils—
Viennese violinists: Franz Clement, Mayseder, Boehm, Ernst and
others—The Belgian school: De Bériot and Vieuxtemps—Other
violinist composers: Wieniawski, Molique, Joachim, Sarasate, Ole
Bull; music of the violinist-composers in general—Violin music of
the great masters.

The art of violin music in the nineteenth century had its head in
Paris. Few violinists with the exception of Paganini developed their
powers without the model set them by the great French violinists at
the beginning of the century. Most of them owed more than can be
determined to the influence of Viotti. Even Spohr, who with more or
less controversial spirit, wrote of the French violinists as old-
fashioned, modelled himself pretty closely upon Rode; and therefore
even Spohr is but a descendant of the old classical Italian school.

The technique of playing the violin was thoroughly understood by the

end of the eighteenth century. Viotti himself was a brilliant virtuoso;
but, trained in the classic style, he laid less emphasis upon external
brilliance than upon expressiveness. The matters of double stops,
trills, runs, skips and other such effects of dexterity were largely
dependent upon the fingers of the left hand; and this part of
technique, though somewhat hampered by holding the violin with the
chin upon the right side of the tailpiece, was clearly mastered within
reasonable limits by the violinists of the middle of the century, Tartini,
Veracini, Nardini, Geminiani, and others. Indeed Geminiani in his
instruction book recommended that the violin be held on the left side;
and in range of fingering gave directions for playing as high as in the
seventh position. Leopold Mozart, however, naturally conservative,
held to the old-fashioned holding of the instrument.

The technique of bowing, upon which depends the art of expression

in violin playing, awaited the perfection of a satisfactory bow. Tartini’s
playing, it will be remembered, was especially admired for its
expressiveness; and this, together with certain of his remarks on
bowing which have been preserved in letters, leads one to think that
he may have had a bow far better than those in the hands of most of
his contemporaries. Whether or not he made it himself, and indeed
just what it may have been, are not known. Certainly it must have
been better than the bows with which Leopold Mozart was familiar.
The clumsy nature of these may be judged by the illustrations in his
instruction book.

The final perfection of the bow awaited the skill of a Frenchman,

François Tourte (1747-1835), who has properly been called the
Stradivari of the bow. It was wholly owing to his improvements that
many modern effects in staccato, as well as in fine shading,
particularly in the upper notes, became possible. He is supposed not
to have hit upon these epoch-making innovations until after 1775;
and there is much likelihood that he was stimulated by the presence
of Viotti in Paris after 1782. No better testimony to the service he
rendered to the art of violin playing can be found than the new
broadening of violin technique and style accomplished by men like
Viotti, Kreutzer, Baillot, Rode, and Lafont, who availed themselves
immediately of the results of his skill.

I
Something may now be said of these men, whose activities have
without exception the glaring background of the horrors of the
French Revolution. Though Kreutzer was of German descent, he
was born in Versailles (1766) and spent the greater part of his life in
and about Paris, intimately associated with French styles and
institutions. Apart from early lessons received from his father, he
seems to have been for a time under the care of Anton Stamitz, son
of Johann Stamitz. At the Chapelle du Roi, to which organization he
obtained admittance through the influence of Marie Antoinette, he
had the occasion of hearing Viotti. The great Italian influenced him
no less than he influenced his young contemporaries in Paris.
Concerning his activities as a composer of operas little need be said,
though one or two of his ballets, especially Paul et Virginie and Le
Carnaval de Venise, held the stage for some years. As a player he
ranks among the most famous of the era. His duets with Rode
roused the public to great enthusiasm. In 1798 he was in Vienna in
the suite of General Bernadotte, and here made the acquaintance of
Beethoven. Subsequently Beethoven dedicated the sonata for violin
and piano (opus 47) to Kreutzer.

By reason of this and his book of forty Études ou caprices pour le

violon, he is now chiefly remembered. His other compositions for the
violin, including nineteen concertos and several airs and variations,
have now been allowed to sink into oblivion. To say that the
concertos are ‘more brilliant than Rode’s, less modern than Baillot’s’
distinguishes them as much as they may be distinguished from the
compositions of his contemporaries. They are dry music, good as
practice pieces for the student, but without musical life. But Kreutzer
was a great teacher. He was one of the original professors of the
violin at the Conservatoire, and with Baillot and Rode prepared the
still famous Méthode which, carrying the authority of that sterling
institution, has remained, almost to the present day, the standard
book of instruction for the young violinist. His own collection of forty
studies likewise holds still a place high among those ‘steps to
Parnassus’ by which the student may climb to the company of
finished artists.