ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2018

Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Medicine 1486

Towards new tools for clinicalevaluation and visualization oftumor growth in patients withglioma

KENNEY ROY ROODAKKER

ISSN 1651-6206ISBN 978-91-513-0410-6urn:nbn:se:uu:diva-356846

Dissertation presented at Uppsala University to be publicly examined in Aula Gunnesalen,Sjukhusvägen 1, ingång 10, Uppsala, SE, Friday, 30 November 2018 at 13:00 for thedegree of Doctor of Philosophy (Faculty of Medicine). The examination will be conductedin English. Faculty examiner: Professor Jan J. Heimans (VU University Amsterdam).

AbstractRoodakker, K. R. 2018. Towards new tools for clinical evaluation and visualization of tumorgrowth in patients with glioma. Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Medicine 1486. 83 pp. Uppsala: Acta Universitatis Upsaliensis.ISBN 978-91-513-0410-6.

Gliomas are derived from glial cells and are the most common type of primary brain tumorsin adults. Gliomas are classified by the World Health Organization (WHO) according to theirmalignancy grade and histological and molecular features. Malignancy grades range from I toIV. WHO grade I tumors are benign tumors, mostly occurring in childhood. High-grade gliomas(WHO grades III and IV) are undifferentiated and fast-growing tumors, with glioblastoma beingthe most common and malignant form. Patients with glioblastomas have a median survivalof only 15 months. Clinical outcomes vary, however, and markers are needed to assist inthe decision-making process and management of these patients. PROX1 is a transcriptionfactor critical for embryonic development, with a role in cell cycle control and progenitor celldifferentiation. Apart from its role in normal central nervous system development, PROX1 hasbeen ascribed both tumor suppressive and oncogenic roles in several human cancers. The roleof PROX1 as a prognostic factor for survival in patients with glioblastomas was the focus ofpaper I.

Gliomas WHO grade II, also called diffuse low-grade gliomas (DLGGs), are well-differentiated tumors that occur mainly in adult life, with a peak incidence at around 30–35years of age and a median survival of 5–10 years. DLGGs grow continuously at a rate of afew mm per year and have a strong tendency to infiltrate the white matter tracts surroundingthe tumor. Eventually these tumors transform into high-grade gliomas, but, as is the casewith glioblastomas, there is a large variety of clinical outcomes. For radiological diagnosis,magnetic resonance imaging (MRI) is routinely used, often in combination with advanced MRI.Positron emission tomography with amino acid tracers provides additional diagnostic accuracy.From a histological as well as imaging point of view, DLGGs are heterogeneous tumors. Theheterogeneity of DLGGs, in particular the correlation between radiological and histologicaltumor features, was the focus of paper II & paper III.

Seizures are amongst the most common presenting symptoms of patients with gliomas.Seizure semiology in patients with brain tumors and other structural brain lesions is closelyrelated to the anatomical location of the lesion and the involvement of functional networks. Apossible dynamic interplay between the anatomical region of seizure onset and connected targetareas within the network was the focus of paper IV.

Keywords: PROX1, Histology, MRI, PET, Epilepsy, Co-registration

Kenney Roy Roodakker, Department of Neuroscience, Neurology, Akademiska sjukhuset,Uppsala University, SE-75185 Uppsala, Sweden.

© Kenney Roy Roodakker 2018

ISSN 1651-6206ISBN 978-91-513-0410-6urn:nbn:se:uu:diva-356846 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-356846)

To my father and mother, for their love and unconditional support.

Jullie liefde en onvoorwaardelijke steun, doet mij bergen verzetten.

Cover and illustrations by Concepción Bueno Galera and Concepción Galera Moya

Researchers cannot transcend the determinism of phenomena; instead, their mission is limited to demonstrating the how, not the why, of observed

changes.

Knowing the conditions under which a phenomenon occurs allows us to re-produce or eliminate it at will, therefore allowing us to control and use it for

the benefit of humanity.

From Advice for a Young Investigator, by Santiago Ramón y Cajal

Examining Committee

Faculty Examiner (opponent) Professor Jan J. Heimans Department of Neurology VU University Amsterdam, Amsterdam, Netherlands

Evaluating Committee

Professor Maria Kristoffersen Wiberg Department of Medical and Health Sciences Linköping University, Linköping, Sweden Associate professor Johan Bengzon Department of Neurosurgery Lund University, Lund, Sweden Docent Ingela Nygren Department of Neuroscience Uppsala University, Uppsala, Sweden

Chairperson Professor Per Enblad Department of Neurosurgery Uppsala University, Uppsala, Sweden

List of Papers

This thesis is based on the following papers, which are referred to in the text by their Roman numerals.

I Roodakker, KR.*, Elsir, T.*, Edqvist, PH., Hägerstrand, D., Carlson, J., Lysiak, M., Henriksson, R., Pontén, F., Rosell, J., Söderkvist, P., Stupp, R., Tchougounova, E., Nistér, M., Malmström, A., Smits, A. PROX1 is a novel pathway-specific biomarker for high-grade astrocytomas; Results from independent glioblastoma cohorts stratified by age and IDH muta-tion status. Oncotarget, 2016 Nov 8; 7(45): 72431-72442

II Zetterling, M., Roodakker, KR., Berntsson, SG., Edqvist, PH., Latini, F., Landtblom, AM., Ponten, F., Alafuzoff, I., Larsson, EM., Smits, A. Extension of diffuse low-grade gliomas beyond radiological borders as shown by the coregistration of histopathological and magnetic resonance imaging data. Journal of Neurosurgery, 2016 Nov; 125(5): 1155-1166

III Roodakker, KR., Alhuseinalkhudhur, A., Jaff, M., Georganaki, M., Zet-terling, M., Berntsson, SG., Danfors, T., Strand, R., Enkvist, P-H., Dim-berg, A., Larsson, E-M., Smits, A. Region-by-region analysis of PET, MRI, and histology in en bloc-resected oligodendrogliomas reveals intra-tumoral heterogeneity. European Journal of Nuclear Medicine and Mo-lecular Imaging. Published online, 2018 Aug.

IV Roodakker, KR., Latini, F., Ezra, B., Gauffin, H., Berntsson, SG., Landtblom, A-M. Functional connectivity changes of the brain related to ecstatic seizures. Submitted, 2018 Aug.

* Authors contributed equally to the work

Article co-authored but not included in this thesis:

Zhang L.*, He L.*, Lugano R., Roodakker KR., Bergqvist M., Smits A., Dimberg A. IDH-mutation status is associated with distinct vascular gene expression in lower grade gliomas. Neuro-oncology, Accepted for publication, 2018 May.

Contents

Proem ............................................................................................................ 13

Part I Epidemiology of gliomas ................................................................... 15 Incidence and survival .............................................................................. 16 Natural course of disease .......................................................................... 18

Low-grade glioma ................................................................................ 18 High-grade gliomas ............................................................................. 18

Part II Histological classification .................................................................. 21 Morphological appearance ....................................................................... 22

Part III Molecular classification ................................................................... 25 Diagnostic markers ................................................................................... 26 Prognostic and predictive markers ........................................................... 29 A new prognostic marker ......................................................................... 30

Part IV Radiological diagnosis ..................................................................... 33 Diagnostic imaging .................................................................................. 34

Magnetic Resonance Imaging .............................................................. 35 Advanced imaging .................................................................................... 35 A novel approach ..................................................................................... 36

Part V Clinical aspects and seizure semiology ............................................. 37 Seizure semiology of gliomas .................................................................. 38

Brain connectivity in epilepsy ............................................................. 39

Part VI Therapeutic management ................................................................. 41 Glioma treatment ...................................................................................... 42

Part VII Present investigations ...................................................................... 45 Aims of this thesis .................................................................................... 46

Scientific questions .............................................................................. 47 Paper I: A pathway-specific prognostic marker ...................................... 48 Paper II: Tumor growth and invasion ..................................................... 52 Paper III: Intra-tumoral heterogeneity .................................................... 55 Paper IV: A dynamic interplay ............................................................... 58

Concluding remarks ...................................................................................... 60

Future Perspectives ....................................................................................... 62

Svensk Sammanfattning ................................................................................ 64

Nederlandse Samenvatting ............................................................................ 66

Words of gratitude ........................................................................................ 68

References ..................................................................................................... 71

Abbreviations

1P19Q 2-HG Alpha-KG ATRX CNS DG DLGG FET FLAIR GBM G-CIMP Gy HGG IDH-1 IHC LGG LOH MET MGMT MRI MAP2 NOS NSCs PCV PET PROX1 rCBV RT SGZ SVZ TCGA TERT TMZ TP53 WHO

Loss of heterozygosity for chromosome arms 1p and 19q 2-hydroxyglutarate Alpha-ketoglutarate Alpha Thalassemia/Mental Retardation Syndrome X-Linked Central nervous system Dentate gyrus Diffuse low grade glioma 18F- fluoroethyl-L-tyrosine Fluid-attenuated inversion recovery Glioblastoma Glioma CpG island methylated phenotype Gray (one joule per one kilogram of matter) High-grade glioma Isocitrate dehydrogenase 1 Immunohistochemistry Low-grade glioma Loss of heterozygosity 11C-methyl-L-methionine O-6-methylguanine-DNA methyltransferase Magnetic resonance imaging Microtubule-associated protein 2 Not otherwise specified Neural stem cells Procarbazine-lomustine-vincristine Positron emission tomography Prospero Homeobox 1 Relative cerebral blood volume Radiotherapy Subgranular zone Subventricular zone The Cancer Genome Atlas Telomerase reverse transcriptase Temozolomide Tumor suppressor protein 53 World Health Organization

Figure legend

Fig 1. Distribution of primary brain and other central p. 16nervous system tumors

Fig 2. Illustrations of mmunohistochemistry (IHC) p. 23

Fig 3. Histological and molecular classification of gliomas p. 28

Fig 4. Illustration of asymmetric cell division p. 31

Fig 5. Illustration of MRI, perfusion MRI and PET p. 36

Fig 6. White matter fiber architecture p. 39

Fig 7. Illustration of growth pattern of diffuse gliomas p. 43

Fig 8. TMA showing PROX1 expression p. 49

Fig 9. En bloc tumor resection p. 52

Fig 10. Tumor cells illustrated outside the radiological border p. 54

Fig 11. Radiological tumor border and histological tumor border p. 54

Fig 12. Co-registration of Histology-MRI-PET p. 56

Fig 13. A schematic visualization of the four papers presented p. 61in this thesis, and their respective overlap

13

Proem

Humans have always been fascinated by the neurobiology of the brain, even as far back as ancient Egypt. Hippocrates wrote about the brain as the seat of intelligence. In the 19th century, the brain and its neurons were described by scientists such as Golgi and Ramón y Cajal, contributing to the founding of modern Neuroscience. Their observations led to the understanding that the human brain consists, among many other things, of neurons and neuroglia that form the building blocks of the central nervous system (CNS).

Neuroglia, or glial cells, were thought to be the glue of the nervous system. Nowadays, we know their importance with regard to support, protection, neu-rotransmission, and synaptic connections in the CNS. Sometimes, these glial cells acquire the ability to escape normal growth-restraining influences due to molecular and genetic alterations. This transformation makes it possible for them to form self-renewing cancer cells or neoplasms in the brain, called gli-omas, which can affect brain function, cause severe symptoms such as epi-lepsy, and can be life-threatening, depending on their location and aggressive-ness. Gliomas are well described in the updated (2016) fourth edition of the World Health Organization (WHO) system. Here, gliomas are classified and graded based on their microscopic appearance, i.e. tumor histology and addi-tional molecular aberrations. Some of these histological and molecular abnor-malities carry important diagnostic, prognostic, and/or predictive information. The use of these prognostic markers could therefore benefit precision medi-cine. In order to determine the location and size of neoplasms in the brain before surgery, noninvasive imaging is needed. In 1895, Wilhelm Röntgen demonstrated x-ray imaging. In 1971, the first medical application using mag-netic signal changes over time was suggested, to distinguish cancer from healthy tissue. Nowadays, neuroimaging has evolved into a comprehensive diagnostic tool. To detect brain lesions, magnetic resonance imaging (MRI) is the gold standard, while new advanced imaging techniques provide additional functional, physiological, and molecular diagnostic information.

The treatment of gliomas is dependent on several factors, with no standard treatment. In the majority of cases, tumor treatment is not curative but can significantly prolong survival and cause symptom relief. The chosen strategy is based on histological diagnosis and additional clinical information. Preci-sion medicine is of growing importance in tumor research and tumor treat-ment.

15

Part I Epidemiology of gliomas

16

Incidence and survival Genetic and molecular alterations can arise in the neuroepithelial tissue of the human brain. These peculiar alterations in glial cells of the brain give rise to gliomas. These neoplasms have been classified by the World Health Organi-zation (WHO), updated in 2016, based on histopathological and molecular fea-tures of the affected cells [1, 2]. The yearly incidence of newly diagnosed gli-omas is around 5.27 per 100,000 persons, based on the European Standard Pop-ulations [3]. In the WHO classification, gliomas are graded from I to IV, with grade IV being the most malignant form. The incidence of the different glioma subtypes, according to the central brain tumor registry of the United States (CBTRUS), is shown in Figure 1. Grade I tumors are benign childhood tumors and are not considered in this thesis. Low-grade gliomas (WHO grade II) are well-differentiated tumors with cytological atypia. High-grade gliomas (WHO grades III and IV), of which glioblastomas (GBMs) are the most frequent, are undifferentiated tumors with anaplasia and mitotic activity (WHO grade III), and endothelial cell proliferation, and/or necrosis in addition to mitosis and nuclear atypia (WHO grade IV) [1]. At present, the median overall survival of patients with gliomas is poor, depending on the tumor grade. For grade II gli-omas, survival ranges from 5 years to >10 years, for grade III gliomas 2–3 years, and for GBMs 1.5 years [3], even with intensive multimodal treatment strategies (surgery, chemoradiation, and adjuvant chemotherapy).

Figure 1. Distribution of primary brain and other central nervous system tu-mors diagnosed in the United States in 2010–2014 [4]. With permission of Oxford University Press.

17

Diffuse gliomas are divided into oligodendrogliomas and astrocytomas based on morphological similarities and molecular tumor features. This means that the predominant population of tumor cells share morphological similarities with either oligodendrocytes or astrocytes [5], and additional genetic driver mutations [2]. As mentioned above, low-grade gliomas have a better progno-sis than high-grade gliomas, with an additional survival difference between astrocytomas and oligodendrogliomas. Oligodendrogliomas (incidence of 0.4 per 100,000 persons, 78% survival at 5 years, 51% at 10 years) have a more favorable prognosis than astrocytomas (incidence of 4.8 per 100,000 persons, 65% survival after 5 years, and 31% after 10 years) [6-8]. Genetic aberrations are associated with clinical relevance such as incidence and survival differ-ences between tumor subtypes. Genetic aberrations have, for example, been associated with sensitivity to some forms of therapy and to improved patient outcome [9, 10].

18

Natural course of disease

Low-grade glioma Diffuse low-grade gliomas (LGGs) are low-malignant brain tumors. As men-tioned previously, they consist of grade II astrocytic and oligodendroglial tu-mors [8], affect young adults with an average age at diagnosis of 40 years, and have an annual incidence of 1–2 per 100,000 adults [11, 12] (Table 1). Diffuse astrocytomas (AIIs) are well-differentiated, slow-growing tumors without ob-vious anaplastic features on histologic examination [13, 14]. Oligodendroglial tumors are highly cellular lesions consisting of relatively small but densely packed tumor cells [10, 15, 16]. Oligodendrogliomas (OIIs) are further char-acterized by branching networks of capillaries (chicken-wire pattern) that sets them apart from astrocytomas. LGGs are commonly located in or close to the so-called eloquent areas of the brain that involve motor, language, visuospa-tial, and memory functions [17-19]. Additionally, LGGs are heterogeneous tumors which possess the ability to grow (average growth rate of 5 mm/year [20, 21]) and invade the brain, complicating surgery. If LGGs remain un-treated, it is just a matter of time before they undergo malignant differentiation (within 5 to 10 years), resulting in progressive neurological disability and ul-timately death [22]. Our accumulating understanding of the biology of LGGs and improved surgical techniques teaches us that early resection of LGGs de-lays the time to malignant progression and therefore improves overall survival [22, 23].

High-grade gliomas High-grade gliomas (HGGs) consist of anaplastic glioma (grade III) and glio-blastoma (grade IV). Typical features of anaplastic gliomas are an increased cellularity, nuclear atypia, and marked mitotic activity, but an absence of mi-cro-vascular proliferation and/or necrosis [1] and consist of either anaplastic oligodendrogliomas (OIIIs) or astrocytomas (AIIIs). GBMs are the most com-mon (4–5 per 100,000 person-years [24]) and lethal type (mean survival around 15 months [25]) of adult brain tumors (Table 1), with histological char-acteristics similar to those of anaplastic gliomas, but show additional endothe-lial proliferation and/or necrosis. Recent advances in cancer genomics have provided support for distinctive molecular correlates of two clinical GBM phenotypes: primary GBMs that arise de novo in predominantly older patients and secondary GBMs in younger patients with a history of prior low-grade gliomas. Clinically defined primary or de novo GBMs are the most prevalent tumor (about 90% of cases), and they are the most predominant tumors in pa-tients over 55 years of age. They develop very rapidly after a short clinical course, without evidence of progression from a prior lesion [1]. Secondary

19

GBM (about 10% of cases) develops slowly by progression from a prior lower grade diffuse glioma and preferentially arises in younger patients. The pro-gression of astrocytomas from AIIs to AIIIs is suggested to be an early stage in the evolution of secondary GBMs [26], sharing similar molecular charac-teristics [27]. Survival of patients with secondary GBMs is longer than that in those with primary GBMs [27]. The division into primary and secondary tu-mors is now done by genetic driver mutations, of which the IDH mutation status is the most used [2]. Table 1. WHO Grade Age of onset Average survival Occurrence (%) Grade II (diffuse glioma)

30 to 40 years 5 to 10 years 7.7

Grade III (anaplastic glioma)

Early 40s 2 to 3 years 6.5

Grade IV (glioblastoma)

Mid 50–60s 14 months 56.1

WHO: World Health Organization [2]

21

Part II Histological classification

22

Morphological appearance At initial diagnosis, brain tumors must be differentiated from benign lesions. In addition to patient information, such as the patient’s age, size of the tumor, and tumor growth rate, the final histological classification and grading are the most important factors that determine treatment strategies [28, 29]. His-tological evaluation and tumor grading of suspicious brain lesions is done by histopathological assessment of tissue samples taken from the tumor area. Mi-croscopic similarities of tumor tissue have been primarily based on light mi-croscopic features in hematoxylin and eosin-stained sections, immunohisto-chemical expression of lineage associated proteins, and ultrastructural charac-terization [2]. Histologically, tumor tissue is an intermingle of tumor cells with normal cells, such as neurons and native glia, and it forms clusters around neurons (satellitosis), blood vessels, and the pia mater [29].

As mentioned in the previous chapter, phenotypical features in tumor classifi-cation are an increased cellular density, nuclear atypia, mitosis, vascular pro-liferation, and necrosis (Table 2). Diffuse astrocytomas are usually well-de-lineated tumors, showing increased cellularity and nuclear atypia. They have enlarged irregularly shaped nuclei or elongated and hyperchromatic nuclei. Anaplastic astrocytomas are histologically similar but have more frequent mi-toses and increased cellularity. GBMs (WHO grade IV) show additional mi-crovascular proliferation and/or necrosis [29].

Table 2. WHO Grade Comments Grade II (astrocytoma)

Increased hypercellularity; no mitosis; no vascular prolif-eration; no necrosis; can recur as a higher-grade tumor

Grade III (anaplastic astrocytoma)

High rate of hypercellularity; high rate of mitosis; no vas-cular proliferation; no necrosis; high rate of tumor recur-rence

Grade IV (glioblastoma)

Very high rate of hypercellularity; very high rate of mito-sis; presence of vascular proliferation; presence of necrosis

WHO: World Health Organization [2]

Immunohistochemistry (IHC) is the gold standard for studying protein dis-tribution in histologic tissue sections, using selective antibody binding [30] (Figure 2). An example is the glial fibrillary acidic protein (GFAP) [31], which provides a way to distinguish astroglial cells from other cell types in the brain. In contrast to astrocytomas, the expression of this protein is absent in oli-godendrogliomas. IHC for oligodendrocyte transcription factor 2 (OLIG2) is a way to highlight the tumor nuclei in oligodendroglial cells, and IHC for neu-

23

rofilament (NF) highlights an infiltrative growth pattern [29]. The microtu-bule-associated protein 2 (MAP2) is expressed in both astrocytic and oli-godendroglial tumors. However, the distinct pattern of MAP2 immunoreac-tivity may be useful as a diagnostic tool in oligodendrogliomas [32]. The an-tigen Ki67 is a cell proliferation-related protein, which increases gradually with higher mitotic activity. Since diffuse LGGs have low mitotic activity, the labeling index is usually less than 5% [33]. For HGGs, the labeling index is usually more than 25% due to the high mitotic activity.

Figure 2. Immunohistochemistry (IHC) of GFAP, MAP2, OLIG and Ki67. Images subtracted from https://www.proteinatlas.org.

Based on histological criteria, diffuse gliomas can be classified as follows: diffuse astrocytomas, anaplastic astrocytomas, and GBMs (WHO grades II, III, and IV, respectively), and oligodendrogliomas (including anaplastic ex-amples) subtypes [29]. However, due to the biological diversity of gliomas it is very difficult to precisely grade tumors using microscopic criteria and avoid any possible misclassification [10].

25

Part III Molecular classification

26

Diagnostic markers Molecular and radiological markers are of increasing importance for tumor assessment. Used together with histological classification, they help to prevent misclassification and guide clinical decision making regarding glioma sub-types. In general, diagnostic markers are based on key molecular alterations in gliomas, which helps to increase the accuracy of classification.

Important examples of diagnostic markers used in glioma classification are the isocitrate dehydrogenase 1 and 2 gene family (IDH1/IDH2), the alpha-thalassemia/mental retardation syndrome X-linked (ATRX) gene, the tumor suppressor protein 53 (TP53), telomerase reverse transcriptase (TERT), and the codeletion of 1p and 19q chromosomes (1p19q codeletion) (Figure 3) [2]. The classification of low-grade glioma, for example, relies on IDH1 mutation status [34], while subtyping of oligodendroglioma is based on the codeletion of the 1p and 19q chromosomes [9]. Therefore, subtyping based on molecular abbreviations leads to more accurate diagnosis and a decreased risk of mis-classification. Which is why genetic profiling, in addition to histological as-sessment, is of clinical importance in reaching more accurate treatment deci-sions for patients with a glioma.

Mutations in the isocitrate dehydrogenase (IDH) gene are found in the cyto-plasm and in peroxisomes. IDH catalyzes the oxidative decarboxylation of isocitrate into alpha-ketoglutarate (alpha-KG) and nicotinamide adenine dinu-cleotide (NADPH), in the Kreb cycle and in the cytoplasm [35, 36]. Most IDH mutations are caused by an amino acid substitution at position 132 in IDH1 from arginine to histidine (R132H) and substitution at position 172 in IDH2 from arginine to lysine (R172K) [36]. Mutant IDH genes inhibit histone de-methylation and induce a glioma hypermethylation phenotype since the mu-tant proteins convert alpha-KG to the putative oncometabolite 2-hydroxyglu-tarate (2-HG). This hypermethylation (glioma CpG island methylated pheno-type or G-CIMP) leads to transcriptional alterations, with prognostic implica-tions for patients with diffuse gliomas [10]. IDH gene mutations (codon R132H) have been identified in astrocytic and oligodendroglial grade II and grade III gliomas. Additionally, they have been identified in secondary GBM, but not in primary or de novo GBM. IDH mutations may pave the way for the development of other mutations such as TP53 and ATRX in astrocytomas and codeletion of 1p/19q in oligodendrogliomas [37].

The alpha-thalassemia/mental retardation syndrome X-linked (ATRX) gene is located on chromosome Xq21.1, and protein expression is found in the cell nuclei. ATRX gene mutations result in a loss of nuclear protein expression in tumor cells, but not in non-tumor cells, and lead to telomere dysfunction [38]. ATRX is associated with mutations such as IDH1 and TP53. Interestingly,

27

ATRX gene mutation is almost mutually exclusive to 1p 19q codeletion, mak-ing it a very specific marker for astrocytic tumors in combination with the IDH mutation status [39]. Additionally, ATRX mutations have been observed in 60%–80% of secondary GBM but in only 5% of primary GBM, making it an additional marker for secondary GBM [40].

The loss of heterozygosity (LOH) for chromosome arms 1p (short arm of chro-mosome 1) and 19q (long arm of chromosome 19) has been known as a diag-nostic marker of oligodendrogliomas since 1998 [9]. The codeletion of these two chromosomal arms is due to an unbalanced translocation that leads to the loss of one hybrid chromosome and thereby LOH [38]. The 1p19q codeletion is one of the molecular alterations with beneficial clinical response to treat-ment [9, 10], which is described in more detail in the following chapter. Gli-omas with either 1p or 19q deletion alone (nonconcurrent deletion) do not share the survival benefit of codeleted tumors [36]. As mentioned previously, 1p19q codeletion is found only in oligodendroglial tumors, therefore the de-tection of codeletion is a useful way to discriminate oligodendroglioma from astrocytic tumors. Telomerase reverse transcriptase (TERT) promoter mutation is located on chromosome 5p15.33, and the promoter region of the gene is the most im-portant regulatory element for telomerase expression [41]. TERT expression is mostly repressed in somatic cells, with the exception of proliferative cells, in self-renewing tissues and cancer. Increased expression of the TERT to maintain telomeres is essential for cancer cells to maintain their potential to proliferate and prevent biological aging (senescence) [36]. Induction of TERT transcription can lead to activation of oncogenes and inactivation of tumor suppressors. An example of a tumor protein that has been shown to downreg-ulate TERT transcription is p53 [41]. The increased TERT expression due to a mutation in the promoter region serves as an important marker of glioma subtypes. TERT mutations are found in IDH wild-type and 1p19q intact pri-mary GBMs. Interestingly, TERT mutations are also found in classical oli-godendrogliomas [42]. Information in Figure 3 is based on previous research [2, 9, 34, 37, 39, 40, 42-45].

28

Figu

re 3

. Hist

olog

ical

and

mol

ecul

ar c

lass

ifica

tion

of g

liom

as, w

ith c

orre

spon

ding

tum

or tr

eatm

ent [

2, 9

, 34,

37,

39,

40,

42-

45].

29

Prognostic and predictive markers Prognostic and predictive markers provide information about the expected benefit of a certain treatment and can therefore assist in selecting the most optimal therapy option. In some cases, molecular genetic subtyping of gliomas has provided information that has proved useful to assure better overall sur-vival in certain subgroups of patients [46, 47] or even to positively affect response to treatment [9, 10], underlining the clinical importance of prognos-tic markers.

A prognostic marker for molecular genetic subtyping of groups with an overall better survival in patients with glioma is the IDH mutation status. In the pre-vious chapter, it is mentioned that the IDH mutation is associated with im-proved survival compared to IDH wild-type tumors [48]. It has been found that a mutated IDH enzyme is unable to generate alpha-KG. This leads to the accumulation of 2-HG in gliomas [38]. Accumulation of this 2-HG results in the previously mentioned CpG island methylator phenotype (G-CIMP). Pa-tients with G-CIMP tumors are younger at the time of diagnosis and experi-ence significantly improved outcome if the G-CIMP tumor is associated with MGMT promoter methylation [49]. An association with survival has been suggested with regard to the differences between primary and secondary GBMs, where overall secondary GBMs have a higher frequency of promoter methylation and better overall survival than primary GBMs [50].

Predictive markers known to have clinical importance with regard to guiding the treatment of patients with glioma are 1p19q codeletion [9, 10], MGMT promoter methylation [36, 44], mutations of TERT [36], and BRAF [44]. As mentioned previously, LOH for chromosome arms 1p and 19q is due to an unbalanced translocation that leads to the loss of one hybrid chromosome and thereby LOH [38]. Allelic losses of chromosomes 1p and 19q are a molecular signature of oligodendrogliomas and occur in 50%–70% of both low-grade and anaplastic tumors [51]. Additionally, it has been shown that 1p19q codele-tion in recurrent anaplastic oligodendrogliomas has a better response to PCV (procarbazine-lomustine-vincristine) chemotherapy. This improved survival in tumors with 1p19q codeletion is directly linked to IDH mutation status, where IDH mutant tumors with 1p19q codeletion have the best outcomes (rep-resenting oligodendroglial tumors), followed by IDH mutant tumors without 1p19q codeletion (astrocytic tumors), while IDH wild-type tumors have the worst outcomes (primary GBM) [9].

MGMT (O6-methylguanine-DNA methyltransferase) is a mismatch repair en-zyme suggested to remove the O6-methylguanine adducts induced by alkyla-tor chemotherapies. Due to the repair capacity of MGMT, higher levels are suggested to cause temozolomide (TMZ) resistance [36]. Conversely this

30

means that tumor cells with low or absent MGMT expression are rendered more sensitive to TMZ [44]. Methylation of the MGMT promoter in the CpG-rich region results in epigenetic silencing (decreased expression) of the MGMT protein. Therefore, MGMT promoter methylation is a predictive marker for longer survival in GBM [52]. Survival almost doubles in patients with IDH wild-type GBM (and high MGMT promoter methylation) receiving both radiation and TMZ. As mentioned before, TERT transcription can lead to activation of oncogenes and inactivation of tumor suppressors, which are found in IDH wild-type and 1p19q intact primary GBMs. TERT has a dichot-omous prognostic effect where it enhances sensitivity to TMZ and improves survival in MGMT methylated tumors, while it is more resistant in unmethyl-ated MGMT tumors [36]. Therefore, TERT promoter mutation is a predictive marker in tumors with MGMT methylation.

A new prognostic marker Through defective regulation of cell growth, differentiation, death and/or sur-vival, cancer cells are capable of self-renewal and sustaining proliferation. It is therefore important to identify key regulatory proteins and signaling path-ways, not only for histological classification or predicting response to treat-ment, but also to unravel oncogenic mechanisms in neurogenesis. The under-lying physiology during neurogenesis in CNS diseases is becoming more and more clear. In the human CNS, neurogenesis occurs mostly in the subgranular zone (SGZ) of the dentate gyrus (DG) in the hippocampus and the subventric-ular zone (SVZ) of the lateral ventricles [53], involving cell proliferation, mi-gration, and neuronal differentiation. A mammalian transcription factor sug-gested to play a key role in embryonic development and adult neurogenesis of the hippocampus is Prox1, which is commonly used as a specific marker for the DG cell lineage [54]. Prox1, whose homolog in Drosophila is Prospero, belongs to the family of homeobox transcription factors characterized by a conserved homeo-domain that binds DNA and allows transcriptional regula-tion of other genes [55].

Prospero is a key determinant of cell fate in the Drosophila CNS [56], and has a 65% similarity with the human Prox1 homeodomain [55]. Prospero is found to be a critical regulator of the balance between self-renewal and differ-entiation in neural stem cells (NSCs) [57], and is expressed in a subset of neu-roblasts, sensory neuron precursors, and glial precursors, but not in mature neurons. Prospero suppresses self-renewal of neuroblast daughter cells and cell cycle progression, while activating neuronal differentiation (Figure 4). On the other hand, absence of Prospero leads to expression of self-renewal pro-teins, while neurons fail to differentiate [58], eventually leading to tumor-like

31

masses. Lack of Prospero in the embryonic stage of Drosophila leads to tumor formation [58].

Figure 4. Illustration of asymmetric cell division [55].

The human Prox1 gene is composed of five exons and four introns localized on chromosome 1q32.2– q32.3 [55]. In vertebrates, Prox1 is a key regulator for the generation of many organs during embryogenesis such as the brain, spinal cord, retina, lens, liver, pancreas, and endothelial lymphatic system [53]. Embryos with knock-out of Prox1 die before birth due to multiple devel-opmental defects [59]. High protein expression of Prox1 in the mammalian brain is found in the previously mentioned areas known to harbor neural pro-genitor and neural stem cells: the SVZ of the lateral ventricle wall and the SGZ in the DG [60, 61]. Due to both the activating and repressive character-istics of the homeodomain, it is not fully clear whether Prox1 has tumor sup-pressive or oncogenic properties. However, just as in Drosophila, alterations in Prox1 expression have been found to be related to a variety of cancer forms in humans, i.e. in neuroblastoma, glioma, small cell lung carcinoma, lung car-cinoid tumor, colon cancer, small intestine adeno-carcinoma, liver carcinoma, and rhabdomyo-sarcoma [55]. Like Prospero, Prox1 is found to promote neu-rogenesis, to inhibit astrogliogenesis and self-renewal, to suppresses cell cycle progression and proliferation [62], and to direct the terminal neuronal differ-entiation of progenitor cells into mature neurons [53]. This balance between

32

progenitor cell differentiation and self-renewal is suggested through suppres-sion of cell division by Sox1 regulating Prox1, or transcriptional regulation by Prox1 of Notch1 [55], which is involved in stem cell self-renewal and stimu-lating neurogenesis. Accordingly, knock-out of Sox1 in mice accumulates Prox1 expression levels [63], while Prox1 represses Notch1 [64].

Additionally, Prox1 has been suggested as a diagnostic marker to distinguish low-grade from high-grade gliomas [65], because a higher malignancy grade in gliomas is correlated to the elevated protein expression of Prox1. However, in a subset of diffuse low-grade gliomas, prox1 protein expression was found to be up-regulated. This subtype shows a more advanced and less differenti-ated phenotype, with worse prognosis and more rapidly progression to high-grade tumors [65, 66]. Recently, Prox1 expression has been suggested not only to be a diagnostic but also a predictive marker in human gliomas [55] due to survival disadvantage of Prox1 expression in patients with low-grade gliomas [66]. Since this has not yet been established in higher grade gliomas, we asked ourselves the question whether Prox1 is a predictive marker in high-grade gli-omas as well. We therefore investigated the prognostic role of the transcrip-tion factor Prox1 in different cohorts of patients with GBMs, described in pa-per I.

33

Part IV Radiological diagnosis

34

Diagnostic imaging Though histopathology is the gold standard for tumor grading, preoperative noninvasive imaging parameters might be helpful in the clinical decision-making process. Conventional MRI provides important tumor information re-garding contrast material enhancement, edema, distant tumor foci, hemor-rhage, necrosis, and mass effect. Therefore, non-invasive anatomic imaging is helpful in characterizing tumor aggressiveness and hence tumor grade [67], showing maximum accuracy for LGG and for HGG subtypes of anaplastic astrocytomas and GBMs of 91, 83, and 88%, respectively [68]. However, sometimes anatomical imaging is unreliable for tumor grading, with a sensi-tivity in different studies ranging from 55.1% to 83.3% [67]. Low-grade glio-mas are sometimes mistaken for a high-grade glioma when the glioma demon-strates peritumoral edema, contrast enhancement, central necrosis, and mass effect. Conversely, high-grade gliomas can sometimes demonstrate minimal edema, no contrast enhancement, no necrosis, and no mass effect, and there-fore be mistaken for a low-grade glioma [67]. Therefore, more reliable nonin-vasive distinction between tumor tissue and the peritumoral brain tissue can improve therapy planning of cerebral gliomas.

In low-grade gliomas, the T2-weighted fluid-attenuated inversion recovery (T2-FLAIR) pulse sequence is the best method for noninvasively mapping of the extent of a glioma. Additionally, a T1-weighted post-gadolinium contrast-enhanced scan may also be obtained to exclude new contrast enhancement, indicative of high-grade transformation [69]. These conventional MRI tech-niques, however, often underestimate the distribution of malignant cells within the brain. Contrast enhancement on T1-weighted MRI reflects only ar-eas of blood–brain barrier breakdown, and therefore the extent of tumor cells infiltrating around the tumor margins is not displayed [69]. T2-FLAIR images are used to identify non-enhancing tumor, but the diffuse infiltration of glioma makes the differentiation of heterogeneous glioma tissue from nonspecific changes unreliable [70], leading to the underestimation of the actual tumor extent of gliomas [71]. Moreover, conventional MRI does not provide reliable information on tumor physiology such as microvascularity, angiogenesis, me-tabolism, micronecrosis, or cellularity [67]. Therefore, additional information is needed. Advanced imaging methods such as perfusion MRI (pMRI) and positron emission tomography (PET) with amino acid tracers provide this functional, physiological, and molecular diagnostic information. The tumor border and the tumor growth pattern are important to define by noninvasive imaging, aiding a more accurate tumor treatment planning in low-grade glio-mas. This is the focus of paper II, while the relation between histological characteristics and conventional and advanced radiological parameters is the focus of paper III.

35

Magnetic Resonance Imaging MRI, including T1‑weighted and T2‑weighted sequences, is currently the standard method for diagnosis and differential diagnosis of brain tumors. MRI has excellent soft-tissue contract and high spatial resolution to detect space-occupying lesions in the brain [70, 72], providing anatomic information. Low-grade gliomas show a low signal on T1- and a high signal on T2-weighted imaging. T1 is the longitudinal relaxation time, showing cerebrospinal fluid (CSF) as a dark area. T2 is the transverse relaxation time, showing fluid like CSF in white matter. T2-weighted images are preferable with regard to defin-ing tumor margins [73]. The T2-FLAIR pulse sequence is the best radiological method for mapping the extent of a glioma (Figure 5a). By carefully choosing the inversion time, the FLAIR method nullifies signals from fluids, such as the CSF. However, as mentioned previously it is still very difficult to deter-mine tumor grade, to discriminate between gliomas (i.e. to separate astrocyto-mas from oligodendrogliomas), or to define the exact boundaries of a glioma due to its infiltrating character [74]. Usually, the presence of larger areas with contrast enhancement seen on MRI is regarded as a sign of high-grade tumor. However, the absence of contrast enhancing areas does not exclude a high-grade tumor. Previous research showed that one-third of non-enhancing tu-mors, presumed to be low-grade gliomas, turned out to be high-grade gliomas after biopsy or surgery [75]. Therefore, correlation of tumor characteristics on MRI with histological findings from stereotactic biopsies of DLGGs has been the focus of several previous studies [76-79]. In these studies, infiltrating tu-mor cells were found in normal appearing T1 and T2 areas, in other words outside the radiologically defined tumor border. These results demonstrate the importance of more sensitive radiological tumor assessment and new imaging techniques.

Advanced imaging Advanced MRI methods such as pMRI and PET using amino acid tracers can provide additional functional, physiological, and molecular diagnostic infor-mation in gliomas. [72, 80]. In pMRI, a paramagnetic contrast agent is injected and the passage of a bolus of the contrast agent is observed via T2-weighted imaging, a technique known as dynamic susceptibility contrast (DSC) [72]. A valuable addition of pMRI to conventional MRI is the assessment of the rela-tive cerebral blood volume (rCBV) due to its proven correlation with the grade of neovascularization [81] and therefore its use in tumor grading and determi-nation of tumor progression (Figure 5b). However, there are some methodo-logical limitations to rCBV imaging such as standardizing image acquisition and the processing and interpretation of rCBV due to its high sensitivity to artifacts [82].

36

A novel approach PET using radiolabeled amino acids provides additional insight, beyond MRI, into the metabolic activity of the tumor (Figure 5c). The use of amino acid PET imaging, in addition to MRI, for brain tumor management is now recom-mended [83]. The uptake of amino acid tracers, representing metabolic activ-ity, is low in normal brain tissue but upregulated in cerebral tumors. This is due to the increased transport of large neutral amino through the plasma mem-brane, via the L-type amino acid transporter (LAT) system [84]. The high tu-mor-to-background contrast can help to differentiate tumor progression from non-specific treatment-related changes [85]. The longest-established amino acid tracer for PET is 11C methyl-l methionine (11C-MET). Hot spot areas of amino acid uptake generally represent the most malignant part of the tumor [86]. This information can improve planning of biopsy, neurosurgical resec-tion, and radiotherapy (RT), to spare healthy tissue [87]. Studies using serial stereotactic biopsies showed that 11C-methionine (MET) PET was superior to CT or MRI with regard to delineating the tumor [88], improving the tumor volume definition in 88% of low-grade gliomas and 78% of high-grade glio-mas [89]. Additional research showed that the combination of PET and MRI during diagnostic biopsy was more accurate than MRI alone (specificity of 94% and 53%, respectively) [90]. Recently, the tracer 18F- fluoroethyl-L-tyro-sine (FET) has received increased attention [91]. FET has logistic advances over MET due to its longer half-life and is now widely used for the clinical management of gliomas [72]. To date, there have been a number of PET stud-ies that compared histological data provided by biopsy or surgery with the amino acid uptake [88, 90, 92, 93], demonstrating that the uptake reflects the solid mass of gliomas and metabolically active tumor areas. Although the combination of MRI and PET has proven to be more specific than MRI alone, , an exact correlation has not yet been described between imaging findings in MRI, the metabolic changes of amino acid uptake, and histological character-istics in these heterogeneous tumors with the use of biopsies. The relation be-tween histological characteristics and conventional and advanced radiological parameters was the focus of paper III, where a new method is introduced for direct local comparison of subsequent areas.

Figure 5. (a) T2-weighted FLAIR MRI shows a tumor in the right frontal lobe. (b) DSC perfusion MRI with rCBV greyscale map corresponding to the PET hotspot (c) MET PET shows the hotspot region of the tumor.

37

Part V Clinical aspects and seizure semiology

38

Seizure semiology of gliomas One of the first tumor-related symptoms of a cerebral lesion is an epileptic seizure [94]. Because of the presence of an enduring alteration in the brain (i.e. the tumor) [95]. The incidence of epilepsy and epileptic symptoms de-pends on several aspects such as tumor subtype, grade, and location. First, seizures are classified into focal onset, generalized onset, and unknown onset [96]. Second, epilepsy types are classified as focal, generalized, combined generalized and focal, and unknown [97]. Epilepsy can include multiple types of seizures, which can be seen on electroencephalography. In focal epilepsies, unifocal and multifocal disorders can include a range of seizure types such as focal aware seizures, focal impaired awareness seizures, focal motor seizures, focal non-motor seizures, and focal to bilateral tonic-clonic seizures [96]. Fo-cal onset seizures are most common (60–95% [94]), and are divided into aware and impaired awareness seizures [97]. Generalized epilepsies have no apparent cause, including a range of seizure types such as absence, myoclonic, atonic, tonic, and tonic-clonic seizures [96]. In regard to tumor grade, slow growing low-grade glioma are more epileptogenic than high-grade gliomas [98]. Seizures with no impairment of consciousness are most common in pa-tients with high-grade gliomas. Impairment of consciousness, followed by in-voluntary movements or automatisms are more frequent in low-grade gliomas [99]. Regarding tumor subtype, epileptic seizures are more common in oli-godendrogliomas than in astrocytomas [100]. The brain has a cortical adaptor mechanism that makes it possible for it to reorganize itself in the case of slow- growing low-grade gliomas owing to its plasticity. In the case of high-grade gliomas, the brain does not have or has only a limited ability to adapt, causing more neurological deficits in patients with fast growing brain tumors, such as GBMs [101, 102]. In general, seizure semiology is said to reflect the specific location of the tu-mor related to the organization of cortical brain functions [103]. The most common location of a LGG is the frontal lobe which is strongly associated with asymmetric tonic seizures [104], causing speech arrest or motor agitation. Patients with temporal lobe seizures typically report symptoms consisting of déjà vu phenomena and visceral sensations such as epigastric rising, gustatory or olfactory auras [99], which are very similar to the epileptic phenomena re-lated to the insular cortex. Sensorimotor deficits, cognitive impairment, and language disorders are often more common in elderly patients with LGG, demonstrating the age-related differences [105] and their importance for on-cological treatment. At present, the exact origin and mechanisms of human glioma-related epilepsy are unknown, although it has been suggested that gli-oma growth stimulates seizures, and that seizures encourage glioma growth [106]. Therefore, antiepileptic drugs, surgery, radiotherapy and chemotherapy can significantly improve seizure control and glioma growth [107].

39

Brain connectivity in epilepsy Understanding the connectivity of the human brain is of importance for pre-dicting tumor behavior and the planning of biopsies and treatment. Studying forms of epilepsy can shed a new light on current perspectives about the brain, and provide information regarding the yet unknown origin and mechanisms of human tumor-related epilepsy. A very rare form of epilepsy is the state of consciousness during focal epilepsy with ecstatic seizures. Patients experienc-ing this form of epilepsy describe feelings of absolute well-being, intense se-renity, bliss, and “enhanced self-awareness” [108]. A feeling not comparable to any other stimuli. At present, there are only 52 known cases described in the literature of patients who have experienced ecstatic seizures [108]. To un-derstand the functional connectivity of the human brain, attempts have been made to identify the anatomical structures of the human brain [109] (Figure 6), as well as the structures involved in this complex semiology of ecstatic seizures [108, 110].

Previous research has provided conflicting results with no clear lateralization of brain lesions to one hemisphere of the brain as the origin of the ecstatic seizures. However, there are studies suggesting an emotional lateralization of the brain with favorable euphoric-left and dysphoric-right distribution in the brain [111]. Previously, the origin of ecstatic seizures was linked to one spe-cific location, the temporal lobe [110, 112]. Modern technical advances in multimodal brain imaging have provided new evidence pointing to other re-gions [113]. Despite the notion that seizure semiology is known to reflect the specific location of the tumor and its organization of cortical brain functions, this phenomenon seems rather to be a combination of several functional net-works (e.g. sensorimotor, auditory, visual, attentional, salience, or default mode networks) [114-118]. This suggests that ictal activity of ecstatic epilep-tic seizures is a dynamic interplay between the anatomical region of seizure onset and connected target areas within the network [119, 120], providing ev-idence for a rather complex functional connectivity behind each symptom [121-127]. A better understanding of human brain connectivity and its relation to specific forms of epilepsy therefore has an impact on tumor treatment plan-ning. In order to provide a better understanding of the anatomical networks of the human brain, we studied this rare form of epilepsy in a unique opportunity described in paper IV.

Figure 6. White matter fiber archi-tecture from the Connectome Scan-ner data set. Image subtracted from http://www.humanconnectomepro-ject.org

41

Part VI Therapeutic management

42

Glioma treatment As previously mentioned, the natural history of diffuse low-grade gliomas var-ies greatly. Therefore, tumor treatment has no optimum management and needs to take into account a range of tumor-related and treatment-related fac-tors that could influence clinical outcomes [128]. Tumor histology based on examination of a biopsy of tumor tissue is a primary predictor of patient prog-nosis. Stereotactic biopsy guided by structural and metabolic imaging (MRI/PET) is today the most common method used due to its high reliability and low-risk. This new method is less invasive and has greater precision than open biopsy via craniotomy [129]. Prognosis according to the histological in-formation determines therapy. Tumor treatment is based on different strate-gies such as watchful waiting, radical surgery, radiotherapy, chemotherapy, or a combination of these [107]. Subtype specific treatment of gliomas is shown in Figure 3.

Surgery Because of their growth pattern (Figure 7), gliomas of grades II, III, or IV cannot be cured by resection. However, surgery can improve neurological def-icits, seizure control, and achieve cytoreduction of the tumor mass with the aim of prolonging patient survival. Nowadays, this is done through the re-moval of all T2/FLAIR-hyperintense tissue (in WHO grade II and III tumors), or of all contrast-enhancing tissue (in GBM), while minimizing neurological risk for the patient [129]. Early and maximum tumor resection has a favorable impact on survival by delaying malignant transformation from low- to high-grade gliomas [130]. Supratotal resection, that is, resection of the tumor to-gether with a margin outside the radiological border, has additional benefits [131]. Tumor cells have been shown to be found up to 26 mm beyond MRI-defined abnormalities, where the supratotal resection is suggested to remove these scattered tumor cells outside the radiological defined tumor border [79]. In primary treatment of WHO grade II and WHO grade III gliomas, the max-imal resection of more than 90% tumor tissue, that is visible in T2/FLAIR weighted MRI, improved the 5- and 8-year overall survival of patients to 97% and 91%, respectively, compared to 76% and 60% with a lesser extent of re-section [129]. However, the infiltrative character of low-grade gliomas to-gether with the limited plasticity of the white matter represents a major chal-lenge to the resection of tumors with infiltration of subcortical pathways. Ad-ditionally, the invasion rate of glioma cells in the white matter is about five times higher than in the gray matter [132]. In GBM, complete removal of tis-sue with a disrupted blood brain barrier, and additional radiotherapy, im-proved overall survival from 12 to 15 months [129, 133]. Despite the benefi-cial effect of early and supratotal resection on survival, brain surgery itself may cause neurological impairment, cognitive disabilities, and impaired qual-ity of life. Especially tumors infiltrating eloquent areas of the brain require

43

selective removal of tumor tissue without affecting the surrounding nonma-lignant brain tissue [128]. After surgical treatment, follow-up MRI is per-formed every 4 to 6 months in patients with WHO grade II or grade III tumors and every 3 months in patients with WHO grade IV tumors, depending on their clinical condition.

Figure 7. Illustration of growth pattern of diffuse gliomas [134]. Left: Histologically diffuse low-grade glioma (Astrocytoma, Oligodendroglioma, Mixed glioma). Right: High-grade diffuse glioma (Glioblastoma, Anaplastic Oligodendroglioma).

Radiotherapy For progressive and inoperable gliomas, as well as to prevent recurrences of gliomas, local radiotherapy is given. The main goal of radiotherapy in general is to achieve cell death by damaging DNA. Local radiotherapy of the tumor region or tumor bed is performed in the immediate vicinity of the primary tumor [28, 129]. Fractionated radiotherapy of gliomas is now a standardized method, with total radiation doses of 50 Gy (for WHO grade II tumors) and 60 Gy (for WHO grade III and IV tumors) [129]. Treatment time is 5 to 6 weeks, with a daily individual doses of 1.8 to 2.9 Gy 5 days per week. De-pending on age, a briefer course of hypofractionated radiotherapy is per-formed. In patients over age 60, a 10 x 3.4 Gy is given over a period of 2weeks [129].

Chemotherapy For gliomas, chemotherapy with the alkylating drug TMZ is the standard chemotherapeutic drug for treatment. TMZ is an oral alkylating agent causing impaired DNA replication and triggers cell death. In addition, a combination of procarbazine, lomustine, and vincristine (PCV treatment) is used. Adjuvant TMZ following radiochemotherapy (RCT) has been showed to be superior to PCV in the treatment of anaplastic astrocytomas and GBMs (the so-called Stupp protocol [135]), while PCV is used mainly for oligodendroglial tumors [129]. Oligodendrogliomas with a 1p/19q codeletion show a high-sensitive response to chemotherapy with PCV [136, 137], underlining the importance of molecular parameters in glioma treatment.

44

Informed decisions combining all possible information are therefore of high importance for tumor treatment. In paper II and paper III, we studied tumor behavior, such as tumor growth pattern and tumor invasion, in order to im-prove tumor treatment planning and to deepen our understanding of advanced imaging techniques. In paper IV, we describe a rare form of epilepsy to pro-vide a deeper understanding of the human brain connectivity.

45

Part VII Present investigations

46

Aims of this thesis This thesis has several aims intended to deepen our understanding of tumor growth, radiological tumor assessment, and anatomical connectivity of the brain, to benefit precision medicine. The aim of this thesis was to:

1. Provide further evidence of the importance of molecular classification by demonstrating PROX1 as a novel pathway-prognostic marker in high-grade astrocytomas.

2. Detect tumor cells outside the radiological border in LGGs.

3. Provide a deeper understanding of the metabolic imaging parameters and perfusion, and compare these with histopathological tumor char-acteristics throughout the whole tumor.

4. Provide evidence for a possible anatomical network behind a complex epileptic syndrome.

47

Scientific questions Paper I PROX1 has been ascribed suppressive as well as oncogenic roles in several human cancers, including brain tumors. Does PROX1 expression correlate with patient survival, not only in low-, but also in high-grade astrocytomas? Is there an age dependency? Are there pathway-specific differences?

Paper II Given that MRI tends to underestimate the extent of DLGGs, how far can cells be found outside the radiological border? How does radiology reflect the tu-mor growth pattern? What are the clinical implications of these findings?

Paper III Oligodendrogliomas are heterogeneous tumors in terms of imaging appear-ances. How is the tumor heterogeneity reflected on 11C-methionine (MET) uptake? Which histological and radiological parameters show strongest corre-lated with MET uptake? What are the clinical implications of these findings?

Paper IV The origin of ecstatic seizures has previously been linked to several areas in the brain. Can ecstatic epilepsy be designated to a single trigger zone? Or is it rather a dynamic interplay of different areas? How does the brain connectivity reflect seizure semiology changes?

48

Paper I: A pathway-specific prognostic marker In the present study, we investigated whether Prox1 is a prognostic factor for patients with high-grade astrocytomas, something that has previously been demonstrated in low-grade gliomas [66]. In three consecutive steps, we showed that PROX1 protein levels correlated with survival for pa-tients with IDH-mutant high-grade astrocytomas, but not for patients with primary GBMs. These findings were confirmed by PROX1 gene ex-pression analysis in tumors extracted from The Cancer Genome Atlas (TCGA) that were 1p19q non-codeleted and stratified by IDH mutation status. Together with the previous studies in low-grade gliomas, we con-clude that PROX1 is a pathway-specific marker for IDH-mutant, 1p19q non-codeleted astrocytomas that may be used to identify high-risk pa-tients. To date, there are few clinically useful markers to predict survival and/or in-dividual response to treatment, or guide clinical management of patients with high-grade gliomas [138]. Like other homeodomains containing proteins, Prox1 is said to be able to both activate and repress transcription of genes in a context- and tissue-dependent manner. Prox1 is shown to be involved in de-velopmental processes such as cell fate determination, gene transcriptional regulation, and progenitor cell regulation [55]. However, the relationship be-tween Prox1 and cancer is complex, and Prox1 can exhibit either tumor sup-pressing or oncogenic properties, depending on cancer type. Patient age is not only associated with glioma incidence [139], but also a pre-dictor for survival in glioma types [140]. Additionally, there is increasing ev-idence for fundamental molecular differences between different age groups with high-grade gliomas. Age-related changes in the microenvironment of neuro-progenitor cells are thought to contribute to the increased cancer risk in the older brain [141]. As mentioned earlier, Prox1 has been assigned a role as both a diagnostic as well as a predictive marker in gliomas. Protein expression of PROX1 is upregulated in high- compared to low-grade gliomas, whereas a higher expression in low-grade gliomas been shown to be associated with poorer survival. Since the survival benefit of Prox1 has not yet been estab-lished in higher grade gliomas, we asked ourselves the question whether Prox1 is a predictive marker in high-grade gliomas as well. Therefore, in this study we investigated the prognostic role of the transcription factor Prox1 in differ-ent cohorts of patients with GBM. In order to study the survival effects of PROX1 gene expression, we examined a total of five different cohorts. 1) 86 unselected high-grade astrocytomas tu-mor samples from a representative group of IDH1 immunopositive (mIDH1R132), and IDH1/2-mutated 1p19q non-codeleted tumors in our tis-sue microarrays (TMAs) (Figure 8). 2) 174 IDH1- R132H1 immunonegative

49

GBMs derived from patients aged 60 years and older enrolled in the Nordic phase III trial of elderly patients with newly diagnosed GBM. 3) 80 IDH-wildtype GBMs from a different cohort of patients aged 18–60 years. 4) A combined focus on the population of IDH-mutant anaplastic astrocytomas and GBMs. We selected 19 IDH-mutant 1p19q non-codeleted high-grade astrocy-tomas from the younger cohort, and 15 high-grade astrocytomas immunoposi-tive for mIDH1R132 from the trial in the elderly. 5) To verify our results, we extracted 34 IDH-mutant 1p19q non-codeleted anaplastic astrocytoma and GBMs from TCGA.

Figure 8. TMA of high-grade astrocytoma showing PROX1 expression.

We initially screened 86 unselected high-grade astrocytomas. As expected, we found the age of the patient to be a prognostic factor for survival in our cohort of unselected high-grade astrocytomas, whereas, surprisingly, PROX1 expres-sion did not seem to have any benefit with regard to survival.

These findings in the screening cohort prompted us to study two age-specific cohorts (younger patients (≤ 60 years) with IDH-wildtype GBM and elderly patients (> 60 years) with newly diagnosed GBM) to determine a possible age-

50

related benefit of PROX1 protein expression on patient survival. In neither of the age-specific cohorts were we able to demonstrate a possible benefit of PROX1 protein expression on survival.

We then questioned whether survival benefit related to PROX1 expression might be pathway dependent, and studied the combined population of IDH-mutant anaplastic astrocytoma and GBMs from both the younger cohort and the trial in the elderly. Here, we found that patients with high PROX1 protein expression showed statistically significantly shorter survival, demonstrating a possible pathway-specific effect of Prox1 on survival. On the other hand, low PROX1 protein and younger age were associated with statistically signifi-cantly longer survival. To validate the findings for PROX1 in IDH-mutated high-grade astrocytomas, we used TCGA database (http://gliovis.bioinfo.cnio.es/). As expected, PROX1 gene expression was identified as an independent prognostic factor together with MGMT promoter methylation and WHO grade in IDH-mutant astrocytomas grade III and IDH-mutant GBMs. Additionally, age was identi-fied as an independent factor for prognosis in IDH-wildtype GBMs derived from TCGA, whereas PROX1 gene expression was not correlated with sur-vival in these tumors.

To summarize, our cohorts provide evidence that PROX1 is a predictor for survival in high-grade 1p19q non-codeleted astrocytomas that are IDH-mutant and are most common in younger patients with a history of prior low-grade gliomas (secondary GBM). Our study does not provide evidence that PROX1 is an age-dependent prognostic marker in primary GBMs.

In conclusion, the transcription factor PROX1 with a widespread role in CNS development is a prognostic marker for IDH-mutant anaplastic astrocytomas (secondary GBMs) that evolve through multi-step genetic alterations over time. The diverting roles of PROX1 in IDH-mutant and IDH-wildtype astro-cytomas may be related to different epigenetic regulatory mechanisms be-tween these tumor types. IDH-wildtype GBMs predominate in elderly patients who generally have a higher degree of genomic methylation [142]. Promoter hypermethylation may be a regulatory mechanism for PROX1 that is tumor type specific, depending on cellular context and specific signaling pathways [143]. As mentioned earlier, although PROX1 protein levels are lower in diffuse low-grade gliomas than in high-grade gliomas, there is a subset of low-grade gliomas with up-regulated PROX1 expression showing worse prognosis and more rapid tumor progression [65, 66], indicating, also here, a pathway-specific effect of PROX1. Together with the previous study in low-grade gliomas [66], we propose that PROX1 is a clinically valuable marker to predict the course of disease in patients with astrocytomas. Future studies are

51

needed to clarify the role of PROX1 and the mechanisms by which PROX1 leads to tumor progression and affects clinical outcome in this patient group. Identification of distinctive molecular pathways can introduce new concepts for therapeutic management based on the clinical diversity of these tumors and provide useful information that will aid in the discovery of novel cellular and molecular mechanisms involved in adult neurogenesis. Additionally, whether PROX1 is a lineage-specific prognostic marker (i.e. for astrocytomas but not for oligodendrogliomas) is still unknown and needs to be studied further. We extracted preliminary data from TCGA to display a statistically significant correlation between PROX1 gene expression and sur-vival for patients with IDH-mutant, 1p19q non-codeleted low-grade astrocy-tomas, but not with IDH-wildtype or 1p19q codeleted low-grade gliomas. These results may indicate that the prognostic impact of PROX1 in low-grade gliomas is also pathway dependent, consistent with our findings in high-grade astrocytomas. TCGA is a collection of comprehensive, multi-dimensional maps of the key genomic changes in 33 types of cancer (http://cancergenome.nih.gov/). It is the result of a collaboration between the National Cancer Institute (NCI) and the National Human Genome Research Institute (NHGRI). TCGA dataset has publicly available genomic data to help the cancer research community im-prove the prevention, diagnosis, and treatment of cancer. Several studies have been conducted based on TCGA database. Here, we used TCGA database as a verification and validation of our patient cohorts.

52

Paper II: Tumor growth and invasion In this study, we developed and applied a new method for the co-registra-tion of MRI and histological data in DLGG. We found that this method is a useful aid in evaluating the growth pattern of these tumors in relation to radiological findings because it illustrates the characteristic tumor in-filtration of the white matter tracts and the extension of tumor cells out-side the radiological border. Agreement between histopathology and noninvasive imaging parameters pre-vents misclassification and undergrading during biopsy or partial tumor resec-tion [144, 145] and is needed in clinical decision making. Additionally, surgi-cal resection, biopsy, and radiation therapy planning in cerebral gliomas de-pend on reliable noninvasive distinction between tumor tissue and the peritu-moral brain tissue. Therefore, with the method presented in this study, the sensitivity of several imaging techniques with regard to detecting infiltrating tumor cells can be further evaluated. Diffuse infiltration of glioma (Figure 7) makes the differentiation of heterogeneous glioma tissue from nonspecific changes or surrounding edema unreliable [70], leading to the underestimation of the actual tumor extent of gliomas using T1‑weighted and T2‑weighted MRI sequences [71]. An attempt to correlate tumor characteristics on MRI and the histological findings in DLGGs has been the focus of previous studies [76-79]. Despite clear radiological tumor borders, tumor cells were found in normal-appearing brain outside these borders. However, these studies were performed using biopsies or partial tumor resection only. In our research, we describe a more direct method for co-registration of histological data with MRI of DLGGs removed en bloc, which has not been previously reported.

We correlated MRI findings with histological data to study the presence of tumor cells out-side the radiological border in patients with suspected DLGGs in whom en bloc resections were performed (Figure 9). Five patients were recruited from a longitudinal follow-up study of adults with DLGGs at our hospital [74]. The complete en bloc tumor resection included a margin of normal appearing brain located outside the radiological border. Neuropathological assessment confirmed DLGG in four patients and revealed progression to WHO Grade III glioma in one patient. Prior to surgery, patients were examined with 3T MRI (Philips Achieva). Mor-phological MRI sequences including sagittal and axial T2-weighted turbo spin echo (SE), coronal and axial T2-FLAIR, axial T1-weighted SE before and af-ter contrast injection, and sagittal T1-weighted 3D turbo field echo after con-trast injection [74]. Radiological tumor borders were manually delineated on

Figure 9. En bloc tumor re-section from our studies.

53

T2-FLAIR sequences at the transition between high signal intensity and nor-mal parenchymal signal intensity based on visual evaluation. In addition to the histopathological diagnostics, immunostaining with IDH1-R132H was per-formed on all histological sections. For the co-registration, pre- and postoper-ative coronal MR images were oriented in the same direction as the histolog-ical sections. Then, the scanned IDH1 images were overlaid on the coronal radiological images using Photoshop CC (Adobe Systems Software), and the radiological tumor border was manually marked on the corresponding IDH1 image. We found that the histological tumor core (area with the highest density of tumor cells) was located in both gray and white matter. Most importantly, we demonstrated that in all samples, tumor cells were found outside the radiolog-ical border at a maximum distance of 1.4 cm from the radiological border (Figure 10). In one sample, an area with high tumor cell density and low signal on T2-FLAIR MRI was found outside the radiological border in the white matter (Figure 11, Left), which had not been detected on preoperative FLAIR sequences (Figure 11, Right). In all cases, the infiltrating tumor cells followed the white matter tracts. Tumor cells were concentrated in the periphery of the tracts, with the highest density in the most peripheral parts. Tumor cells did not cross the pia mater in the sulcus, resulting in a notable difference in tumor cell density in the gray matter of two adjacent gyri. The co-registration of histological data with MRI of DLGGs removed en bloc as described here has not been previously reported, but the correlation between the tumor characteristics on MRI and the histological findings in DLGGs has been the focus of previous studies [76-79]. In these studies, tumor cells have been demonstrated outside the radiological tumor borders, which is confirmed in our study using a detailed comparison between the histological properties of specific intra- and peritumoral regions and the corresponding signal changes on MRI. We showed a relatively large area of high-density tumor cells outside the radiological border, which had appeared normal on preoper-ative T2-weighted FLAIR images (Figure 11), supporting the notion that FLAIR sequences are not sensitive enough to accurately detect the infiltration of tumor cells. Infiltrating tumor cells were detected in the white matter tracts in all cases in the present study, in line with the previous studies [146-148]. Interestingly, in two of the cases, the white matter tracts were only partly infiltrated, mostly in the peripheral fibers of the white matter tracts. It is likely that this limited infiltration of the white matter reflects an early stage of tumor invasion. To spread to the adjacent gyrus, tumor cells followed the white matter tracts. They reached the molecular layer of cortex but did not infiltrate the pia mater. It is discussed whether short and intracortical U-fibers are the most probable way of diffusion for tumor cells between two adjacent gyri [149-151]. Differences

54

in tumor genotype and phenotype are associated with different invasion pat-terns in DLGG [152]. It has previously been shown that oligodendrogliomas have a preference for the white matter regions, while low-grade astrocytomas are more often located close to the lateral ventricles [153]. In addition, the site of origin is also thought to affect the growth properties of the tumor [148, 154, 155]. Previous research demonstrated that different proteins can guide the mi-gration of cells [152, 156, 157]. For example, metalloproteinases, expressed by the tumor cells, may enhance the spread of glioma cells in the white matter [158, 159]. Tumor treatment is dependent on a better understanding of the behavior of tumor cells beyond the radiological border, enhancing the possibility of per-forming a radical tumor resection. Our findings indicate that the extension of surgery for DLGGs should not be limited to the radiological border as deline-ated on T2-FLAIR MRI sequences, and they support the concept of supratotal resection, which is also supported by previous studies [160]. Performing su-pratotal resection is crucial for improving the prognosis of patients with DLGG because it has a positive influence on the time to anaplastic transfor-mation [23, 107, 161-164]. Finally, large tumor resections provide more reli-able representative samples for histopathological tumor diagnosis. When tis-sue loss is minimized and the whole tumor can be analyzed, the probability of a correct diagnosis with respect to tumor grade and focal signs of anaplastic transformation is augmented, preventing misclassification [165]. These results show the importance of a combination of radiological and histological tumor characteristics. Molecular and radiological parameters may become more im-portant during radiological tumor assessment and tumor treatment. Additionally, the tumor is expected to be visible on T2-FLAIR MRI if the tumor cell density is above a certain value, that is, the “visibility threshold.” In addition to the cell density, the MRI signal is influenced by several other factors, including the water content of the tissue [166]. Another potential tool for identifying infiltrative tumor cells is PET using amino acid tracers. Due to the low background activity of 11C-l-methionine, these tracers are considered reliable options in detecting the most malignant portion of the tu-mor.

Figure 11. Radiological tumor border (blue line) and histological tumor border (dotted line).

Figure 10. Tumor cells outside the radiological border.

55

Paper III: Intra-tumoral heterogeneity In this study, we further developed our method of co-registration, and investigated intra-tumoral 11C-methionine (MET) uptake in relation to tumor perfusion and the protein expression of histological cell markers in corresponding areas of oligodendrogliomas. Using PET-to-MRI-to-Histology co-registration, we demonstrate that the MET uptake through-out the whole tumor reflects the tumor cell density of oligodendrogliomas. Vessel density and the density of proliferating cells are less stringently correlated with MET uptake and are probably more susceptible to arti-facts caused by larger blood vessels surrounding the tumor.

Oligodendrogliomas are heterogeneous tumors in terms of imaging appear-ances, whereas diffuse infiltration of glioma makes the differentiation of het-erogeneous glioma tissue using conventional MRI unreliable [70]. PET with amino acid tracers circumvents this problem since the uptake of amino acid tracers is low in normal brain tissue but upregulated in cerebral tumors [84]. This high tumor-to-background contrast can help to differentiate tumor pro-gression from non-specific, treatment-related changes [85]. Several studies have tried to compare histological data with the amino acid uptake [88, 90, 92, 93], these studies demonstrate that the uptake reflects the solid mass of glio-mas and metabolically active tumor areas. Additionally, the uptake of amino acid tracers measured by PET has been suggested as an indirect measure of microvessel density [167], proliferation, tumor cell density, and vessel density [168-175]. However, these studies are contradictory and mostly based on bi-opsies, while an exact correlation between imaging findings in MRI, meta-bolic changes of amino acid uptake, and histological characteristics in these heterogeneous tumors has not yet been described.

In order to provide the proof-of-concept of how MET uptake and rCBV reflect the cellular heterogeneity in oligodendrogliomas, we evaluated the entire vol-ume in en bloc resected oligodendrogliomas. We further developed our method used in paper II to perform a region-by-region comparison of histo-logical and advanced radiological parameters. We used consecutive histolog-ical sections in a series of en bloc resected oligodendrogliomas, described pre-viously. We analyzed the intra-tumoral uptake of the amino acid tracer 11C methyl-l methionine (11C-MET) uptake in PET, tumor perfusion (rCBV) on MRI and the protein expression of histological cell markers. First, we immunostained all sections of four tumors covering the entire en bloc removed tumor, with antibodies against IDH1-mutated protein (tumor cells), Ki67 (proliferating cells) and CD34 (blood vessels). The subsequent histolog-ical images were co-registered with each other, using an in-house developed software, to quantify protein expression of the histological cell markers. Here-after, the histological images were overlaid on corresponding MET PET scans

56

and MRI perfusion maps. We then selected regions of interest (ROIs) on PET throughout the entire tumor volume. The ROIs covered hot spot areas (known to generally represent the most malignant part of the tumor [86]), areas adja-cent to hot spots, and tumor borders with infiltrating zone. We also determined MET uptake (tumor-to-normal tissue ratios) and mean relative cerebral blood volume (rCBV) in these ROIs, which were compared to the protein expression of histological cell markers in the corresponding regions.

We first investigated whether MET uptake throughout the whole tumor re-flects tumor cell density using a total of 84 ROIs in four oligodendrogliomas. In our study, we found that MET uptake was indeed correlated with tumor cell density in each of the tumors studied. By using the antibody against IDH1-mutated protein as a marker for tumor cells, we confirmed previously estab-lished correlations between MET and cell density [86], and extended these findings to a correlation between MET and tumor cell density in oligodendro-gliomas. Since the development and maintenance of an adequate blood supply are es-sential for tumor growth and invasion [175-177], cell proliferation and angio-genesis have been suggested to explain the generally higher MET uptake in oligodendrogliomas compared to astrocytomas [86, 178]. Indeed, we showed that MET uptake correlated with vessel cell density, as well as proliferating cell density, in two tumors with consistently decreasing rCBV values in rela-tion to the tumor core. Our results confirm previous findings and show that MET uptake is a measure for not only tumor cell density, but also microvessel density in regions outside the hot spot and in peritumoral areas. In the other two tumors, we noticed an unexpected increase of rCBV values in the peritu-moral region. The lack of significant correlations between MET-CD34 and MET-Ki67 was due to partial volume effects from adjacent cortex and large vessels in this region. It is likely that the increased count for microvessels and proliferating cells in the tumor periphery are not true histological representa-tions, explaining the lack of statistically significant correlations between MET-CD34 and MET-Ki67 in these tumors.

Figure 12. Co-registration of MRI, PET and corresponding histological images with defined ROIs (ROI1, red and orange; ROI2 yellow, ROI3 green)

57

To conclude, we complemented previous reports correlating MET uptake with histological parameters. We demonstrate that the MET uptake in hot spots, outside hotspots, and in infiltrating tumor edges unanimously reflects the tu-mor cell density of oligodendrogliomas. The correlation between MET uptake and vessel density and density of proliferating cells is less stringent in infil-trating tumor edges and is probably more susceptible to artifacts caused by larger blood vessels surrounding the tumor. A better understanding of nonin-vasive tumor distinction could improve the planning of biopsy, neurosurgical resection, and radiotherapy and to a greater extent to the sparing of healthy tissue [87].

The novelty of the present study lies in the use of en bloc resected tumors, representing the concept of supra-total tumor resections, whereas previous studies have all been based on limited, less representative tumor tissue [93]. Our model also illustrates a novel way for evaluating imaging parameters with respect to the intra-tumoral heterogeneity of gliomas.

58

Paper IV: A dynamic interplay In this study, we investigated a possible network involved in a rare form of epilepsy. We describe changes in ecstatic seizure semiology and MRI connectivity during three different time periods, prior to and after the surgical removal of a hypothalamic hamartoma. We carefully suggest that a balance of the networks between the thalamus, the brain stem, and the frontal lobe is crucial with regard to experiencing gelastic and ecstatic seizures. We hypothesize a possible connection between gelastic and ec-static seizures, where it seems that the frontal lobe functions as a negative reinforcer, while activation of the brain stem in combination with the gelastic component is a possible positive reinforcer of an ecstatic seizure. This new information about the possible organization supports the hy-pothesis that the complex pattern of symptoms seen in patients with ec-static epileptic seizures is a network, rather than just a single trigger zone. As mentioned in section V, epileptic seizures are one of the first tumor-related symptoms [94] that are due to an alteration in the brain (i.e. the tumor) [95]. Complex focal seizures involve impairment of consciousness, followed by in-voluntary movements or automatisms [99], generally reflecting the specific location of the tumor related to the organization of cortical brain functions [103]. For example, the frontal lobe is strongly associated with speech arrest or motor agitation [104]; the temporal lobe with symptoms consisting of déjà vu phenomena and visceral sensations such as epigastric rising and gustatory or olfactory auras [99], similar to the insular cortex. At present, the exact origin and mechanisms of human tumor-related epilepsy is unknown, and re-search is difficult due to the plasticity of the brain, which makes it possible for the brain to can reorganize itself in the event of slow growing tumors. We have mentioned before that it is important to understand the connectivity of the human brain in order to predict tumor behavior, planning of biopsy, and tumor treatment. Therefore, rare forms of epilepsy can shed a new light on current perspectives about the brain and provide information on the yet un-known origin and mechanisms of human tumor-related epilepsy. In this paper, we studied a unique opportunity provided by a case of ecstatic epilepsy to reveal a possible network behind this epilepsy. We describe a clin-ical/neurological longitudinal follow-up study of a patient with a history of gelastic and ecstatic seizures. A gelastic seizure is a “desire or pressure to laugh”, that often occurs without any “funny” emotion, in contrast to ecstatic seizures [179]. The uniqueness of this study lies in the disappearance and later reappearance of the ecstatic phenomenon in relation to the treatment of a hy-pothalamic hamartoma. We merged data from MRI investigations performed at three different time points with clinical information on presurgical semiol-ogy and its development 1 year and 11 years after surgery. Normally, neural

59

tracts could be visually represented using a 3D modeling technique called trac-tography based on diffusion-weighted images (DWI). Artifacts caused by a vagus nerve stimulator prevented us from using this technique. To circumvent this problem, we used a previously described tractographic template [109] to reconstruct the white matter connectivity in relation to the tumor lesion. Therefore, we created a ROI of the lesion volume from the native MRI se-quences within the Montreal Neurological Institute (MNI) space. We then re-constructed the white matter bundles, according to the ROIs in our MRI in-vestigations, with the white matter architecture of an average brain from the Human Connectome Project (HCP) template. In our connectivity analysis, we found that three potential areas were possibly involved in the clinical course of the gelastic and ecstatic seizures: the left superior frontal gyrus, the brainstem, and the left thalamus. During the pres-ence of the seizures, the white matter pathways between hypothalamus-thala-mus and brainstem were particularly evident. The absence of the seizures in the postoperative period seemed to be connected with the predominant in-volvement of the fronto-thalamic-brainstem network. Therefore, we carefully suggest that a balance of the networks between the thalamus, the brain stem, and the frontal lobe seemed crucial with regard to experiencing gelastic and ecstatic seizures. Our results suggest a possible connection between gelastic and ecstatic seizures, where the frontal lobe is involved in the inhibitory ef-fect/negative reinforcement, and the activation of the brain stem in combina-tion with the gelastic component is a possible positive trigger/reinforcement of an ecstatic seizure [180]. These findings support the hypothesis that these phenomenon seem to be a combination of several functional networks (e.g. sensorimotor, auditory, visual, attentional, salience, or default mode net-works) [114-118], with a complex functional connectivity behind each symp-tom [121-127]. This new information about the possible organization supports the notion that it is a dynamic interplay rather than just a single trigger zone that produces the complex pattern of symptoms seen in patients with ecstatic epileptic seizures. The opportunity to study this rare form of epilepsy may add to a better understanding of the human brain connectivity and its relation to specific forms of epilepsy, possibly impacting tumor treatment planning. The Human Connectome Project (HPC) is a project to construct a map of the complete structural and functional neural connections in vivo within and across individuals (Figure 6). In this map, the connectivity of 500 healthy sub-jects is reconstructed [109, 181]. In our project, we used reconstructed data from 488 subjects, which were averaged to create a representative HCP-488 atlas. Hereafter, a deterministic fiber tracking algorithm was used [181]. ROIs from our own MRI investigations were placed within this template and used for the segmentation of the HPC’s white matter connectivity, which may have been influenced by the tumor area of our patient at three different times inves-tigated.

60

Concluding remarks

The work in the four studies forming this thesis touches on several concepts related directly or indirectly to the clinical management of gliomas (Figure 13). This thesis provides further insight into the use of intelligent integration of morphological, molecular, radiological, and clinical information for preci-sion medicine. The research presented here is built upon a wide spectrum of interest, tackling a variety of different methodological challenges in order to provide answers to the questions raised. First, we provide additional information on the importance of molecular gli-oma classification. We demonstrate that Prox1, a transcription factor involved in developmental processes, may have novel pathway-specific benefits as a marker. We provided proof for the use of Prox1 to predict the course of disease in patients with astrocytomas and concluded that the identification of distinc-tive molecular pathways can introduce new concepts for therapeutic manage-ment, based on the clinical diversity of these tumors. Second, we defined the diffuse tumor characteristics of low grade gliomas by studying the presence of tumor cells outside the radiological border in en bloc tumor resections. We support the notion that T2-FLAIR sequences are not sensitive enough to accurately detect the infiltration of tumor cells because they consistently show that tumor cells appear normal outside the radiological tumor border on preoperative images. Interestingly, we found that tumor cells followed the white matter tracts to spread to the adjacent gyrus, instead of infiltrating the pia mater. This information shows the importance of a combi-nation of radiological and histological tumor characteristics, leading to a better understanding of the behavior of tumor cells beyond the radiological border. This knowledge could improve tumor treatment planning in favor of radical and supratotal tumor resection. Advanced imaging with amino acid uptake can be favorably used to differen-tiate tumor progression from non-specific treatment related changes, due to its high tumor-to-background contrast. Here, we provide a deeper understanding of the imaging parameters in relation to the corresponding histopathological tumor characteristics. By evaluating the entire tumor volume of en bloc re-sected oligodendrogliomas using a PET-to-MRI-to-Histology co-registration, we were able to demonstrate that intra-tumoral 11C-methionine (MET) uptake

61

reflects tumor cell density throughout the whole tumor. Additional infor-mation on the histological and radiological heterogeneity of gliomas and a better understanding of noninvasive tumor distinction could improve tumor treatment planning and possibly spare healthy tissue.

By using a new approach in a complex and fascinating epileptic syndrome, we provide evidence for a possible anatomical network in the human brain that is involved in epilepsy. We investigated seizure semiology and MRI connectiv-ity changes during three different time periods, supporting the hypothesis that the complex pattern of epileptic symptoms seen in patients is a dynamic inter-play, rather than just a single trigger zone. Information provided by this unique study may contribute to a better understanding of human brain connectivity and its relation to specific forms of epilepsy, impacting precision medicine.

Figure 13. A schematic visualization of the four papers presented in this thesis, and their respective overlap.

62

Future Perspectives

This thesis provides further insights regarding the use of intelligent integration of morphological, molecular, radiological, and clinical information. We raised questions and provided evidence by presenting new methodologies. Results presented in this thesis may help to improve precision medicine. The methods presented in this thesis can be improved and used for similar studies. In paper I we showed the role of Prox1 as a prognostic factor for patients with specific genetic mutations. The different genetic characteristics between gli-oma types originated in diagnostic and predictive markers, which may give further insight into tumor-specific treatment planning. This study is the first to suggest Prox1 as a pathway-specific prognostic marker in high grade glio-mas (i.e. for astrocytomas but not for oligodendrogliomas). However, the mechanism behind this is still unknown and should be studied further. As mentioned previously, preliminary data from TCGA provided additional evi-dence for a pathway-dependent effect of PROX1 related to 1p19q codeletion in low-grade gliomas. This pathway-specific effect is of interest for future studies. This could lead to the discovery of novel cellular and molecular mech-anisms involved in adult neurogenesis. In paper II and paper III, we demonstrated a method for direct comparison between radiological and histological tumor characteristics, and its im-portance for tumor assessment and treatment. Especially tumor growth and its spread locally by moving into nearby normal tissue is a topic of intense inter-est. Previous studies have showed differences in cell migration regarding gli-oma type [182], based on the inhibitory effects of proteins. The large en-bloc tumor resections described in this study can be used to study protein expres-sions involved in the cell migration and tumor spread in the brain to reveal intra- and inter-tumoral differences. We therefore further developed our method of co-registration in order to provide a deeper understanding of the imaging parameters in relation to the corresponding histopathological tumor characteristics. For this, we used 11C-methionine (MET), which is the longest-established amino acid tracer. However, the tracer 18F- fluoroethyl-L-tyrosine (FET) has been proven to have logistic advances over MET due to its longer half-life and is now widely used for the clinical management of gliomas [72]. Therefore, it is of interest to look at the PET-to-MRI-to-Histology co-regis-tration using FET as the amino acid tracer of choice. In our study, we stated

63

that vessel density and the density of proliferating cells are less stringently correlated with MET uptake in infiltrating tumor edges. However, we discuss whether this is due to the susceptibility to artifacts caused by larger blood ves-sels surrounding the tumor. Blood vessels in low-grade glioma are morpho-logically similar to normal brain vasculature [164]; however, higher-grade gli-omas are characterized by abnormal vessels. It is hypothesized that the linear growth in low-grade gliomas is due to this absence of neovascularization [183], whereas the exponential growth and shorter survival of high-grade gli-oma are caused by the presence of hypervascularization [184]. This difference in neovascularization between low- and high-grade glioma is called the “an-giogenic switch” [185], which has recently been shown to be associated with IDH mutation status [186]. It has been suggested that targeting angiogenesis-promoting factors could be used therapeutically. [187, 188]. Therefore, a deeper understanding of the molecular mechanisms underlying vascular ab-normalization in gliomas and their relation to advanced imaging parameters would be of interest for future studies. This knowledge could contribute to the development of efficient treatment regimens for normalizing vascular function in the brain [189]. In paper IV we describe the importance of unique data to study seizure semi-ology changes in relation to MRI connectivity. Our results point toward a dy-namic interplay of the human brain rather than just a single trigger zone in epilepsy. The origin of specific types of epilepsy is the subject of a current debate on which we tried to shine new light. However, our study is based on only one case because of it’s the uniqueness of the condition. Additionally, our study showed technical limitations due to the use of a vagus nerve stimu-lator in the patient. In future studies, it would be beneficial to find a way to perform MR tractography to study the anatomical connectivity involved in epilepsy. Though challenging, more cases with rare forms of epilepsy such as ecstatic epilepsy need to be recognized, studied, and described in the literature to deepen our knowledge.

64

Svensk Sammanfattning

Denna Doktorsavhandling baseras på fyra studier som berör områden och kon-cept som direkt eller indirekt är relaterade till den kliniska handläggningen av gliom (Figur 13). Syftet med avhandlingen är att bidra med ytterligare insikter inom användningsområdet “precision medicin” med smart integration av mor-fologisk, molekylär, radiologisk och klinisk informations karaktär. Forsk-ningen som presenteras är baserat på ett stort intresseområde med varierande metodologiska utmaningar som genom bearbetning i sin tur ger svar på den ställda frågan. Det första som vi gjorde var att presentera information som är viktigt för klas-sificeringen av olika molekylära gliom. Vi har demonstrerat att Prox1, en transkriptions faktor som är involverad i utvecklingsstadiets processer, kan ha en ny signalvägs-specifik nytta som biomarkör. Genom att använda sig av Prox 1 har vi bevisat att vi kan förutspå sjukdom hos patienter med astrocy-tom. Detta öppnar upp möjligheten att det går att identifiera distinkta mole-kylära signalvägar som kan användas för att introducera ett nytt koncept av terapeutiska behandlingar som kan baseras på tumörers diversitet inom klini-ken. Det andra vi gjorde var att definiera den diffusa tumörkarakteristiken av låg grads gliom genom att studera närvaron av tumörceller utanför den radiolo-giska gränsen i en ”en bloc ” tumörresektion. Vi stödjer åsikten att FLAIR sekvenser inte är känsliga nog för att korrekt upptäcka infiltration av tumör-celler. Detta visas genom att tumörceller i upprepade fall påvisats utanför den radiologiska gränsen på preoperativa foton. Intressant nog så har vi hittat tumörceller som följer den vita hjärnsubstansen gångar som sprider sig till närliggande gyrus istället för att infiltrera pia matern. Detta fynd visar vikten av att kombinera radiologiska och histologiska tumör karakterisering/diagnostik för att bättre förstå hur tumörceller beter sig utanför den radiologiska gränsen. I framtiden skulle detta eventuellt kunna leda till en förbättrad tumörbehandlings planering där en radikal och supra total tumörresektion är att föredra. För att differentiera tumörprogression i en icke-specifik behandling föredrar man användningen av metoden “advanced imaging” där man mäter upptaget

65

av aminosyror. Metoden är föredragen på grund av att den ger en hög tumör-kontrast till bakgrunds återgivning. Här har vi bidragit med en djupare förstå-else för återspeglings parametrar i relation till den korresponderande histopa-tologiska tumörkarakteristiken. Med hjälp av PET, MRI, och histologisk co-registrering har vi undersökt tumörvolymen i en ”en bloc” resekterad Oligodendrogliom och visat att intra-tumörala C-methionine (MET) upptag reflekterar tumörcellers densitet genom hela tumören. Ytterligare information om histologiska och radiologiska heterogeniteten av gliom, samt en bättre för-ståelse för icke invasiv tumör distinktion skulle kunna förbättra tumörbehand-lings planering och eventuellt spara frisk vävnad vid en resektion. Slutligen har vi genom att använda en ny infallsvinkel i ett komplext och fa-scinerande epileptiskt syndrom kunnat bidra med bevis för ett möjligen nytt anatomiskt nätverk i hjärnan kopplat till epilepsi. Vi undersökte anfallssemi-ologi och MRI- konektivitets förändringar vid tre olika tidpunkter för att stödja hypotesen som föreslår att de komplexa mönstren av epileptiska sym-tom som setts i patienter handlar mer om ett dynamiskt mellanspel än en en-sam ”trigger zone”. Information från denna unika studie skulle eventuellt kunna bidra med en bättre förståelse för hjärnans konektivitet och dess relation till specifika former av epilepsi som skulle kunna ha en inverkan på precisions medicin.

66

Nederlandse Samenvatting

In deze thesis beschrijf ik vier studies die direct of indirect aansluiten op de klinische behandeling van gliomas (Figuur 13). Vandaag de dag krijgen kankerpatiënten niet altijd een effectieve behandeling. Een behandeling die voor de ene patiënt wel werkt, werkt voor een andere patiënt misschien veel minder en vice versa. In sommige onderzoekscentra wordt daarom aanvullend op een diagnose met de microscoop, tijdens een moleculair onderzoek gekeken wat er precies misgaat bij de patiënt, op het niveau van het DNA en RNA. Aan de hand van deze analyse, kan een arts in sommige gevallen bepalen welke behandeling wel of niet effectief is. Dit wordt “precisie-geneeskunde” genoemd. Tijdens het onderzoek dat hier wordt gepresenteerd worden vanuit een breed draagvlak uiteenlopende wetenschappelijke vraagstukken aangehaald. De onderzoeken hebben als doel bij te dragen aan de verbetering van de tumor behandeling, en zijn een combinatie van verschillende aspecten. Zo wordt er in de thesis dieper ingegaan op de morfologische, moleculaire, radiologische en klinische achtergrond van gliomas, om een klein steentje bij te dragen aan de verbetering van de precisie-geneeskunde. Ten eerste worden de belangen van de moleculaire classificatie van gliomas aangehaald. In dit onderzoek laten wij zien dat Prox1, een transcriptie factor betrokken bij ontwikkelingsprocessen van het centraal zenuwstelsel, mogelijk nieuwe pathway-specifieke voordelen heeft als een marker. Wij verschaffen bewijs voor het gebruik van Prox1, om het verloop van de ziekte te voorspellen bij patiënten met astrocytomas. De identificatie van moleculaire pathways kan nieuwe concepten verschaffen voor therapeutische behandeling, gebaseerd op de klinische diversiteit van deze tumoren. Ten tweede worden, door het bestuderen van tumorcellen buiten de radiologische grens de tumor karakteristieken van laag-graad gliomas verder in kaart gebracht. Dit hebben wij gedaan door het gebruik van en-bloc verwijderde resecties, waarbij de gehele tumor wordt verwijderd, te vergelijken met radiologische scans van de tumor vóór de operatie. Door nauwkeurig tumorcellen buiten de radiologische grens aan te tonen die niet te zien waren op de scans, bevestigen wij eerdere bevindingen dat T2-FLAIR

67

niet gevoelig genoeg is om nauwkeurig infiltrerende tumorcellen te meten. Verder vonden wij dat tumorcellen zich spreiden langs de witte banen in het brein, in plaats van de pia mater te infiltreren. Deze informatie toont de belangen en mogelijkheden van het combineren van radiologische- met histologische informatie. Uiteindelijk draagt ons onderzoek bij tot het verschaffen van een beter inzicht in het gedrag van tumorcellen buiten de radiologische grens. Dit leidt tot het verbeteren van tumor behandeling, waarbij ons onderzoek een voorkeur laat zien voor radicale en supra-totale verwijdering van de tumor. Advanced imaging waarbij de aminozuur opname in het brein wordt gemeten is een goede manier om tumor progressie te onderscheiden van niet specifieke veranderingen in het brein. Dit komt vanwege het hoge contrast tussen de tumor en achtergrondinformatie op de scans. In dit onderzoek gaan wij dieper in op de radiologische meetwaarden in relatie tot de histologie van de tumor. Wij hebben de gehele tumor volume van en bloc verwijderde oligodendrogliomas geëvalueerd door het gebruik van PET-op-MRI-op-Histologie co-registraties. Op deze manier laten wij zien dat in de gehele tumor 11C-methionine (MET) opname correspondeert met de dichtheid van tumorcellen. Het verschaffen van informatie omtrent histologische- en radiologische heterogeniteit van gliomas kan leiden tot het beter onderscheiden van tumoren aan de hand van radiologische scans en het verbeteren van de tumor behandeling door bijvoorbeeld (zo min mogelijk) gezond hersenweefsel te verwijderen. Uiteindelijk hebben wij gebruik gemaakt van een complex en fascinerend epileptisch syndroom om een eventueel anatomisch netwerk in het menselijk brein te onderzoeken die betrokken is bij epilepsie. Wij beschrijven een nieuwe methode waarbij we de semiologie van de epileptische aanvallen hebben vergeleken met veranderingen in MRI-connectiviteit, wanneer deze specifieke aanval af- en aanwezig was. Wij bevestigen aan de hand van onze resultaten dat het complex patroon in patiënten met epilepsie eerder een dynamische wisselwerking is tussen verschillende gebieden in het brein, in plaats van een enkele triggerzone. De informatie uit dit onderzoek kan mogelijk leiden tot een beter begrip van de connectiviteit van het menselijk brein, in relatie tot specifieke vormen van epilepsie.

68

Words of gratitude

There are many to whom I am grateful for being part of my journey in any way possible. Ever since I was a young boy, I have been the odd one out. I have never been the best listener, most of the time a little absent, slightly stub-born, and a bit quirky. But it was those personality traits that led me to the path of science. My intense fascination with life, science, and first and fore-most the human brain was fed by events throughout life. The birth of a younger sister, where my older brother and I could not stop yelling proudly to the world, we have a younger sister (How can you feel such a strong bond with someone you have just met?). The passing of a beloved great-aunt, who, more than anyone and anything else, is the inspiration that led me to question life and become the scientist I am today (What happens to the brain after you die, when you are nothing but a memory in other people’s minds?). The pain fol-lowing the ending of a romantic relationship, but the intense pleasure of falling in love again (I actually did study this effect on the human brain). I could go on endlessly, and wish a lifetime was long enough to study the whole world, but for now I leave it at this. I wish I could thank everyone personally that has been of support in some way or another during my doctoral thesis. First of all, I want to thank my main supervisor Anja Smits, for providing me with the possibility to follow my passion in science. Particularly, I have always been very impressed by the trust you showed me, by letting me develop my independence. It caused a healthy pressure to prove myself to you, and I hope that I have succeeded. I wish that one day I can match your impressive scientific writing skills. I appreciate the support you gave me to follow my dreams for medical school after the doctoral thesis. Also, thanks to my co-supervisors. Elna-Marie Larsson, without you I would never have undertaken the adventure to come to Uppsala. I still re-member the moment I drove all the way to Sweden for a 5-minute talk. How you tried to warn me about the challenges ahead, but nothing could change my enthusiasm for a research project abroad. The Erasmus time that has given me so much. I am grateful to Anne-Marie Landtblom and her family who adopted me in Sweden. Talking for hours about history, science and shared interests, while experiencing the Swedish culture during a “smörgåsbord” and “kräftskiva”. I am thankful for everything that you, my Swedish family, has done for me.

69

A special mention to all my collaborators in the studies from this thesis. In particular Maria Zetterling, without your surgical skills we would have never been able to perform our studies. Shala Berntsson, for including the patients. Irina Alafuzoff, for your neuropathological expertise. Per-Henrik Edqvist and the Human Protein Atlas group at Rudbeck Laboratory, for all the help and support. Annika Malmström and Johan Rosell from Linköping Univer-sity, providing your cohort of patients. As well as Tamador Elsir and Daniel Hägerstrand, from Stockholm University. Torsten Danfors, for your help at the PET center. To Anna Dimberg, for your help and letting me experience the lab work within the IGP group, with Lei Zhang, Maria Georganaki and Roberta Lugano. To Robin Strand, from the Image Analysis Center, with your help on the methodological parts. To Francesco Latini, for your contri-bution in the final paper. Ali Alhuseinalkhudhur and Mohammed Al-Jaff, for your time and expertise. Your collaboration has been a pleasant social ad-dition to my studies over the last year. To Sari Thunberg, for always being there when I needed either administrative help or a small chat. To Alexander Franzén, for his Swedish translation. To Nils Weber for the portrait in the back-cover. To Concepción Bueno Galera and Concepción Galera Moya, for the cover and illustrations in this thesis. A special thanks to all participants in the studies presented in this thesis, and the financial support this thesis has received from; Bissen Brain Walk Stiftelse; Erik, Karin och Gösta Selanders Stiftelse; Lions Stiftelse för Can-cerforskning vid Akademiska Sjukhuset; Hanna Eklund Stiftelse; and Uppsala County Council (ALF-avtal). Thanks to old and new friendships. The unexpected turn of events ending up with a sojourn in Sweden for nearly five years has showed me the importance of friendships. Harold, who bought a Toyota Vitara five years ago to explore Scandinavia with me so I could meet a professor in Sweden to talk about a research project abroad. Without you, all this might have never happened. And for that, I am always thankful. Johan, who shares the same addiction for sim-ulating football matches and who, while distances became longer, grew closer as a friend. And the rest of the “Zwitsal” group; Chiljon, Milo, Eric, and Arend. Joop, our special friendship goes back as far as I can remember. Dis-tance did not change that. And guess what, I´m coming back boys. Over the past years, I have made a lot of new friendships. If I ever want to travel the world, I am confident that I have a place in all of your homes, and likewise my door will always be open to you. I wish I could mention all of you, but if you read these words and you feel that you were part of my adventure in Swe-den, then know that I would have loved to put your name right here “______”. In particular Collin, who during my first year made sure the Dutch in me never really left. It has been tough to replace your absence after you left. Then, Fa-bian, Verena, Morten, Laura, Mark, Kristina, Nils, Nadja, and Alex for

70

your friendship and moral support in my last two years. Of course, also you Emma. I have given you a hard time about Swedish hospitality. And you proved me wrong.

Furthermore, I want to thank the entire International Committee that I have worked closely with over the past years. It has been both a pleasure and an honor to have been your president for a whole year. Working with so many enthusiastic people to create a new home for the international students has been a welcome distraction. The Committee will always have a place in my heart, hoping that my efforts made a small positive difference for the new stu-dents to come in future semesters. And it was then, while I was busy welcom-ing the new students to Uppsala, that I met you, Conchi Bueno Galera, my Spanish beauty. I am happy that you decided to come on Erasmus to Uppsala. Thank you for always being there whenever I need you. How you and your family believed in me, no matter what, more than I sometimes did myself.

Last, but definitely not least, family Roodakker. Leo and Anja, the reason of my existence and biggest motivational support in my life. You have always been there for me, and always will, through good and bad times. My older brother Marvin and younger sister Eline, who both walked their own path in the medical world. For me, this adventure is almost over, knowing that I made the best of most situations. Life is all about taking chances, small steps, to get closer to your goal, your passion, your dream. One problem at a time. I’m ready for the next step, a new adventure, in my home country where I will attend medical school. I will close this thesis with an anecdote.

The very first Delta Force mission was to free hostages out of Tehran, in 1979, Iran. One of the members of this elite group was fast asleep when the plane lands in the desert somewhere to refuel. While it's on the ground, a Navy hel-icopter crashes into it due to the heavy winds. The person wakes up, and sees the plane is on fire. Thinking they’re still 20,000 feet up, he can't find a para-chute. So, he just flings himself through the smoke, right out the door. Later, the rest of the members ask him: "If you thought we were 20,000 feet in the air, why would you fling yourself out the door without a parachute?". And he answers "All I knew was the plane was on fire, and I figured one prob-lem at a time". From Inside Delta Force, by Eric L. Haney

71

References

1. Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, Burger PC, Jouvet A, et al. The 2007 WHO classification of tumours of the central nervous system. Acta Neuropathol. 2007;114:97-109. doi:10.1007/s00401-007-0243-4.

2. Louis DN, Perry A, Reifenberger G, von Deimling A, Figarella-Branger D, Cavenee WK, et al. The 2016 World Health Organization Classification of Tu-mors of the Central Nervous System: a summary. Acta Neuropathol. 2016;131:803-20. doi:10.1007/s00401-016-1545-1.

3. Ohgaki H, Kleihues P. Population-based studies on incidence, survival rates, and genetic alterations in astrocytic and oligodendroglial gliomas. J Neuropathol Exp Neurol. 2005;64:479-89.

4. Ostrom QT, Gittleman H, Liao P, Vecchione-Koval T, Wolinsky Y, Kruchko C, et al. CBTRUS Statistical Report: Primary brain and other central nervous sys-tem tumors diagnosed in the United States in 2010-2014. Neuro Oncol. 2017;19:v1-v88. doi:10.1093/neuonc/nox158.

5. Bailey OT. Genesis of the Percival Bailey-Cushing classification of gliomas. Pe-diatr Neurosci. 1985;12:261-5.

6. Crocetti E, Trama A, Stiller C, Caldarella A, Soffietti R, Jaal J, et al. Epidemi-ology of glial and non-glial brain tumours in Europe. Eur J Cancer. 2012;48:1532-42. doi:10.1016/j.ejca.2011.12.013.

7. Okamoto Y, Di Patre PL, Burkhard C, Horstmann S, Jourde B, Fahey M, et al. Population-based study on incidence, survival rates, and genetic alterations of low-grade diffuse astrocytomas and oligodendrogliomas. Acta Neuropathol. 2004;108:49-56. doi:10.1007/s00401-004-0861-z.

8. Bauchet L. Epidemiology of Diffuse Low-Grade Gliomas. In: Duffau H, editor. Diffuse Low-Grade Gliomas in Adults. London: Springer; 2013. p. 9-30.

9. Cairncross JG, Ueki K, Zlatescu MC, Lisle DK, Finkelstein DM, Hammond RR, et al. Specific genetic predictors of chemotherapeutic response and survival in patients with anaplastic oligodendrogliomas. J Natl Cancer Inst. 1998;90:1473-9.

10. Wesseling P, van den Bent M, Perry A. Oligodendroglioma: pathology, molec-ular mechanisms and markers. Acta Neuropathol. 2015;129:809-27. doi:10.1007/s00401-015-1424-1.

11. Bampoe J, Bernstein M. The role of surgery in low grade gliomas. J Neurooncol. 1999;42:259-69.

12. Davis FG, McCarthy BJ. Epidemiology of brain tumors. Curr Opin Neurol. 2000;13:635-40.

13. Schiff D, Brown PD, Giannini C. Outcome in adult low-grade glioma: the impact of prognostic factors and treatment. Neurology. 2007;69:1366-73. doi:10.1212/01.wnl.0000277271.47601.a1.

14. Reifenberger G, Collins VP. Pathology and molecular genetics of astrocytic gli-omas. J Mol Med (Berl). 2004;82:656-70. doi:10.1007/s00109-004-0564-x.

72

15. Miller CR, Dunham CP, Scheithauer BW, Perry A. Significance of necrosis in grading of oligodendroglial neoplasms: a clinicopathologic and genetic study of newly diagnosed high-grade gliomas. J Clin Oncol. 2006;24:5419-26. doi:10.1200/JCO.2006.08.1497.

16. Miller CR, Perry A. Glioblastoma. Arch Pathol Lab Med. 2007;131:397-406. doi:10.1043/1543-2165(2007)131[397:G]2.0.CO;2.

17. Sanai N, Chang S, Berger MS. Low-grade gliomas in adults. J Neurosurg. 2011;115:948-65. doi:10.3171/2011.7.JNS101238.

18. Berger MS, Rostomily RC. Low grade gliomas: functional mapping resection strategies, extent of resection, and outcome. J Neurooncol. 1997;34:85-101.

19. Duffau H, Capelle L. Preferential brain locations of low-grade gliomas. Cancer. 2004;100:2622-6. doi:10.1002/cncr.20297.

20. Pallud J, Blonski M, Mandonnet E, Audureau E, Fontaine D, Sanai N, et al. Ve-locity of tumor spontaneous expansion predicts long-term outcomes for diffuse low-grade gliomas. Neuro Oncol. 2013;15:595-606. doi:10.1093/neuonc/nos331.

21. Goze C, Blonski M, Le Maistre G, Bauchet L, Dezamis E, Page P, et al. Imaging growth and isocitrate dehydrogenase 1 mutation are independent predictors for diffuse low-grade gliomas. Neuro Oncol. 2014;16:1100-9. doi:10.1093/neu-onc/nou085.

22. Hayhurst C. Contemporary management of low--grade glioma: a paradigm shift in neuro-oncology. Pract Neurol. 2017;17:183-90. doi:10.1136/practneurol-2017-001604.

23. Duffau H, Lopes M, Arthuis F, Bitar A, Sichez JP, Van Effenterre R, et al. Con-tribution of intraoperative electrical stimulations in surgery of low grade glio-mas: a comparative study between two series without (1985-96) and with (1996-2003) functional mapping in the same institution. J Neurol Neurosurg Psychia-try. 2005;76:845-51. doi:10.1136/jnnp.2004.048520.

24. Wrensch M, Minn Y, Chew T, Bondy M, Berger MS. Epidemiology of primary brain tumors: current concepts and review of the literature. Neuro Oncol. 2002;4:278-99.

25. Appin CL, Brat DJ. Molecular pathways in gliomagenesis and their relevance to neuropathologic diagnosis. Adv Anat Pathol. 2015;22:50-8. doi:10.1097/PAP.0000000000000048.

26. Marko N, Weil, RJ. The Molecular Biology of Diffuse Low-Grade Gliomas. In: Duffau H, editor. Diffuse Low-Grade Gliomas in Adults. London: Springer; 2013. p. 99-116.

27. Ohgaki H, Kleihues P. The definition of primary and secondary glioblastoma. Clin Cancer Res. 2013;19:764-72. doi:10.1158/1078-0432.CCR-12-3002.

28. Baumert BG, Hegi ME, van den Bent MJ, von Deimling A, Gorlia T, Hoang-Xuan K, et al. Temozolomide chemotherapy versus radiotherapy in high-risk low-grade glioma (EORTC 22033-26033): a randomised, open-label, phase 3 intergroup study. Lancet Oncol. 2016. doi:10.1016/S1470-2045(16)30313-8.

29. Ferris SP, Hofmann JW, Solomon DA, Perry A. Characterization of gliomas: from morphology to molecules. Virchows Arch. 2017;471:257-69. doi:10.1007/s00428-017-2181-4.

30. Paavilainen L, Wernerus H, Nilsson P, Uhlen M, Hober S, Wester K, et al. Eval-uation of monospecific antibodies: a comparison study with commercial analogs using immunohistochemistry on tissue microarrays. Appl Immunohistochem Mol Morphol. 2008;16:493-502. doi:10.1097/PAI.0b013e31817c645e.

73

31. Reifenberger G, Szymas J, Wechsler W. Differential expression of glial- and neuronal-associated antigens in human tumors of the central and peripheral nerv-ous system. Acta Neuropathol. 1987;74:105-23.

32. Blumcke I, Becker AJ, Normann S, Hans V, Riederer BM, Krajewski S, et al. Distinct expression pattern of microtubule-associated protein-2 in human oli-godendrogliomas and glial precursor cells. J Neuropathol Exp Neurol. 2001;60:984-93.

33. Faria MH, Goncalves BP, do Patrocinio RM, de Moraes-Filho MO, Rabenhorst SH. Expression of Ki-67, topoisomerase IIalpha and c-MYC in astrocytic tu-mors: correlation with the histopathological grade and proliferative status. Neu-ropathology. 2006;26:519-27.

34. Watanabe T, Nobusawa S, Kleihues P, Ohgaki H. IDH1 mutations are early events in the development of astrocytomas and oligodendrogliomas. Am J Pathol. 2009;174:1149-53. doi:10.2353/ajpath.2009.080958.

35. Balss J, Meyer J, Mueller W, Korshunov A, Hartmann C, von Deimling A. Anal-ysis of the IDH1 codon 132 mutation in brain tumors. Acta Neuropathol. 2008;116:597-602. doi:10.1007/s00401-008-0455-2.

36. Chen R, Smith-Cohn M, Cohen AL, Colman H. Glioma Subclassifications and Their Clinical Significance. Neurotherapeutics. 2017;14:284-97. doi:10.1007/s13311-017-0519-x.

37. Wakimoto H, Tanaka S, Curry WT, Loebel F, Zhao D, Tateishi K, et al. Tar-getable signaling pathway mutations are associated with malignant phenotype in IDH-mutant gliomas. Clin Cancer Res. 2014;20:2898-909. doi:10.1158/1078-0432.CCR-13-3052.

38. Siegal T. Clinical impact of molecular biomarkers in gliomas. J Clin Neurosci. 2015;22:437-44. doi:10.1016/j.jocn.2014.10.004.

39. Liu XY, Gerges N, Korshunov A, Sabha N, Khuong-Quang DA, Fontebasso AM, et al. Frequent ATRX mutations and loss of expression in adult diffuse astrocytic tumors carrying IDH1/IDH2 and TP53 mutations. Acta Neuropathol. 2012;124:615-25. doi:10.1007/s00401-012-1031-3.

40. Simon M, Hosen I, Gousias K, Rachakonda S, Heidenreich B, Gessi M, et al. TERT promoter mutations: a novel independent prognostic factor in primary gli-oblastomas. Neuro Oncol. 2015;17:45-52. doi:10.1093/neuonc/nou158.

41. Heidenreich B, Rachakonda PS, Hemminki K, Kumar R. TERT promoter muta-tions in cancer development. Curr Opin Genet Dev. 2014;24:30-7. doi:10.1016/j.gde.2013.11.005.

42. Masui K, Komori T, Kato Y, Masutomi K, Ichimura K, Ogasawara S, et al. El-evated TERT Expression in TERT-Wildtype Adult Diffuse Gliomas: Histologi-cal Evaluation with a Novel TERT-Specific Antibody. Biomed Res Int. 2018;2018:7945845. doi:10.1155/2018/7945845.

43. Picca A, Berzero G, Sanson M. Current therapeutic approaches to diffuse grade II and III gliomas. Ther Adv Neurol Disord. 2018;11:1756285617752039. doi:10.1177/1756285617752039.

44. Reifenberger G, Wirsching HG, Knobbe-Thomsen CB, Weller M. Advances in the molecular genetics of gliomas - implications for classification and therapy. Nat Rev Clin Oncol. 2017;14:434-52. doi:10.1038/nrclinonc.2016.204.

45. Reuss DE, Kratz A, Sahm F, Capper D, Schrimpf D, Koelsche C, et al. Adult IDH wild type astrocytomas biologically and clinically resolve into other tumor entities. Acta Neuropathol. 2015;130:407-17. doi:10.1007/s00401-015-1454-8.

74

46. Jenkins RB, Blair H, Ballman KV, Giannini C, Arusell RM, Law M, et al. A t(1;19)(q10;p10) mediates the combined deletions of 1p and 19q and predicts a better prognosis of patients with oligodendroglioma. Cancer Res. 2006;66:9852-61. doi:10.1158/0008-5472.CAN-06-1796.

47. Suzuki H, Aoki K, Chiba K, Sato Y, Shiozawa Y, Shiraishi Y, et al. Mutational landscape and clonal architecture in grade II and III gliomas. Nat Genet. 2015;47:458-68. doi:10.1038/ng.3273.

48. Thompson CB. Metabolic enzymes as oncogenes or tumor suppressors. N Engl J Med. 2009;360:813-5. doi:10.1056/NEJMe0810213.

49. van den Bent MJ, Gravendeel LA, Gorlia T, Kros JM, Lapre L, Wesseling P, et al. A hypermethylated phenotype is a better predictor of survival than MGMT methylation in anaplastic oligodendroglial brain tumors: a report from EORTC study 26951. Clin Cancer Res. 2011;17:7148-55. doi:10.1158/1078-0432.CCR-11-1274.

50. Ohgaki H, Kleihues P. Genetic pathways to primary and secondary glioblas-toma. Am J Pathol. 2007;170:1445-53. doi:10.2353/ajpath.2007.070011.

51. von Deimling A, Louis DN, von Ammon K, Petersen I, Wiestler OD, Seizinger BR. Evidence for a tumor suppressor gene on chromosome 19q associated with human astrocytomas, oligodendrogliomas, and mixed gliomas. Cancer Res. 1992;52:4277-9.

52. Hegi ME, Diserens AC, Gorlia T, Hamou MF, de Tribolet N, Weller M, et al. MGMT gene silencing and benefit from temozolomide in glioblastoma. N Engl J Med. 2005;352:997-1003. doi:10.1056/NEJMoa043331.

53. Stergiopoulos A, Elkouris M, Politis PK. Prospero-related homeobox 1 (Prox1) at the crossroads of diverse pathways during adult neural fate specification. Front Cell Neurosci. 2014;8:454. doi:10.3389/fncel.2014.00454.

54. Lavado A, Oliver G. Prox1 expression patterns in the developing and adult mu-rine brain. Dev Dyn. 2007;236:518-24. doi:10.1002/dvdy.21024.

55. Elsir T, Smits A, Lindstrom MS, Nister M. Transcription factor PROX1: its role in development and cancer. Cancer Metastasis Rev. 2012;31:793-805. doi:10.1007/s10555-012-9390-8.

56. Doe CQ, Chu-LaGraff Q, Wright DM, Scott MP. The prospero gene specifies cell fates in the Drosophila central nervous system. Cell. 1991;65:451-64.

57. Li L, Vaessin H. Pan-neural Prospero terminates cell proliferation during Dro-sophila neurogenesis. Genes Dev. 2000;14:147-51.

58. Choksi SP, Southall TD, Bossing T, Edoff K, de Wit E, Fischer BE, et al. Pros-pero acts as a binary switch between self-renewal and differentiation in Drosoph-ila neural stem cells. Dev Cell. 2006;11:775-89. doi:10.1016/j.devcel.2006.09.015.

59. Wigle JT, Oliver G. Prox1 function is required for the development of the murine lymphatic system. Cell. 1999;98:769-78.

60. Steiner B, Zurborg S, Horster H, Fabel K, Kempermann G. Differential 24 h responsiveness of Prox1-expressing precursor cells in adult hippocampal neuro-genesis to physical activity, environmental enrichment, and kainic acid-induced seizures. Neuroscience. 2008;154:521-9. doi:10.1016/j.neurosci-ence.2008.04.023.

61. Galeeva A, Treuter E, Tomarev S, Pelto-Huikko M. A prospero-related homeo-box gene Prox-1 is expressed during postnatal brain development as well as in the adult rodent brain. Neuroscience. 2007;146:604-16. doi:10.1016/j.neurosci-ence.2007.02.002.

75

62. Foskolou IP, Stellas D, Rozani I, Lavigne MD, Politis PK. Prox1 suppresses the proliferation of neuroblastoma cells via a dual action in p27-Kip1 and Cdc25A. Oncogene. 2013;32:947-60. doi:10.1038/onc.2012.129.

63. Elkouris M, Balaskas N, Poulou M, Politis PK, Panayiotou E, Malas S, et al. Sox1 maintains the undifferentiated state of cortical neural progenitor cells via the suppression of Prox1-mediated cell cycle exit and neurogenesis. Stem Cells. 2011;29:89-98. doi:10.1002/stem.554.

64. Kaltezioti V, Kouroupi G, Oikonomaki M, Mantouvalou E, Stergiopoulos A, Charonis A, et al. Prox1 regulates the notch1-mediated inhibition of neurogene-sis. PLoS Biol. 2010;8:e1000565. doi:10.1371/journal.pbio.1000565.

65. Elsir T, Eriksson A, Orrego A, Lindstrom MS, Nister M. Expression of PROX1 Is a common feature of high-grade malignant astrocytic gliomas. J Neuropathol Exp Neurol. 2010;69:129-38. doi:10.1097/NEN.0b013e3181ca4767.

66. Elsir T, Qu M, Berntsson SG, Orrego A, Olofsson T, Lindstrom MS, et al. PROX1 is a predictor of survival for gliomas WHO grade II. Br J Cancer. 2011;104:1747-54. doi:10.1038/bjc.2011.162.

67. Law M, Yang S, Wang H, Babb JS, Johnson G, Cha S, et al. Glioma grading: sensitivity, specificity, and predictive values of perfusion MR imaging and pro-ton MR spectroscopic imaging compared with conventional MR imaging. AJNR Am J Neuroradiol. 2003;24:1989-98.

68. Fouke SJ, Benzinger T, Gibson D, Ryken TC, Kalkanis SN, Olson JJ. The role of imaging in the management of adults with diffuse low grade glioma: A sys-tematic review and evidence-based clinical practice guideline. J Neurooncol. 2015;125:457-79. doi:10.1007/s11060-015-1908-9.

69. Whitfield GA, Kennedy SR, Djoukhadar IK, Jackson A. Imaging and target vol-ume delineation in glioma. Clin Oncol (R Coll Radiol). 2014;26:364-76. doi:10.1016/j.clon.2014.04.026.

70. Jansen EP, Dewit LG, van Herk M, Bartelink H. Target volumes in radiotherapy for high-grade malignant glioma of the brain. Radiother Oncol. 2000;56:151-6.

71. Jenkinson MD, du Plessis DG, Smith TS, Joyce KA, Warnke PC, Walker C. Histological growth patterns and genotype in oligodendroglial tumours: correla-tion with MRI features. Brain. 2006;129:1884-91. doi:10.1093/brain/awl108.

72. Langen KJ, Galldiks N, Hattingen E, Shah NJ. Advances in neuro-oncology im-aging. Nat Rev Neurol. 2017;13:279-89. doi:10.1038/nrneurol.2017.44.

73. Tozer DJ, Jager HR, Danchaivijitr N, Benton CE, Tofts PS, Rees JH, et al. Ap-parent diffusion coefficient histograms may predict low-grade glioma subtype. NMR Biomed. 2007;20:49-57. doi:10.1002/nbm.1091.

74. Berntsson SG, Falk A, Savitcheva I, Godau A, Zetterling M, Hesselager G, et al. Perfusion and diffusion MRI combined with (1)(1)C-methionine PET in the pre-operative evaluation of suspected adult low-grade gliomas. J Neurooncol. 2013;114:241-9. doi:10.1007/s11060-013-1178-3.

75. Price SJ. The role of advanced MR imaging in understanding brain tumour pa-thology. Br J Neurosurg. 2007;21:562-75. doi:10.1080/02688690701700935.

76. Iwama T, Yamada H, Sakai N, Andoh T, Nakashima T, Hirata T, et al. Correla-tion between magnetic resonance imaging and histopathology of intracranial gli-oma. Neurol Res. 1991;13:48-54.

77. Kelly PJ, Daumas-Duport C, Kispert DB, Kall BA, Scheithauer BW, Illig JJ. Imaging-based stereotaxic serial biopsies in untreated intracranial glial neo-plasms. J Neurosurg. 1987;66:865-74. doi:10.3171/jns.1987.66.6.0865.

78. Lunsford LD, Martinez AJ, Latchaw RE. Magnetic resonance imaging does not define tumor boundaries. Acta Radiol Suppl. 1986;369:154-6.

76

79. Pallud J, Varlet P, Devaux B, Geha S, Badoual M, Deroulers C, et al. Diffuse low-grade oligodendrogliomas extend beyond MRI-defined abnormalities. Neu-rology. 2010;74:1724-31. doi:10.1212/WNL.0b013e3181e04264.

80. Herholz K, Coope D, Jackson A. Metabolic and molecular imaging in neuro-oncology. Lancet Neurol. 2007;6:711-24. doi:10.1016/S1474-4422(07)70192-8.

81. Aronen HJ, Gazit IE, Louis DN, Buchbinder BR, Pardo FS, Weisskoff RM, et al. Cerebral blood volume maps of gliomas: comparison with tumor grade and histologic findings. Radiology. 1994;191:41-51. doi:10.1148/radiol-ogy.191.1.8134596.

82. Patel P, Baradaran H, Delgado D, Askin G, Christos P, John Tsiouris A, et al. MR perfusion-weighted imaging in the evaluation of high-grade gliomas after treatment: a systematic review and meta-analysis. Neuro Oncol. 2017;19:118-27. doi:10.1093/neuonc/now148.

83. Albert NL, Weller M, Suchorska B, Galldiks N, Soffietti R, Kim MM, et al. Response Assessment in Neuro-Oncology working group and European Associ-ation for Neuro-Oncology recommendations for the clinical use of PET imaging in gliomas. Neuro Oncol. 2016;18:1199-208. doi:10.1093/neuonc/now058.

84. Okubo S, Zhen HN, Kawai N, Nishiyama Y, Haba R, Tamiya T. Correlation of L-methyl-11C-methionine (MET) uptake with L-type amino acid transporter 1 in human gliomas. J Neurooncol. 2010;99:217-25. doi:10.1007/s11060-010-0117-9.

85. Galldiks N, Stoffels G, Filss C, Rapp M, Blau T, Tscherpel C, et al. The use of dynamic O-(2-18F-fluoroethyl)-l-tyrosine PET in the diagnosis of patients with progressive and recurrent glioma. Neuro Oncol. 2015;17:1293-300. doi:10.1093/neuonc/nov088.

86. Langen K-J, Rapp M, Sabel M, Galldiks N. Positron-Emission-Tomography in Diffuse Low-Grade Gliomas. Diffuse Low-Grade Gliomas in Adults: Springer, Cham; 2017. p. 263-86.

87. Galldiks N, Langen KJ. Amino acid PET in neuro-oncology: applications in the clinic. Expert Rev Anticancer Ther. 2017;17:395-7. doi:10.1080/14737140.2017.1302799.

88. Mosskin M, Ericson K, Hindmarsh T, von Holst H, Collins VP, Bergstrom M, et al. Positron emission tomography compared with magnetic resonance imaging and computed tomography in supratentorial gliomas using multiple stereotactic biopsies as reference. Acta Radiol. 1989;30:225-32.

89. Pirotte B, Goldman S, Dewitte O, Massager N, Wikler D, Lefranc F, et al. Inte-grated positron emission tomography and magnetic resonance imaging-guided resection of brain tumors: a report of 103 consecutive procedures. J Neurosurg. 2006;104:238-53. doi:10.3171/jns.2006.104.2.238.

90. Lopez WO, Cordeiro JG, Albicker U, Doostkam S, Nikkhah G, Kirch RD, et al. Correlation of (18)F-fluoroethyl tyrosine positron-emission tomography uptake values and histomorphological findings by stereotactic serial biopsy in newly diagnosed brain tumors using a refined software tool. Onco Targets Ther. 2015;8:3803-15. doi:10.2147/OTT.S87126.

91. Langen KJ, Stoffels G, Filss C, Heinzel A, Stegmayr C, Lohmann P, et al. Im-aging of amino acid transport in brain tumours: Positron emission tomography with O-(2-[(18)F]fluoroethyl)-L-tyrosine (FET). Methods. 2017;130:124-34. doi:10.1016/j.ymeth.2017.05.019.

92. Kracht LW, Miletic H, Busch S, Jacobs AH, Voges J, Hoevels M, et al. Deline-ation of Brain Tumor Extent with [11C]l-Methionine Positron Emission Tomog-raphy: Local Comparison with Stereotactic Histopathology. Clinical Cancer Re-search. 2004;10:7163-70. doi:10.1158/1078-0432.CCR-04-0262.

77

93. Pauleit D, Floeth F, Hamacher K, Riemenschneider MJ, Reifenberger G, Muller HW, et al. O-(2-[18F]fluoroethyl)-L-tyrosine PET combined with MRI im-proves the diagnostic assessment of cerebral gliomas. Brain. 2005;128:678-87. doi:10.1093/brain/awh399.

94. Pallud J, Audureau E, Blonski M, Sanai N, Bauchet L, Fontaine D, et al. Epilep-tic seizures in diffuse low-grade gliomas in adults. Brain. 2014;137:449-62. doi:10.1093/brain/awt345.

95. Fisher RS, van Emde Boas W, Blume W, Elger C, Genton P, Lee P, et al. Epi-leptic seizures and epilepsy: definitions proposed by the International League Against Epilepsy (ILAE) and the International Bureau for Epilepsy (IBE). Epi-lepsia. 2005;46:470-2. doi:10.1111/j.0013-9580.2005.66104.x.

96. Scheffer IE, Berkovic S, Capovilla G, Connolly MB, French J, Guilhoto L, et al. ILAE classification of the epilepsies: Position paper of the ILAE Commission for Classification and Terminology. Epilepsia. 2017;58:512-21. doi:10.1111/epi.13709.

97. Falco-Walter JJ, Scheffer IE, Fisher RS. The new definition and classification of seizures and epilepsy. Epilepsy Res. 2018;139:73-9. doi:10.1016/j.eplepsyres.2017.11.015.

98. van Breemen MS, Wilms EB, Vecht CJ. Epilepsy in patients with brain tumours: epidemiology, mechanisms, and management. Lancet Neurol. 2007;6:421-30. doi:10.1016/S1474-4422(07)70103-5.

99. Smits A, Jakola AS. Clinical Presentation in Diffuse Low-Grade Gliomas. In: Duffau H, editor. Diffuse Low-Grade Gliomas in Adults. Cham: Springer Inter-national Publishing; 2017. p. 199-213.

100. Chang EF, Potts MB, Keles GE, Lamborn KR, Chang SM, Barbaro NM, et al. Seizure characteristics and control following resection in 332 patients with low-grade gliomas. J Neurosurg. 2008;108:227-35. doi:10.3171/JNS/2008/108/2/0227.

101. Duffau H. Diffuse low-grade gliomas and neuroplasticity. Diagn Interv Imaging. 2014;95:945-55. doi:10.1016/j.diii.2014.08.001.

102. Mukand JA, Blackinton DD, Crincoli MG, Lee JJ, Santos BB. Incidence of neu-rologic deficits and rehabilitation of patients with brain tumors. Am J Phys Med Rehabil. 2001;80:346-50.

103. Vercueil L. Brain tumor epilepsy: a reappraisal and six remaining issues to be debated. Rev Neurol (Paris). 2011;167:751-61. doi:10.1016/j.neu-rol.2011.08.007.

104. Jobst BC, Siegel AM, Thadani VM, Roberts DW, Rhodes HC, Williamson PD. Intractable seizures of frontal lobe origin: clinical characteristics, localizing signs, and results of surgery. Epilepsia. 2000;41:1139-52.

105. Kaloshi G, Psimaras D, Mokhtari K, Dehais C, Houillier C, Marie Y, et al. Su-pratentorial low-grade gliomas in older patients. Neurology. 2009;73:2093-8. doi:10.1212/WNL.0b013e3181c6781e.

106. Huberfeld G, Vecht CJ. Seizures and gliomas--towards a single therapeutic ap-proach. Nat Rev Neurol. 2016;12:204-16. doi:10.1038/nrneurol.2016.26.

107. Soffietti R, Baumert BG, Bello L, von Deimling A, Duffau H, Frenay M, et al. Guidelines on management of low-grade gliomas: report of an EFNS-EANO Task Force. Eur J Neurol. 2010;17:1124-33. doi:10.1111/j.1468-1331.2010.03151.x.

108. Gschwind M, Picard F. Ecstatic Epileptic Seizures: A Glimpse into the Multiple Roles of the Insula. Frontiers in behavioral neuroscience. 2016;10:21. doi:10.3389/fnbeh.2016.00021.

78

109. Meola A, Yeh FC, Fellows-Mayle W, Weed J, Fernandez-Miranda JC. Human Connectome-Based Tractographic Atlas of the Brainstem Connections and Sur-gical Approaches. Neurosurgery. 2016;79:437-55. doi:10.1227/NEU.0000000000001224.

110. Asheim Hansen B, Brodtkorb E. Partial epilepsy with "ecstatic" seizures. Epi-lepsy Behav. 2003;4:667-73.

111. Guillory SA, Bujarski KA. Exploring emotions using invasive methods: review of 60 years of human intracranial electrophysiology. Social cognitive and affec-tive neuroscience. 2014;9:1880-9. doi:10.1093/scan/nsu002.

112. Mulder DW, Daly D. Psychiatric symptoms associated with lesions of temporal lobe. Journal of the American Medical Association. 1952;150:173-6.

113. Isnard J, Guenot M, Sindou M, Mauguiere F. Clinical manifestations of insular lobe seizures: a stereo-electroencephalographic study. Epilepsia. 2004;45:1079-90. doi:10.1111/j.0013-9580.2004.68903.x.

114. Engel J, Jr., Thompson PM, Stern JM, Staba RJ, Bragin A, Mody I. Connec-tomics and epilepsy. Current opinion in neurology. 2013;26:186-94. doi:10.1097/WCO.0b013e32835ee5b8.

115. Gotman J. Epileptic networks studied with EEG-fMRI. Epilepsia. 2008;49 Suppl 3:42-51. doi:10.1111/j.1528-1167.2008.01509.x.

116. Laufs H, Hamandi K, Salek-Haddadi A, Kleinschmidt AK, Duncan JS, Lemieux L. Temporal lobe interictal epileptic discharges affect cerebral activity in "de-fault mode" brain regions. Human brain mapping. 2007;28:1023-32. doi:10.1002/hbm.20323.

117. Richardson MP. Large scale brain models of epilepsy: dynamics meets connec-tomics. J Neurol Neurosurg Psychiatry. 2012;83:1238-48. doi:10.1136/jnnp-2011-301944.

118. Varotto G, Tassi L, Franceschetti S, Spreafico R, Panzica F. Epileptogenic net-works of type II focal cortical dysplasia: a stereo-EEG study. Neuroimage. 2012;61:591-8. doi:10.1016/j.neuroimage.2012.03.090.

119. Picard F. State of belief, subjective certainty and bliss as a product of cortical dysfunction. Cortex. 2013;49:2494-500. doi:10.1016/j.cortex.2013.01.006.

120. Picard F, Craig AD. Ecstatic epileptic seizures: a potential window on the neural basis for human self-awareness. Epilepsy & behavior : E&B. 2009;16:539-46. doi:10.1016/j.yebeh.2009.09.013.

121. Spencer SS. Neural networks in human epilepsy: evidence of and implications for treatment. Epilepsia. 2002;43:219-27.

122. Bullmore E, Sporns O. Complex brain networks: graph theoretical analysis of structural and functional systems. Nat Rev Neurosci. 2009;10:186-98. doi:10.1038/nrn2575.

123. Wig GS. Segregated Systems of Human Brain Networks. Trends Cogn Sci. 2017;21:981-96. doi:10.1016/j.tics.2017.09.006.

124. Besson P, Bandt SK, Proix T, Lagarde S, Jirsa VK, Ranjeva JP, et al. Anatomic consistencies across epilepsies: a stereotactic-EEG informed high-resolution structural connectivity study. Brain. 2017;140:2639-52. doi:10.1093/brain/awx181.

125. Catani M, Dell'acqua F, Vergani F, Malik F, Hodge H, Roy P, et al. Short frontal lobe connections of the human brain. Cortex. 2012;48:273-91. doi:10.1016/j.cortex.2011.12.001.

126. Catani M, ffytche DH. The rises and falls of disconnection syndromes. Brain. 2005;128:2224-39. doi:10.1093/brain/awh622.

79

127. De Benedictis A, Duffau H. Brain hodotopy: from esoteric concept to practical surgical applications. Neurosurgery. 2011;68:1709-23; discussion 23. doi:10.1227/NEU.0b013e3182124690.

128. Sanai N, Berger MS. Surgical oncology for gliomas: the state of the art. Nat Rev Clin Oncol. 2018;15:112-25. doi:10.1038/nrclinonc.2017.171.

129. Goldbrunner R, Ruge M, Kocher M, Lucas CW, Galldiks N, Grau S. The Treat-ment of Gliomas in Adulthood. Dtsch Arztebl Int. 2018;115:356-64. doi:10.3238/arztebl.2018.0356.

130. Jakola AS, Myrmel KS, Kloster R, Torp SH, Lindal S, Unsgard G, et al. Com-parison of a strategy favoring early surgical resection vs a strategy favoring watchful waiting in low-grade gliomas. JAMA. 2012;308:1881-8. doi:10.1001/jama.2012.12807.

131. Yordanova YN, Moritz-Gasser S, Duffau H. Awake surgery for WHO Grade II gliomas within "noneloquent" areas in the left dominant hemisphere: toward a "supratotal" resection. Clinical article. J Neurosurg. 2011;115:232-9. doi:10.3171/2011.3.JNS101333.

132. Swanson KR, Alvord EC, Jr., Murray JD. A quantitative model for differential motility of gliomas in grey and white matter. Cell Prolif. 2000;33:317-29.

133. Pichlmeier U, Bink A, Schackert G, Stummer W, Group ALAGS. Resection and survival in glioblastoma multiforme: an RTOG recursive partitioning analysis of ALA study patients. Neuro Oncol. 2008;10:1025-34. doi:10.1215/15228517-2008-052.

134. Claes A, Idema AJ, Wesseling P. Diffuse glioma growth: a guerilla war. Acta Neuropathol. 2007;114:443-58. doi:10.1007/s00401-007-0293-7.

135. Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher B, Taphoorn MJ, et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med. 2005;352:987-96. doi:10.1056/NEJMoa043330.

136. Hoang-Xuan K, Capelle L, Kujas M, Taillibert S, Duffau H, Lejeune J, et al. Temozolomide as initial treatment for adults with low-grade oligodendrogliomas or oligoastrocytomas and correlation with chromosome 1p deletions. J Clin On-col. 2004;22:3133-8. doi:10.1200/JCO.2004.10.169.

137. Buckner JC, Shaw EG, Pugh SL, Chakravarti A, Gilbert MR, Barger GR, et al. Radiation plus Procarbazine, CCNU, and Vincristine in Low-Grade Glioma. N Engl J Med. 2016;374:1344-55. doi:10.1056/NEJMoa1500925.

138. Aldape K, Zadeh G, Mansouri S, Reifenberger G, von Deimling A. Glioblas-toma: pathology, molecular mechanisms and markers. Acta Neuropathol. 2015;129:829-48. doi:10.1007/s00401-015-1432-1.

139. Ohgaki H, Kleihues P. Epidemiology and etiology of gliomas. Acta Neuropa-thol. 2005;109:93-108. doi:10.1007/s00401-005-0991-y.

140. Kita D, Ciernik IF, Vaccarella S, Franceschi S, Kleihues P, Lutolf UM, et al. Age as a predictive factor in glioblastomas: population-based study. Neuroepi-demiology. 2009;33:17-22. doi:10.1159/000210017.

141. Wang Y, Yang J, Zheng H, Tomasek GJ, Zhang P, McKeever PE, et al. Expres-sion of mutant p53 proteins implicates a lineage relationship between neural stem cells and malignant astrocytic glioma in a murine model. Cancer Cell. 2009;15:514-26. doi:10.1016/j.ccr.2009.04.001.

142. Christensen BC, Houseman EA, Marsit CJ, Zheng S, Wrensch MR, Wiemels JL, et al. Aging and environmental exposures alter tissue-specific DNA methylation dependent upon CpG island context. PLoS Genet. 2009;5:e1000602. doi:10.1371/journal.pgen.1000602.

80

143. Nagai H, Li Y, Hatano S, Toshihito O, Yuge M, Ito E, et al. Mutations and ab-errant DNA methylation of the PROX1 gene in hematologic malignancies. Genes Chromosomes Cancer. 2003;38:13-21. doi:10.1002/gcc.10248.

144. Azar S, Leventoux N, Ripoll C, Rigau V, Goze C, Lorcy F, et al. Cellular and molecular characterization of IDH1-mutated diffuse low grade gliomas reveals tumor heterogeneity and absence of EGFR/PDGFRalpha activation. Glia. 2018;66:239-55. doi:10.1002/glia.23240.

145. Jackson RJ, Fuller GN, Abi-Said D, Lang FF, Gokaslan ZL, Shi WM, et al. Lim-itations of stereotactic biopsy in the initial management of gliomas. Neuro On-col. 2001;3:193-200. doi:10.1093/neuonc/3.3.193.

146. Giese A, Westphal M. Glioma invasion in the central nervous system. Neurosur-gery. 1996;39:235-50; discussion 50-2.

147. Giese A, Bjerkvig R, Berens ME, Westphal M. Cost of migration: invasion of malignant gliomas and implications for treatment. J Clin Oncol. 2003;21:1624-36. doi:10.1200/JCO.2003.05.063.

148. Mandonnet E, Capelle L, Duffau H. Extension of paralimbic low grade gliomas: toward an anatomical classification based on white matter invasion patterns. J Neurooncol. 2006;78:179-85. doi:10.1007/s11060-005-9084-y.

149. Magnain C, Augustinack JC, Reuter M, Wachinger C, Frosch MP, Ragan T, et al. Blockface histology with optical coherence tomography: a comparison with Nissl staining. Neuroimage. 2014;84:524-33. doi:10.1016/j.neu-roimage.2013.08.072.

150. Oishi K, Zilles K, Amunts K, Faria A, Jiang H, Li X, et al. Human brain white matter atlas: identification and assignment of common anatomical structures in superficial white matter. Neuroimage. 2008;43:447-57. doi:10.1016/j.neuroimage.2008.07.009.

151. Palm C, Axer M, Grassel D, Dammers J, Lindemeyer J, Zilles K, et al. Towards ultra-high resolution fibre tract mapping of the human brain - registration of po-larised light images and reorientation of fibre vectors. Front Hum Neurosci. 2010;4:9. doi:10.3389/neuro.09.009.2010.

152. Hassani Z, Hugnot, JP. Deciphering the molecular and cellular basis for dissem-ination of diffuse low-grade gliomas. In: Duffau H, editor. Diffuse Low-Grade Gliomas in Adults. London: Springer; 2013. p. 91-8.

153. Persson AI, Petritsch C, Swartling FJ, Itsara M, Sim FJ, Auvergne R, et al. Non-stem cell origin for oligodendroglioma. Cancer Cell. 2010;18:669-82. doi:10.1016/j.ccr.2010.10.033.

154. Zlatescu MC, TehraniYazdi A, Sasaki H, Megyesi JF, Betensky RA, Louis DN, et al. Tumor location and growth pattern correlate with genetic signature in oli-godendroglial neoplasms. Cancer Res. 2001;61:6713-5.

155. Guichet P, Hugnot, JP. Cellular origin of Grade II gliomas. In: Duffau H, editor. Diffuse Low-Grade Gliomas in Adults. London: Springer; 2013. p. 75-90.

156. de Castro F, Bribian A. The molecular orchestra of the migration of oligoden-drocyte precursors during development. Brain Res Brain Res Rev. 2005;49:227-41. doi:10.1016/j.brainresrev.2004.12.034.

157. Miyamoto Y, Yamauchi J, Tanoue A. Cdk5 phosphorylation of WAVE2 regu-lates oligodendrocyte precursor cell migration through nonreceptor tyrosine ki-nase Fyn. J Neurosci. 2008;28:8326-37. doi:10.1523/JNEUROSCI.1482-08.2008.

158. Fillmore HL, VanMeter TE, Broaddus WC. Membrane-type matrix metallopro-teinases (MT-MMPs): expression and function during glioma invasion. J Neu-rooncol. 2001;53:187-202.

81

159. Mentlein R, Hattermann K, Held-Feindt J. Lost in disruption: role of proteases in glioma invasion and progression. Biochim Biophys Acta. 2012;1825:178-85. doi:10.1016/j.bbcan.2011.12.001.

160. Duffau H, Taillandier L. New concepts in the management of diffuse low-grade glioma: Proposal of a multistage and individualized therapeutic approach. Neuro Oncol. 2015;17:332-42. doi:10.1093/neuonc/nou153.

161. Capelle L, Fontaine D, Mandonnet E, Taillandier L, Golmard JL, Bauchet L, et al. Spontaneous and therapeutic prognostic factors in adult hemispheric World Health Organization Grade II gliomas: a series of 1097 cases: clinical article. J Neurosurg. 2013;118:1157-68. doi:10.3171/2013.1.JNS121.

162. McGirt MJ, Chaichana KL, Attenello FJ, Weingart JD, Than K, Burger PC, et al. Extent of surgical resection is independently associated with survival in pa-tients with hemispheric infiltrating low-grade gliomas. Neurosurgery. 2008;63:700-7; author reply 7-8. doi:10.1227/01.NEU.0000325729.41085.73.

163. Sanai N, Berger MS. Glioma extent of resection and its impact on patient out-come. Neurosurgery. 2008;62:753-64; discussion 264-6. doi:10.1227/01.neu.0000318159.21731.cf.

164. Smith JS, Chang EF, Lamborn KR, Chang SM, Prados MD, Cha S, et al. Role of extent of resection in the long-term outcome of low-grade hemispheric glio-mas. J Clin Oncol. 2008;26:1338-45. doi:10.1200/JCO.2007.13.9337.

165. Pedeutour-Braccini Z, Burel-Vandenbos F, Goze C, Roger C, Bazin A, Costes-Martineau V, et al. Microfoci of malignant progression in diffuse low-grade gli-omas: towards the creation of an intermediate grade in glioma classification? Virchows Arch. 2015;466:433-44. doi:10.1007/s00428-014-1712-5.

166. Mandonnet E. Biomathematical modeling of DLGG behavior. In: Duffau H, ed-itor. Diffuse Low-Grade Gliomas in Adults. London: Springer; 2013. p. 447-55.

167. Kracht LW, Friese M, Herholz K, Schroeder R, Bauer B, Jacobs A, et al. Methyl-[11C]- l-methionine uptake as measured by positron emission tomography cor-relates to microvessel density in patients with glioma. Eur J Nucl Med Mol Im-aging. 2003;30:868-73. doi:10.1007/s00259-003-1148-7.

168. Dhermain FG, Hau P, Lanfermann H, Jacobs AH, van den Bent MJ. Advanced MRI and PET imaging for assessment of treatment response in patients with gli-omas. Lancet Neurol. 2010;9:906-20. doi:10.1016/S1474-4422(10)70181-2.

169. Chung JK, Kim YK, Kim SK, Lee YJ, Paek S, Yeo JS, et al. Usefulness of 11C-methionine PET in the evaluation of brain lesions that are hypo- or isometabolic on 18F-FDG PET. Eur J Nucl Med Mol Imaging. 2002;29:176-82.

170. Langen KJ, Muhlensiepen H, Holschbach M, Hautzel H, Jansen P, Coenen HH. Transport mechanisms of 3-[123I]iodo-alpha-methyl-L-tyrosine in a human gli-oma cell line: comparison with [3H]methyl]-L-methionine. J Nucl Med. 2000;41:1250-5.

171. Tanaka K, Yamamoto Y, Maeda Y, Yamamoto H, Kudomi N, Kawai N, et al. Correlation of 4'-[methyl-(11)C]-thiothymidine uptake with Ki-67 immuno-histochemistry and tumor grade in patients with newly diagnosed gliomas in comparison with (11)C-methionine uptake. Ann Nucl Med. 2016;30:89-96. doi:10.1007/s12149-015-1035-x.

172. Kinoshita M, Arita H, Okita Y, Kagawa N, Kishima H, Hashimoto N, et al. Comparison of diffusion tensor imaging and (11)C-methionine positron emis-sion tomography for reliable prediction of tumor cell density in gliomas. J Neu-rosurg. 2016;125:1136-42. doi:10.3171/2015.11.JNS151848.

82

173. Arita H, Kinoshita M, Kagawa N, Fujimoto Y, Kishima H, Hashimoto N, et al. (1)(1)C-methionine uptake and intraoperative 5-aminolevulinic acid-induced fluorescence as separate index markers of cell density in glioma: a stereotactic image-histological analysis. Cancer. 2012;118:1619-27. doi:10.1002/cncr.26445.

174. Okita Y, Kinoshita M, Goto T, Kagawa N, Kishima H, Shimosegawa E, et al. (11)C-methionine uptake correlates with tumor cell density rather than with mi-crovessel density in glioma: A stereotactic image-histology comparison. Neu-roimage. 2010;49:2977-82. doi:10.1016/j.neuroimage.2009.11.024.

175. Sadeghi N, D'Haene N, Decaestecker C, Levivier M, Metens T, Maris C, et al. Apparent diffusion coefficient and cerebral blood volume in brain gliomas: re-lation to tumor cell density and tumor microvessel density based on stereotactic biopsies. AJNR Am J Neuroradiol. 2008;29:476-82. doi:10.3174/ajnr.A0851.

176. Shweiki D, Neeman M, Itin A, Keshet E. Induction of vascular endothelial growth factor expression by hypoxia and by glucose deficiency in multicell sphe-roids: implications for tumor angiogenesis. Proc Natl Acad Sci U S A. 1995;92:768-72.

177. Fischer I, Gagner JP, Law M, Newcomb EW, Zagzag D. Angiogenesis in glio-mas: biology and molecular pathophysiology. Brain Pathol. 2005;15:297-310.

178. Chan AS, Leung SY, Wong MP, Yuen ST, Cheung N, Fan YW, et al. Expression of vascular endothelial growth factor and its receptors in the anaplastic progres-sion of astrocytoma, oligodendroglioma, and ependymoma. Am J Surg Pathol. 1998;22:816-26.

179. Sturm JW, Andermann F, Berkovic SF. "Pressure to laugh": an unusual epileptic symptom associated with small hypothalamic hamartomas. Neurology. 2000;54:971-3.

180. Kameyama S, Masuda H, Murakami H. Ictogenesis and symptomatogenesis of gelastic seizures in hypothalamic hamartomas: an ictal SPECT study. Epilepsia. 2010;51:2270-9. doi:10.1111/j.1528-1167.2010.02739.x.

181. Ghaziri J, Tucholka A, Girard G, Houde JC, Boucher O, Gilbert G, et al. The Corticocortical Structural Connectivity of the Human Insula. Cereb Cortex. 2017;27:1216-28. doi:10.1093/cercor/bhv308.

182. Amberger VR, Hensel T, Ogata N, Schwab ME. Spreading and migration of human glioma and rat C6 cells on central nervous system myelin in vitro is cor-related with tumor malignancy and involves a metalloproteolytic activity. Can-cer Res. 1998;58:149-58.

183. Mandonnet E, Delattre JY, Tanguy ML, Swanson KR, Carpentier AF, Duffau H, et al. Continuous growth of mean tumor diameter in a subset of grade II gli-omas. Ann Neurol. 2003;53:524-8. doi:10.1002/ana.10528.

184. Rees J, Watt H, Jager HR, Benton C, Tozer D, Tofts P, et al. Volumes and growth rates of untreated adult low-grade gliomas indicate risk of early malignant trans-formation. Eur J Radiol. 2009;72:54-64. doi:10.1016/j.ejrad.2008.06.013.

185. Bergers G, Benjamin LE. Tumorigenesis and the angiogenic switch. Nat Rev Cancer. 2003;3:401-10. doi:10.1038/nrc1093.

186. Zhang L, He L, Lugano R, Roodakker K, Bergquist M, Smits A, et al. IDH-mutation Status is Associated with Distinct Vascular Gene Expression Signa-tures in Lower Grade Gliomas. Neuro Oncol. 2018. doi:10.1093/neuonc/noy088.

187. Anderson JC, McFarland BC, Gladson CL. New molecular targets in angiogenic vessels of glioblastoma tumours. Expert Rev Mol Med. 2008;10:e23. doi:10.1017/S1462399408000768.

83

188. Miletic H, Niclou SP, Johansson M, Bjerkvig R. Anti-VEGF therapies for ma-lignant glioma: treatment effects and escape mechanisms. Expert Opin Ther Tar-gets. 2009;13:455-68. doi:10.1517/14728220902806444.

189. Dieterich LC, Mellberg S, Langenkamp E, Zhang L, Zieba A, Salomaki H, et al. Transcriptional profiling of human glioblastoma vessels indicates a key role of VEGF-A and TGFbeta2 in vascular abnormalization. J Pathol. 2012;228:378-90. doi:10.1002/path.4072.

Acta Universitatis UpsaliensisDigital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Medicine 1486

Editor: The Dean of the Faculty of Medicine

A doctoral dissertation from the Faculty of Medicine, UppsalaUniversity, is usually a summary of a number of papers. A fewcopies of the complete dissertation are kept at major Swedishresearch libraries, while the summary alone is distributedinternationally through the series Digital ComprehensiveSummaries of Uppsala Dissertations from the Faculty ofMedicine. (Prior to January, 2005, the series was publishedunder the title “Comprehensive Summaries of UppsalaDissertations from the Faculty of Medicine”.)

Distribution: publications.uu.seurn:nbn:se:uu:diva-356846

ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2018