Gliomas

Visit the link below to download the full version of this book:

https://medidownload.com/product/gliomas/

Click Download Now

vi AVAILABLE TITLES (Continued)
Vol. 124, Clinical neuroendocrinology, E. Fliers, M. Korbonits and J.A. Romijn, eds. ISBN 9780444596024
Vol. 125, Alcohol and the nervous system, E.V. Sullivan and A. Pfefferbaum, eds. ISBN 9780444626196
Vol. 126, Diabetes and the nervous system, D.W. Zochodne and R.A. Malik, eds. ISBN 9780444534804
Vol. 127, Traumatic brain injury Part I, J. Grafman and A.M. Salazar, eds. ISBN 9780444528926
Vol. 128, Traumatic brain injury Part II, J. Grafman and A.M. Salazar, eds. ISBN 9780444635211
Vol. 129, The human auditory system: Fundamental organization and clinical disorders, G.G. Celesia
and G. Hickok, eds. ISBN 9780444626301
Vol. 130, Neurology of sexual and bladder disorders, D.B. Vodušek and F. Boller, eds. ISBN 9780444632470
Vol. 131, Occupational neurology, M. Lotti and M.L. Bleecker, eds. ISBN 9780444626271
Vol. 132, Neurocutaneous syndromes, M.P. Islam and E.S. Roach, eds. ISBN 9780444627025
Vol. 133, Autoimmune neurology, S.J. Pittock and A. Vincent, eds. ISBN 9780444634320
Foreword

This volume of the Handbook of Clinical Neurology provides a comprehensive account of gliomas in 25 chapters
contributed by some of the leading authorities in the field. An initial section is made up of 11 chapters on diagnosis
and treatment and includes chapters on epidemiology, clinical and imaging features, surgical treatment, histologic
and molecular classification, radiation and pharmacologic therapy, immunotherapy, and experimental treatments.
The advance of knowledge on all fronts is exciting, and the emerging gene therapies and new treatments with oncolytic
viruses cannot fail to be of interest and spark the hopes of all who care for patients with these tumors. The second
section is made up of eight chapters that focus on the complications arising in patients with gliomas as a result of the
tumor or its treatment, and on symptom management, including a discussion of rehabilitation, palliative care, and
quality-of-life issues. The final section (six chapters) contains chapters on specific types of gliomas.
The volume thus provides a seamless and comprehensive account that will be an essential compendium for all
healthcare providers who manage patients with gliomas as well as for neuroscientists who are attempting to advance
knowledge about primary brain tumors. It provides a detailed account of their diagnosis and clinical management
according to the highest contemporary standards. It also points the way to the future, by summarizing the newest
fundamental studies that extend our understanding of their molecular biology. For example, new molecular techniques
may permit the biologic diversity of these tumors to be better recognized and allow better matching of treatment to
individual tumor type. Again, increasing insight into the molecular biology of gliomas, and improvements in genetic
engineering techniques, are facilitating laboratory-based approaches to developing novel antiglioma therapies. New
viral therapies, involving nonreplicative viral vectors and replicating oncolytic viruses, offer exciting prospects for the
future. Novel drug carrier systems and nanoparticles for targeting brain tumors are under development.
The encyclopedic coverage provided in the present volume has been guided by the experience and skills of the two
volume editors. Mitchel S. Berger is Professor and Chairman of the Department of Neurological Surgery at the Uni-
versity of California San Francisco, where he also directs the Brain Tumor Surgery Program and the Brain Tumor
Research Center. Michael Weller is Chairman of the Department of Neurology at the University Hospital Zurich
in Switzerland and has received several awards in recognition of his contributions to cancer research, including
the German Cancer Award in 2007. Both are thus internationally recognized experts in the field. They have gathered
together a group of outstanding contributors and ensured that the various chapters complement each other to provide
an important synthesis of the literature. We are grateful to them and to all the contributors.
We have reviewed and commented on each of the chapters in our capacity as series editors and believe that both
clinicians and basic scientists will find this volume especially appealing. The availability of the volume electronically on
Elsevier’s Science Direct site should ensure its ready accessibility and facilitate searches for specific information.
As always, it is a pleasure to thank Elsevier, our publishers – and in particular Michael Parkinson in Scotland and
Mara Conner and Kristi Anderson in California – for their assistance in the development and production of this
volume.
Michael J. Aminoff
François Boller
Dick F. Swaab
Preface

Gliomas are the most common intrinsic brain tumor. They can occur in any age group and at any location of the central
nervous system, and cause major morbidity and mortality. Glioblastoma, the most common variant of glioma, is
among the most lethal cancers. Advances in neurosurgical techniques, increased safety and tolerability of radiother-
apy, and advanced neuroimaging have improved care and outcome on a population level in recent years. Although
therapeutic measures for most gliomas are still not curative, major advances in the understanding of the pathogenesis
and molecular genetic alterations characterizing gliomas have had a major impact on the development of novel exper-
imental treatments.
We have joined forces in co-editing the present volume of the Handbook of Clinical Neurology as representatives
of the neurosurgical and neuro-oncologic approaches to gliomas. Our goal is to provide an update on all relevant areas
of the disease, thus providing a comprehensive reference source.
All chapters are written by international leaders in the field, partnering experts from North America and Europe
across the entire volume, often co-authored by junior members of our research community. The general chapters on
diagnosis and treatment of gliomas in part I of the volume are followed in part II by a series of general chapters addres-
sing complications and symptomatic management. Finally, the third part of this volume includes disease-specific chap-
ters on the various types of glioma. We are grateful to all authors for devoting their time and effort to make this a
unique source of knowledge in one of the most active research fields in neuro-oncology.

M.S. Berger
M. Weller
Contributors

M. Ahluwalia J.E.C. Bromberg

Brain Tumor and Neuro-Oncology Center, Cleveland Neuro-Oncology Unit, The Brain Tumor Center
Clinic, Cleveland, OH, USA at Erasmus MC Cancer Center, Rotterdam,
The Netherlands
K. Aldape
Department of Pathology, University Health Network J. Buckner
and Princess Margaret Cancer Centre, Toronto, Ontario, Department of Oncology, Mayo Clinic, Rochester, MN,
Canada USA

A. Alentorn E.F. Chang

Department of Neurology 2, Groupe Hospitalier UCSF Comprehensive Epilepsy Center and
Pitié-Salp^etriere and Inserm Experimental Neuro- Department of Neurological Surgery, University
Oncology Laboratory, Institut du Cerveau et de of California San Francisco, San Francisco,
la Moelle, Université Pierre et Marie Curie, California, USA
Paris, France
S. Chang
T.S. Armstrong Department of Neurological Surgery, University of
Department of Family Health, School of Nursing, California San Francisco, San Francisco, CA, USA
University of Texas Health Science Center, Houston,
TX, USA E.A. Chiocca
Harvey Cushing Neuro-Oncology Laboratories,
I.J. Barani
Department of Neurosurgery, Brigham and
Department of Radiation Oncology, University of
Women’s Hospital, Harvard Medical School, Boston,
California San Francisco, San Francisco, CA, USA
MA, USA
T. Batchelor
C-F. Cho
Department of Neurology, Harvard Medical School and
Harvey Cushing Neuro-Oncology Laboratories,
Catherine Pappas Center for Neuro-Oncology,
Department of Neurosurgery, Brigham and Women’s
Massachusetts General Hospital, Boston, MA, USA
Hospital, Harvard Medical School, Boston, MA, USA
M.S. Berger
Department of Neurological Surgery, University F. Dhermain
of California, San Francisco, CA, USA Radiation Oncology Department, Gustave Roussy
University Hospital, Villejuif, France
R. Bjerkvig
NorLux Neuro-Oncology Laboratory, Department of I. Djoukhadar
Oncology, Luxembourg Institute of Health, Wolfson Molecular Imaging Centre, University of
Luxembourg and Department of Biomedicine, Manchester, Manchester, UK
University of Bergen, Norway
C. Dorfer
M. Bornhorst Division of Neurosurgery, Department of Surgery,
Gilbert Family Neurofibromatosis Institute and Brain Hospital for Sick Children, University of Toronto,
Tumor Institute, Children’s National Health System, Toronto, Canada and Department of Neurosurgery,
Washington, DC, USA Medical University of Vienna, Vienna, Austria
xii CONTRIBUTORS
D.J. Englot E. Le Rhun
UCSF Comprehensive Epilepsy Center and Department Neuro-oncology, Department of Neurosurgery,
of Neurological Surgery, University of California University Hospital and Breast Unit, Department
San Francisco, San Francisco, California, USA of Medical Oncology, Oscar Lambret Center, Lille,
France
G. Finocchiaro
Department of Neuro-oncology, IRCCS Istituto B.D. Liebelt
Neurologico Besta, Milan, Italy Department of Neurosurgery, University of Texas MD
Anderson Cancer Center and Houston Methodist
D. Frappaz
Neurological Institute, Houston, TX, USA
Department of Pediatric and Adult Neuro-oncology,
Centre Léon Bérard and Institute of Pediatric
Hematology and Oncology, Lyon, France K. Masui
Department of Pathology, Tokyo Women’s Medical
E. Galanis University, Shinjku-ku, Tokyo, Japan and Ludwig
Departments of Oncology and Molecular Medicine, Institute for Cancer Research, University of California
Mayo Clinic, Rochester, MN, USA San Diego, La Jolla, CA, USA

W. Grisold M. Mehta
Department of Neurology, Kaiser-Franz-Josef Hospital Department of Radiation Oncology, Marlene
and Medical University of Vienna, Vienna, Austria and Stewart Greenebaum Cancer Center,
University of Maryland Medical Center, Baltimore,
A.B. Heimberger MD, USA
Department of Neurosurgery, University of Texas MD
Anderson Cancer Center, Houston, TX, USA T. Mikkelsen
Hermelin Brain Tumor Center and Departments of
R. Henriksson
Neurology and Neurosurgery, Henry Ford Hospital and
Regional Cancer Centre, Stockholm Gotland,
Wayne State University, Detroit, MI, USA
Stockholm, Sweden

S. Hervey-Jumper P.S. Mischel

Department of Neurological Surgery, Taubman Health Ludwig Institute for Cancer Research, University of
Center, Ann Arbor, MI, USA California San Diego, La Jolla, CA, USA

K. Hoang-Xuan L.B. Nabors

Department of Neurology 2, Groupe Hospitalier Pitié- Department of Neurology, University of Alabama at
Salp^etriere and Inserm Experimental Neuro-Oncology Birmingham, Birmingham, AL, USA
Laboratory, Institut du Cerveau et de la Moelle,
Université Pierre et Marie Curie, Paris, France H. Ohgaki
Section of Molecular Pathology, International Agency
M.M. Hulou for Research on Cancer, Lyon, France
Harvey Cushing Neuro-Oncology Laboratories,
Department of Neurosurgery, Brigham and Women’s R.J. Packer
Hospital, Harvard Medical School, Boston, MA, USA Gilbert Family Neurofibromatosis Institute and Brain
Tumor Institute, Children’s National Health System,
A. Jackson
Washington, DC, USA
Wolfson Molecular Imaging Centre, University of
Manchester, Manchester, UK
A. Perry
C. Jackson Departments of Pathology and Neurological Surgery,
Department of Neurosurgery, Johns Hopkins Hospital, University of California San Francisco, San Francisco,
Baltimore, MD, USA CA, USA

J.A.F. Koekkoek J.R. Perry

Department of Neurology, Leiden University Medical Division of Neurology, Odette Cancer Centre and
Center and Department of Neurology, Medical Center Sunnybrook Health Science Centre, University of
Haaglanden, The Hague, The Netherlands Toronto, Toronto, Canada
CONTRIBUTORS xiii
S.M. Pfister R. Soffietti
Division of Pediatric Neurooncology, German Cancer Department of Neuro-Oncology, University and City of
Research Center and Department of Hematology and Health and Science Hospital, Turin, Italy
Oncology, Heidelberg University Hospital, Heidelberg,
Germany B. Surboeck
Department of Neurology, Kaiser-Franz-Josef Hospital,
W.B. Pope Vienna, Austria
Radiological Sciences, Ronald Reagan Medical Center,
Los Angeles, CA, USA M.J.B. Taphoorn
Department of Neurology, Leiden University
A. Quiñones-Hinojosa Medical Center and Department of Neurology,
Department of Neurosurgery, Johns Hopkins Hospital, Medical Center Haaglanden, The Hague,
Baltimore, MD, USA The Netherlands

D. Reardon J. Tonn
Center for Neuro-Oncology, Harvard Medical Department of Neurosurgery, University Clinic of
School and Dana-Farber Cancer Institute, Ludwig-Maximilians-University Munich-Großhadern,
Boston, USA Munich, Germany

G. Reifenberger M.J. van den Bent

Department of Neuropathology, Heinrich Heine Neuro-Oncology Unit, The Brain Tumor Center
University, Düsseldorf, Germany at Erasmus MC Cancer Center, Rotterdam,
The Netherlands
J.C. Reijneveld
Department of Neurology, VU University Medical M. Vargo
Center, Amsterdam, The Netherlands Department of Physical Medicine and Rehabilitation,
Case Western Reserve University, MetroHealth
Medical Center, Cleveland, OH, USA
J. Rodon
Vall D’Hebron Institute of Oncology and Universitat
Autonoma de Barcelona, Barcelona, Spain C.J. Vecht
Service Neurologie Mazarin, Groupe Hospitalier
Pitié-Salp^etriere, Paris, France
R. Rudà
Department of Neuro-Oncology, University and City of
K.M. Walsh
Health and Science Hospital, Turin, Italy
Division of Neuroepidemiology, Department of
Neurological Surgery, University of California
J.T. Rutka San Francisco and UCSF Helen Diller Family Cancer
Division of Neurosurgery, Department of Surgery, Center, San Francisco, CA, USA
Hospital for Sick Children, University of Toronto,
Toronto, Canada C. Watts
Department of Clinical Neurosciences, Division of
P. Salander Neurosurgery, University of Cambridge, Cambridge, UK
Department of Social Work, Umeå University, Umeå,
Sweden M. Weller
Department of Neurology, University Hospital Zurich,
N. Sanai Zurich, Switzerland
Barrow Brain Tumor Research Center, Barrow
Neurological Institute, Phoenix, AZ, USA P. Wesseling
Department of Pathology, VU University
J.W. Snider, III Medical Center, Amsterdam, Department of
Department of Radiation Oncology, Marlene and Pathology, Canisius Wilhelmina Hospital,
Stewart Greenebaum Cancer Center, University Nijmegen, and Department of Pathology,
of Maryland Medical Center, Baltimore, Radboud University Medical Center, Nijmegen,
MD, USA The Netherlands
xiv CONTRIBUTORS
M. Westphal H-G. Wirsching
Department of Neurosurgery, University Department of Neurology, University Hospital Zurich,
Medical Center Hamburg-Eppendorf, Hamburg, Zurich, Switzerland
Germany
M.R. Wrensch
Division of Neuroepidemiology, Department of
W. Wick Neurological Surgery, University of California
Department of Neurooncology, University Clinic of San Francisco and UCSF Helen Diller Family Cancer
Heidelberg, Heidelberg, Germany Center, San Francisco, CA, USA
Handbook of Clinical Neurology, Vol. 134 (3rd series)
Gliomas
M.S. Berger and M. Weller, Editors
© 2016 Elsevier B.V. All rights reserved

Chapter 1

Epidemiology
KYLE M. WALSH1, HIROKO OHGAKI2, AND MARGARET R. WRENSCH1*
1
Division of Neuroepidemiology, Department of Neurological Surgery, University of California San Francisco
and UCSF Helen Diller Family Cancer Center, San Francisco, CA, USA
2
Section of Molecular Pathology, International Agency for Research on Cancer, Lyon, France

Abstract
More than 250 000 new cases of primary malignant brain tumors are diagnosed annually worldwide, 77%
of which are gliomas. A small proportion of gliomas are caused by the inheritance of rare high-penetrance
genetic variants or high-dose radiation. Since 2009, inherited genetic variants in 10 regions near eight dif-
ferent genes have been consistently associated with glioma risk via genome-wide association studies. Most
of these variants increase glioma risk by 20–40%, but two have higher relative risks. One on chromosome
8 increases risk of IDH-mutated gliomas sixfold and another that affects TP53 function confers a 2.5-fold
increased risk of glioma. Functions of some of the other risk variants are known or suspected, but future
research will determine functions of other risk loci. Recent progress also has been made in defining sub-
groups of glioma based on acquired alterations within tumors. Allergy history has been consistently asso-
ciated with reduced glioma risk, though the mechanisms have not yet been clarified. Future studies will
need to be large enough so that environmental and constitutive genetic risk factors can be examined within
molecularly defined, etiologically homogeneous subgroups.

DESCRIPTIVE EPIDEMIOLOGY1 USA more than 19 000 new cases of glioma are diag-
nosed each year with an age-adjusted average annual
Glioma incidence by age and histologic incidence rate of 6.24 per 100 000 population (Ostrom
subtype et al., 2014b). Brain tumor incidence rates have increased
Primary malignant brain tumors are the 17th most com- over the past three decades, with improved reporting,
mon cancer type worldwide, with more than 250 000 new increased use of diagnostic imaging, and changing atti-
cases diagnosed annually (Forman et al., 2013). Approx- tudes toward diagnosis in the elderly suspected to
imately 77% of these are gliomas, which include pilocytic account for much of this observed increase (Ohgaki
astrocytoma (World Health Organization (WHO) grade and Kleihues, 2005). Although primary brain tumors
I), diffuse astrocytoma (WHO grade II), anaplastic are relatively rare compared with metastatic brain
astrocytoma (WHO grade III), glioblastoma (WHO tumors or more common primary cancer sites such as
grade IV), oligodendroglioma (WHO grade II), anaplas- lung, breast, prostate, and colorectal, they constitute
tic oligodendroglioma (WHO grade III), ependymomas an important source of morbidity and mortality. In chil-
(WHO grade II), anaplastic ependymoma (WHO grade dren, brain tumors cause one-quarter of all cancer deaths
III), and mixed gliomas (Louis et al., 2007). In the (Ostrom et al., 2015a).

1
Abbreviations used in the chapter are listed at the end of the chapter before the References section.

*Correspondence to: Margaret R. Wrensch, PhD, Department of Neurological Surgery, University of California, San Francisco,
1450 Third Street, Room HD273, San Francisco CA 94158, USA. Tel: +1-415-476-1979, E-mail: [email protected]
4 K.M. WALSH ET AL.

Fig. 1.2. Incidence of common glioma histologies by age at

Fig. 1.1. Percentages of glioma by histologic type in children diagnosis, in children (A) and adults (B). (Data from
(A) and adults (B). NOS, not otherwise specified. (Data from Ostrom et al., 2014b, Tables 10 and 15, are age-adjusted to
Ostrom et al., 2014b, Table 19.) the 2000 US standard population.)

Differences in glioma incidence by sex,

The histologic distribution of glioma is very different
ethnicity, and geographic location
in adults versus that observed in children (Fig. 1.1).
Pilocytic astrocytoma (WHO grade I), the most common Men have higher incidences rates of glioma, embryonal
glioma in children, accounts for 37% of all glioma diag- tumors, germ cell tumors, and primary central nervous
noses among those under age 20 (Fig. 1.1A). Glioblas- system (CNS) lymphoma, whereas women have higher
toma (WHO grade IV), the most common glioma in incidence rates of meningioma and pituitary tumors.
adults, accounts for 67% of all adult glioma diagnoses The 1.3-fold increased risk of glioma in males versus
(Fig. 1.1B) (Ostrom et al., 2014b). females is among the most consistent findings in brain
The incidence rates of both pilocytic astrocytoma and tumor epidemiology (Ostrom et al., 2014b, 2015a).
ependymoma decrease throughout childhood and into Because of the consistency of this finding, even in pedi-
adolescence (Fig. 1.2A). In adults, overall incidence of atric populations, a comprehensive understanding of gli-
glioma dramatically increases with advancing age. How- oma etiology must account for this observation.
ever, the incidence of oligodendroglioma and ependy- However, this important epidemiologic observation
moma peaks in middle age and there is a decline in the remains unexplained.
incidence of glioblastoma among those 85 years and Interpretations of ethnic and geographic variation in
older (Fig. 1.2B) (Ostrom et al., 2014b). As with other can- the occurrence of brain tumors are complicated by prob-
cers, the increased incidence of glioma with age could be lems in ascertainment and reporting. Regions with the
due to the length of time required for malignant trans- highest reported rates of primary malignant brain
formation, the necessity of many genetic alterations tumors (e.g., Northern Europe, US white population,
prior to the onset of clinical disease, and/or diminished and Israel: rates of 11–20 per 100 000 population) gener-
immune surveillance. ally have better access to medical imaging than areas
EPIDEMIOLOGY 5

Fig. 1.4. Estimated distribution of five molecular groups of

glioma based on presence or absence of TERT promoter muta-
tion, IDH1/2 mutation, and 1p19q co-deletion among adults
diagnosed with glioma annually in the USA.

groups expected among incident infiltrating grade II–

Fig. 1.3. Primary glioma and meningioma incidence, by eth-
IV adult gliomas, estimated at the population level
nicity. (Data from Ostrom et al., 2014b, Table 4, are age-
adjusted to the 2000 US standard population.) (Rice et al., 2015). Further work will help to establish
the precise distribution and utility of these and other
molecular classifications currently being considered.
with the lowest rates (e.g., India and the Philippines: rates
of 2–4 per 100 000 people) (Inskip et al., 1995). However,
some of the variation suggests ethnic differences in INHERITED RISK FACTORS FORGLIOMA
inherited susceptibility and/or cultural or geographic dif- Inherited genetic predisposition to glioma has long been
ferences in risk factors (Jacobs et al., 2012; Dubrow suspected because of increased familial risk and the exis-
et al., 2013). Most notably, the rate of malignant brain tence of glioma in rare familial cancer syndromes
tumors in Japan, an economically prosperous country, (Malmer et al., 2007). Below, we briefly review both
is less than half the rate in Northern Europe (Forman the hereditary syndromes and the more common inher-
et al., 2013). Furthermore, in the USA, whites have higher ited variants associated with increased risk of glioma.
rates of glioma than African Americans but lower rates
of meningioma (Fig. 1.3) (Ostrom et al., 2014b). These
observations would be difficult to attribute solely to dif-
Hereditary cancer syndromes associated with
ferences in access to medical care or diagnostic
glioma risk
practices. Gliomas are thought to arise through the progressive
The absolute variation in the occurrence of brain accumulation of genetic and epigenetic alterations
tumors between high-risk and low-risk areas is on the that permit cells to evade normal regulatory mechanisms
order of fourfold, compared with the 20-fold difference and escape destruction by the immune system (Wrensch
observed for lung cancer or the 150-fold difference et al., 2005a; Schwartzbaum et al., 2006). Rare inherited
observed for melanoma (Forman et al., 2013). Thus, genetic mutations conferring increased glioma risk
for glioma, it seems unlikely that there are strong envi- within families have long been known (Table 1.1), but
ronmental risk factors associated with geography. they explain only a small proportion of brain tumor inci-
dence at the population level (Hemminki et al., 2009).
Li–Fraumeni syndrome, caused by a constitutive loss-
Incidence of molecular subgroups of glioma
of-function mutation in the TP53 gene, is the familial
Recent studies have focused on using molecular markers tumor syndrome most frequently associated with glioma.
to help clarify glioma classification (Cancer Genome However, numerous other rare Mendelian disorders
Atlas Research Network, 2008; Jiao et al., 2012; Liu increase risk of glioma, including neurofibromatosis 1
et al., 2012; Brennan et al., 2013; Wiestler et al., 2013; and neurofibromatosis 2, tuberous sclerosis, Lynch syn-
Eckel-Passow et al., 2015; Mur et al., 2015; Spiegl- drome, and melanoma-neural system tumor syndrome
Kreinecker et al., 2015). For example, three acquired (Ostrom et al., 2015b).
molecular alterations (IDH mutation, 1p19q co-deletion, A recent paper from the Gliogene Consortium dem-
and TERT promoter mutation) found in glioma tumor onstrated that rare loss-of-function mutations in the
cells define five etiologically and clinically distinct POT1 gene are associated with greatly increased risk
groups of glioma patients that account for over 95% of glioma in families (Bainbridge et al., 2015). This is
of grade II–IV gliomas (Eckel-Passow et al., 2015). the first monogenic cause of glioma to be discovered
Figure 1.4 shows the proportions of these molecular since CHK2 mutations were identified as an alternative
6 K.M. WALSH ET AL.
Table 1.1
Hereditary syndromes associated with increased risk of glioma

Gene Syndrome Inheritance Features Associated histology

NF1 Neurofibromatosis 1 Dominant Neurofibromas, schwannomas, Astrocytoma, optic

cafe-au-lait macules nerve glioma
NF2 Neurofibromatosis 2 Dominant Acoustic neuromas, Spinal ependymoma
meningiomas, neurofibromas
TSC1, TSC2 Tuberous sclerosis Dominant Multisystem nonmalignant Subependymal giant
tumors cell astrocytoma
MSH2, MLH1, Lynch syndrome Dominant Gastrointestinal, endometrial, Glioblastoma,
MSH6, PMS2 and other cancers astrocytoma
TP53 Li–Fraumeni syndrome Dominant Numerous cancers, especially Glioblastoma,
breast, brain, and soft-tissue astrocytoma, choroid
sarcoma plexus tumor
POT1 Melanoma- Dominant with Predisposition to melanoma and Oligodendroglioma and
oligodendroglioma reduced oligodendroglial tumors mixed
susceptibility syndrome penetrance oligoastrocytoma
CDKN2A Melanoma-neural system Dominant Predisposition to melanoma and Astrocytoma
tumor syndrome astrocytic tumors
IDH1/IDH2 Ollier disease/ Postzygotic Intraosseous benign Glioma
Maffucci syndrome mosaicism/ cartilaginous tumors, cancer
dominant predisposition
with reduced
penetrance

cause of Li–Fraumeni syndrome at the turn of the century (Table 1.2) (Shete et al., 2009; Wrensch et al., 2009;
(Bell et al., 1999). Interestingly, POT1 mutations seem to Sanson et al., 2011; Stacey et al., 2011; Walsh et al.,
confer greater risk for oligodendroglioma than for astro- 2014). These germline SNPs, that are more frequent in
cytoma. Of eight POT1 mutation carriers diagnosed with glioma cases than controls, also are called glioma risk
a glioma in the Gliogene study, six had oligodendroglial loci, risk alleles, or risk variants. Although some of these
tumors (four oligodendroglioma and two mixed oligoas- risk loci are in or near genes or chromosomal regions not
trocytoma) (Bainbridge et al., 2015). Other studies have previously associated with glioma (i.e., TERC, TERT,
shown that inherited mutations of POT1 also underlie RTEL1, involved in telomere maintenance and
some cases of familial melanoma (Robles-Espinoza CCDC26, and PHLDB1, of unknown function), several
et al., 2014; Shi et al., 2014). Three such families also con- of the glioma risk genes identified through GWAS have
tained individuals with an adult-onset malignant brain previously been identified in glioma tumor studies (i.e.,
tumor (Robles-Espinoza et al., 2014), strongly supporting TP53, CDKN2B, EGFR) and glioma-associated familial
the findings of the Gliogene Consortium. cancer syndromes (i.e., CDKN2B, TP53). Interestingly,
the inherited risk variants are not within the exonic
(protein-coding) portions of these genes. This suggests
Genome-wide association studies of glioma
that inherited differences in gene regulation, not protein
Prior to the advent of genome-wide association studies structure, confer glioma risk at these loci.
(GWAS), attempts to identify specific common inherited Four of these eight glioma risk regions identified by
variants associated with glioma did not yield consistent GWAS contain variants that appear to contribute to
findings (Walsh et al., 2013). In a GWAS, individuals development of all glioma grades and histologies (TERT,
with the disease of interest and healthy controls are gen- RTEL1, EGFR, TP53) (Table 1.2) (Jenkins et al., 2011;
otyped at hundreds of thousands of single-nucleotide Walsh et al., 2013; Rice et al., 2015). However, the other
polymorphisms (SNPs) to discover inherited variants four regions contain variants associated with the devel-
which are significantly more common in those with dis- opment of certain glioma grades, histologies, or molec-
ease than in those without. The glioma GWAS identified ular subtypes (TERC, CDKN2B, PHLDB1, CCDC26)
10 independently significant SNP associations located in (Table 1.2) (Jenkins et al., 2011, 2012; Rice et al., 2013;
eight gene regions, including near TERC, TERT, EGFR, Walsh et al., 2013). SNPs near CDKN2B on chromosome
CCDC26, CDKN2B, PHLDB1, TP53, and RTEL1 9 increase risk of astrocytomas, regardless of grade, but
EPIDEMIOLOGY 7
Table 1.2
Validated glioma risk loci from genome-wide association studies, hypothesized functional impact, and associated glioma
subtypes

Lead single-
nucleotide
Gene polymorphism Risk allele Odds Associated histology/
(chromosome) (risk allele) frequency ratio Hypothesized function molecular subtype

TERC rs1920116 (G) 0.71 1.27 Increased telomere length/ Astrocytoma III–IV
(3q26.2) telomerase activity
TERT rs2736100 (C) 0.50 1.33 Increased telomere length/ All glioma subtypes
(5p15.33) telomerase activity
EGFR rs2252586 (A) 0.28 1.15 Undetermined Astrocytoma III–IV
(7p11.2)
EGFR rs11979158 (A) 0.81 1.22 Undetermined Astrocytoma III–IV
(7p11.2)
CCDC26 rs55705857 (G) 0.07 6.10 Undetermined Oligodendroglial tumors,
(8q24.21) IDH-mutated
astrocytomas
CDKN2B rs1412829 (G) 0.41 1.43 Increased ANRIL Astrocytoma II–IV
(9p21.3) expression
PHLDB1 rs498872 (A) 0.28 1.52 Undetermined IDH-mutated glioma
(11q23.3)
TP53 rs78378222 (C) 0.01 2.65 Alteration of TP53 All glioma subtypes
(17p13.1) polyadenylation signal
RTEL1 rs6010620 (G) 0.77 1.42 Alteration of RTEL1– All glioma subtypes
(20q13.33) PCNA interaction
domain
RTEL1 rs4809324 (C) 0.10 1.66 Increased telomere length/ Astrocytoma III–IV
(20q13.33) telomerase activity

are not associated with risk of oligodendroglial tumors The sixfold increased risk of IDH-mutated glioma
(Walsh et al., 2013). SNPs in PHLDB1 increase risk of and oligodendroglioma associated with rs55705857 is
IDH-mutated gliomas, regardless of grade or histology the same magnitude of effect as that of BRCA1 mu-
(Rice et al., 2013). SNPs in CCDC26 on chromosome tations on breast cancer risk (Jenkins et al., 2012).
8q24 increase risk of IDH-mutated astrocytomas and However, glioma is sufficiently infrequent that the abso-
also of oligodendroglial tumors, regardless of IDH lute risk, even with very high relative risk, remains small
mutation status (Jenkins et al., 2012). (about 6/1000 individuals) and insufficient to justify
This risk region on chromosome 8q24 is a striking genetic screening (Rice et al., 2015). Because breast can-
example where GWAS initially identified variants that cer is a much more common cancer, the lifetime risk for
only modestly increase disease risk, but these variants women who have inherited deleterious variants in
were tagging or marking a much less common variant BRCA1 increases to about 65/100 (Gail et al., 1989;
that confers a very high risk of specific glioma subtypes Antoniou et al., 2003). Even so, BRCA1 testing is still
(Jenkins et al., 2012; Enciso-Mora et al., 2013). By only recommended for women who have an indication
sequencing the genomic region and conducting exten- of inherited risk due to family history or early age at
sive validation genotyping, researchers showed that the onset because screening the general population would
G allele of SNP rs55705857 confers an approximately identify many women as high-risk who will not develop
sixfold increased risk of IDH-mutated astrocytoma breast cancer in their lifetime. Because glioma is much
and oligodendroglial tumors (Jenkins et al., 2012). The rarer than breast cancer, rs55705857 currently has little
mechanism by which rs55705857 confers risk of glioma utility in genetic testing due to its low specificity. For
is not understood. One possibility is that rs55705857 may glioma screening based on inherited risk variants to be
be involved in long-range interactions with MYC, also feasible, one would need to identify a subgroup of peo-
located on 8q24.2, but this is still speculative (Tseng ple with substantially higher glioma risk than that of the
et al., 2014). general population.
8 K.M. WALSH ET AL.

Fig. 1.5. Hypothesized pathways of adult glioma development, including inherited risk variants and acquired mutations and chro-
mosomal changes. Established risk variants in TP53 and EGFR were not included because it remains unclear which branches of
gliomagenesis they influence. GBM, glioblastoma multiforme. (Reproduced from Rice et al., 2015, with permission from Oxford
University Press.)

The recently defined molecular subgroups of glioma TELOMERE MAINTENANCE AND

based on presence or absence of tumor TERT promoter GLIOMA
mutation, IDH mutation, and/or 1p19q co-deletion have
Recent studies of both inherited glioma risk variants and
distinct associations with the known glioma risk variants
of acquired mutations found in glial tumors identify a
(Eckel-Passow et al., 2015). In addition to the previously
substantial role for telomere maintenance in gliomagen-
discussed relationship between certain risk variants and
esis. Below we briefly review telomere biology and the
IDH mutation, there is some evidence that variants near
emerging role of telomeres in glioma development.
TERC and RTEL1 are more strongly associated with gli-
omas harboring somatic mutation of the TERT promoter
(Eckel-Passow et al., 2015). Figure 1.5 summarizes our
Telomere biology and heritability
current understanding of the hypothesized pathways
of telomere length:
of adult glioma development based on inherited and
acquired mutations (Rice et al., 2015). For some molec- Human telomeres, located at the ends of chromatids, are
ular subgroups, our understanding of the interplay composed of tandem hexanucleotide DNA repeats
between inherited variation and acquired tumor alter- (TTAGGG) and several associated telomere-binding pro-
ations is based on a relatively small number of patients. teins, including the shelterin complex (Griffith et al.,
Therefore, further investigations will be needed to ver- 1999). The primary function of telomeres is to compen-
ify, refine, and expand our understanding of inherited sate for incomplete DNA replication at chromosome
variation in gliomagenesis. ends, caused by the eukaryotic “end replication
EPIDEMIOLOGY 9
problem” (Shore and Bianchi, 2009). Human telomeres Unlike those near TERC and TERT, glioma risk alleles
are initially several kilobases in length but shorten with near RTEL1 are not consistently associated with LTL and
each mitotic division. When the telomere is depleted, suggest the presence of multiple causal alleles (Walsh
further division leads to loss of integral genomic content et al., 2014). Although rs6010620 is associated with sig-
and genomic instability, initiating apoptosis (McEachern nificantly shorter LTL (p ¼ 1.1 103), rs4809324 is asso-
et al., 2000). Thus, telomere depletion ultimately induces ciated with a modest increase in LTL (p ¼ 0.039) (Walsh
replicative senescence and limits the proliferative capac- et al., 2014). Thus, genetic variation near RTEL1 may
ity of cells. impact gliomagenesis through multiple mechanisms,
Telomere length is strongly influenced by age, and not all of which are necessarily telomere-dependent.
more subtly influenced by other factors, including: oxi- Another possibility is that the presence of longer versus
dative stress, smoking, gender, diet, exercise, and both shorter telomeres may have different relevance at dif-
inherited and acquired genetic variation (Denham ferent stages of the oncogenic process.
et al., 2013; Huzen et al., 2014). Mean leukocyte telomere
length (LTL) is a highly heritable trait, with heritability
Telomere lengthening and cancer
estimates as high as 80% (Slagboom et al., 1994;
Vasa-Nicotera et al., 2005). Mean LTL is known to be In cells requiring continuous renewal (e.g., germ and
independently associated with inherited variation in stem cells), telomere length is maintained by telomerase.
eight genomic regions located in or near ACYP2, TERC, Telomerase is inactive in most adult cells, but is often
NAF1, TERT, OBFC1, CTC1, ZNF208, and RTEL1 reactivated in cancer cells (Blackburn et al., 2006). The
(Mangino et al., 2012; Codd et al., 2013). telomerase enzyme adds nucleotides to telomeres and
is composed of a reverse transcriptase (encoded by
TERT) and an RNA template (encoded by TERC)
(Wang and Meier, 2004). Telomere maintenance in gli-
Telomere length and risk of glioma
oma occurs in at least two ways: through reactivation
Two case-control studies have attempted to directly of telomerase or through a homologous recombination-
measure LTL and determine its association with glioma based mechanism known as alternative lengthening of
risk (Walcott et al., 2013; Wang et al., 2014). The first of telomeres (ALT) (Chang et al., 2003; Heaphy et al.,
these studies, with 101 glioma patients and 198 healthy 2011b). How telomerase is reactivated in tumor cells
controls, did not identify a significant association was recently discovered when sequencing of germline
between glioma risk and LTL, but a larger study of DNA in melanoma-prone families identified two differ-
467 glioma patients and 467 controls showed that glioma ent activating mutations in the promoter of TERT (Horn
patients had significantly longer LTL than control sub- et al., 2013). Somatic mutations at these same positions
jects (Wang et al., 2014). Individuals in the upper tertile were identified in tumor DNA from sporadic melanoma
of LTL had increased risk of glioma relative to individ- cases (Huang et al., 2013). Soon after, targeted sequenc-
uals in the middle tertile (odds ratio, 3.5). Paradoxically, ing of the TERT promoter was carried out in a diverse set
those in the lowest tertile of LTL also had increased risk of human tumors (Killela et al., 2013). Mutations of the
of glioma compared to those with intermediate LTL. TERT promoter are observed in approximately 75% of
As discussed above, GWAS have implicated SNPs glioblastomas and 20% of grade II and III astrocytomas
near TERC, TERT, and RTEL1 in gliomagenesis (Shete and generate a novel GA-binding protein transcription
et al., 2009; Wrensch et al., 2009; Walsh et al., 2014). factor-binding site that upregulates TERT mRNA
Interestingly, the top glioma risk alleles near TERC expression (Bell et al., 2015). Strikingly, 75% of oligo-
and TERT also are very significantly associated with lon- dendrogliomas also harbor TERT promoter mutations,
ger LTL (p ¼ 5.5 10–20 and 4.4 10–19, respectively) despite otherwise being largely molecularly and clini-
(Walsh et al., 2014). Although these findings do not rule cally dissimilar from glioblastomas (Killela et al., 2013;
out the possibility that other genes in these two regions Eckel-Passow et al., 2015).
may underlie the genetic association with glioma, they Gliomas that do not maintain telomere length through
strongly support the idea that inherited risk for glioma activation of telomerase frequently activate the ALT
is, to some degree, mediated through longer telomere pathway via mutation of either ATRX or DAXX
length. This is further supported by the family-based (Heaphy et al., 2011a). More than half of all adult grade
study, mentioned above, that recently identified POT1 II and III astrocytomas have mutations in ATRX, and this
mutations as a cause of glioma, since dysfunction of this proportion is even higher when limited to IDH-mutated
shelterin complex protein was also associated with astrocytoma (Heaphy et al., 2011b; Kannan et al., 2012).
increased LTL in mutation carriers (Bainbridge DAXX mutations are primarily limited to pediatric
et al., 2015). gliomas (Heaphy et al., 2011b).
10 K.M. WALSH ET AL.
A growing body of epidemiologic and tumor genomic are less frequently reported by glioma cases than con-
research has identified an important role for telomere trols, suggesting that these conditions reduce glioma risk
maintenance in glioma predisposition, initiation, and prog- (Schoemaker et al., 2006; Wigertz et al., 2007; Scheurer
nosis. Though further research is necessary, several points et al., 2008; Berg-Beckhoff et al., 2009; Il’yasova et al.,
seem clear: (1) inherited variants in or near telomere- 2009; Wiemels et al., 2009; McCarthy et al., 2011).
related genes (TERC, TERT, RTEL1, POT1) are associ- A formal meta-analysis concluded that allergies reduce
ated with glioma risk; (2) mutations affecting telomere glioma risk by nearly 40% (Linos et al., 2007). The asso-
maintenance pathways (TERT promoter, ATRX, DAXX) ciation between increased allergies and reduced brain
are among the most recurrent acquired somatic events tumor risk was recently validated using prospectively
observed in gliomas; and (3) these inherited variants and collected data from US veterans, minimizing the poten-
acquired somatic mutations primarily cause lengthening, tial for recall bias (Cahoon et al., 2014). Further support
not shortening, of telomeres. However, as noted above, for this inverse association has been contributed by five
the importance of longer versus shorter telomeres may studies showing that glioma patients have lower levels of
vary during the oncogenic process. an atopy biomarker, immunoglobulin E (IgE) (Wiemels
et al., 2007, 2009; Calboli et al., 2011; Schlehofer et al.,
2011; Schwartzbaum et al., 2012).
ENVIRONMENTAL AND
Although mechanisms governing the potential anti-
DEVELOPMENTAL RISK FACTORS
glioma effects of allergy have not been identified, they
Below we briefly review nongenetic risk factors that may arise from the anti-inflammatory effects of
have been studied for gliomas, which are summarized interleukin-4 (IL-4) and IL-13 cytokines involved in aller-
in Table 1.3. gic and autoimmune disease, or from increased tumor
immunosurveillance in those with allergies and autoim-
mune disease (Dunn et al., 2002; Dinarello, 2003). It is
Allergies, infections, and immunologic risk
also possible that the inverse association results from
factors
immune suppression by the preclinical tumor, but valida-
Numerous studies have shown that allergic conditions, tion in prospective data sources makes this explanation
including asthma, hayfever, eczema, and food allergies, less likely (Cahoon et al., 2014).
Reduced glioma risk has also been attributed to a
Table 1.3 reported history of varicella-zoster virus (VZV) infec-
Nongenetic risk factors studied as possible glioma risk tions (i.e., chickenpox and shingles) and positive IgG
factors to VZV (Wrensch et al., 1997b, 2001, 2005b). With rela-
tive consistency, results suggest that history of chicken-
Association pox or shingles and anti-VZV IgG levels are inversely
(magnitude associated with adult glioma risk (Wrensch et al.,
Exposure and direction) 1997b, 2001, 2005b). Given the ubiquity of VZV expo-
sure, it may be the specific nature of a person’s VZV-
Established risk factors associated immune response, and not exposure to the
High-dose radiation +++ virus itself, that is responsible for this inverse association
Male vs. female gender +
with glioma (Wiemels et al., 2011). Strong anti-VZV reac-
White vs. African American ethnicity +
Increasing age +++
tions in highly allergic individuals may be a biomarker of
Probable risk factors effective CNS immunosurveillance (Lee et al., 2014).
Allergies/asthma At present, there is no strong epidemiologic evidence
Elevated IgE suggesting that human cytomegalovirus (HCMV) plays
Chickenpox/antivaricella-zoster virus IgG a role in the development of glioma. However, HCMV
Probably not risk factors nucleic acids and proteins are found in the tumors of
Diagnostic radiation some glioblastoma patients, and HCMV DNA has also
Head injury been found in the peripheral blood of glioblastoma
Residential power lines/electromagnetic patients (Poltermann et al., 2006). However, others
fields report that antibody positivity to HCMV in glioma
Cigarette smoking
patients is not different from that in controls or the gen-
Alcohol consumption
Cell phone use
eral population (Wrensch et al., 2005b; Poltermann et al.,
2006). The presence of HCMV gene products in blood or
+++ strong risk factor, relative risk >3; + risk factor >1; risk factor tumor tissue may result from reactivation of infection or
<1; no consistent associations. from infected tumor cells shedding viral DNA (Mitchell