**The Role of Stem Cells in the Glioma Growth**

Sergio Garcia, Vinicius Kannen and Luciano Neder

*Faculty of Medicine of Ribeirao Preto University of Sao Paulo Brazil* 

#### **1. Introduction**

188 Glioma – Exploring Its Biology and Practical Relevance

Weick, J.P., Austin Johnson, M. & Zhang, S.C. (2009). Developmental regulation of human

potential channels. *Stem Cells*, Vol. 27, No. 12, pp. 2906-2916, ISSN 1549-4918 Wisnoskey, B.J., Sinkins, W.G. & Schilling, W.P. (2003). Activation of vanilloid receptor type

Wissenbach, U., Niemeyer, B.A., Fixemer, T., Schneidewind, A., Trost, C., Cavalie, A., Reus,

*of Biological Chemistry*, Vol. 276, No. 22, pp. 19461–19468, ISSN 0021-9258 Wondergem, R. & Bartley, J.W. (2009). Menthol increases human glioblastoma intracellular

Wondergem, R., Ecay, T.W., Mahieu, F., Owsianik, G. & Nilius, B. (2008). HGF/SF and

Xu, X.Z., Moebius, F., Gill, D.L. & Montell, C. (2001). Regulation of melastatin, a TRP-related

Yang, H., Mergler, S., Sun, X., Wang, Z., Lu, L., Bonanno, J.A., Pleyer, U. & Reinach, P.S.

*Biological Chemistry* Vol. 280, No. 37, pp. 32230-32237, ISSN 0021-9258 Yu, P.C., Gu, S.Y., Bu, J.W. & Du, J.L. (2010). TRPC1 is essential for in vivo angiogenesis in zebrafish. *Circulation Research*, Vol. 106, No. 7, pp. 1221-1232, ISSN 0009-7330 Zhang, X.P., Zheng, G., Zou, L., Liu, H.L., Hou, L.H., Zhou, P., Yin, D.D., Zheng, Q.J., Liang,

*Biochemical Journal*, Vol. 372, No. 2, pp. 517–528, ISSN 0264-6021

No. 1, pp. 90, ISSN 1021- 7770

ISSN 0006-291X

ISSN 0027-8424

2, pp. 101-108, ISSN 0300-8177

embryonic stem cell-derived neurons by calcium entry via transient receptor

1 in the endoplasmic reticulum fails to activate store-operated Ca2+ entry.

K., Meese, E., Bonkhoff, H. & Flockerzi, V. (2001). Expression of CaT-like, a novel calcium-selective channel, correlates with the malignancy of prostate cancer. *Journal* 

Ca2+, BK channel activity and cell migration. *Journal of Biomedical Science*, Vol. 16,

menthol increase human glioblastoma cell calcium and migration. *Biochemical and Biophysical Research Communications*, Vol. 372, No. 1, pp. 210-215, ISSN 0006-291X Wu, H.M., Yuan, Y., Zawieja, D.C., Tinsley, J. & Granger, H.J. (1999). Role of phospholipase

C, protein kinase C, and calcium in VEGF-induced venular hyperpermeability. *American Journal of Physiology*, Vol. 276, No. 2, pp. H535–H542, ISSN 0363-6135 Xin, H., Tanaka, H., Yamaguchi, M., Takemori, S., Nakamura, A. & Kohama, K. (2005).

Vanilloid receptor expressed in the sarcoplasmic reticulum of rat skeletal muscle. *Biochemical and Biophysical Research Communications*, Vol 332, No. 3, pp. 756–762,

protein, through interaction with a cytoplasmic isoform. *Proceedings of the National Academy of Sciences of the United States of America*, Vol. 98, No. 19, pp. 10692–10697,

(2005). TRPC4 knockdown suppresses epidermal growth factor-induced storeoperated channel activation and growth in human corneal epithelial cells. *Journal of* 

L., Zhang, S.Z., Feng, L., Yao, L.B., Yang, A.G., Han, H. & Chen, J.Y. (2008). Notch activation promotes cell proliferation and the formation of neural stem cell-like colonies in human glioma cells. *Molecular and Cellular Biochemistry,* Vol. 307, No. 1Malignant glioma is the most common type of primary brain tumor and represents one of the most lethal cancers. In contrast to the long-standing and well-defined histopathology, the underlying molecular and genetic bases for gliomas are less known. (Collins, 2004; Dai & Holland, 2001).

As some other human cancers, particularly central nervous tumors are highly heterogeneous. Primarily because of its diffuse nature, relatively little is known about the processes by which they develop (Hulleman & Helin, 2005). Thus, the traditional evolution concept of tumors arising from a single mutated cell has limitations in explaining the heterogeneity observed in a single tumor nest.

Recent decades have seen only limited progress in treatment trials and basic research on human glioma, the most common central nervous malignancy (Huang et al., 2008). Unfortunately, for such gliomas, tumor recurrence after treatment is the rule due to the infiltrative nature of these tumors and the presence of cellular populations with ability to escape therapies and drive tumor recurrence and progression. At least in some cases, these resistant cells exhibit stem cell properties (Frosina, 2011). For these reasons the comprehension of the current knowledge of cancer stem cells (CSC) in relation to gliomas origin, growth and treatment is crucial. As the stem cells (for glioma, neuronal stem cells) are more susceptible to mutation, they become altered easily for their genetic composition and therefore act as the source of cancer/glioma cells. They are not actually a separate cell type and in most cases they are misinterpreted as cancer stem cells (in brain, they are glioma stem cells).

#### **2. Glioma and the concept of cancer stem cells**

For a long time it has been known that there are subpopulations of cells within solid tumors that contain different biological behaviors. Among these subpopulations, accumulating evidence supports the existence of the so-called cancer stem cells (CSCs), because these tumor cells possess stem cell properties, possibly being responsible for the initiation, growth and recurrence of tumors. Apparent similarities with non-transformed stem cells, including high self-renewal capacity and the ability to generate differentiated progeny of several cellular lineages, have led to the proposal that stem cell-like cancer cells may either originate from adult undifferentiated stem and progenitor cells or that these properties are being

The Role of Stem Cells in the Glioma Growth 191

cell markers, including the transcription factors as Oct4, Sox2, Nanog and Klf4. In line with these findings, in high grade gliomas, mesodermal- and endodermal-specific transcription factors were detected together with neural proteins, a combination of lineage markers not normally present in the central nervous system. These findings demonstrate a general deregulated expression of neural and pluripotent stem cell traits in malignant human

Primarily because of the diffuse nature of gliomas, relatively little is known about the processes by which they develop (Hulleman & Helin, 2005). The concept of stem cells originating gliomas is gaining increased recognition in neuro-oncology (Richj & Eyler, 2008). Until recently, the paradigm of a tumor-initiating stem cell was confined to hematopoietic malignancies where the hierarchical lineages of stem progenitor cells are well established. Nevertheless, the demonstration of persistent stem cells and cycling progenitors in the adult brain is coupled with the expansion of the cancer stem cell concept to solid tumors, leading to the exploration of "stemness" within gliomas. Emerging data are highly suggestive of the subsistence of transformed multipotential cells within a glioma, with a subfraction of cells exhibiting increased efficiency at tumor initiation stage. However, data in support of the true glioma stem cells are inconclusive to date, particularly in respect to the functional

Thus, it may be considered that currently it is conceivable thought that malignant gliomas may arise from neural stem cells and appear to contain tumor stem cells. It is thought that normal stem cells live in protected pockets of the body called *niches*, where they divide infrequently to avoid accumulating damaging mutations. Upon injury or in response to normal stimuli, stem cells are mobilized to divide (Gilbertson, 2006). Hence, parallel to the role that normal stem cells play in organogenesis, stem cells are thought to be crucial for

The normal adult neural stem cells (NSCs) arise from radial glia (RG) within the central nervous system (Weiner**,** 2008). The RG progeny includes all the main lineages of the CNS: neurons, astrocytes, oligodendrocytes, ependymocytes and adult neural stem cells (Malatesta et al. 2003). By comparing the gene expression profiles of ependymomas with those of cells in the normal developing nervous system, it was possible to identify the RG as candidate stem cells of this brain tumor (Gilbertson, 2006). Furthermore, RG cells produce neurons in addition to glia during central nervous system development in all vertebrates

Until recently, it was thought that ependymomas originated from neuroepithelial cells, glioblastomas from abnormal astrocytes, and medulloblastomas from primitive cells in the external granular layer, but there is now evidence that all tumors can originate from a special type of stem cell called "radial glial cell" (RGC). It is interesting to note that in the human brain, most of stem cells are located in the subventricular zones (SVZ). Both supraand infratentorially and when stimulated with carcinogens, cells in the SVZ become tumorigenic faster than those located elsewhere. In the SVZ, stem cells exist in the form of RGCs, which remain quiescent until they receive transformational signals. It is not clear whether RGCs, after receiving transformation signals, return to their initial stem cell configuration and then become tumorigenic or they transform to tumor progenitor cells

gliomas.

tumorigenesis.

**3. Stem cells and the origin of gliomas** 

characterization of these cells. (Panagiotakos & Tabar, 2007).

and are also involved in reparative process (Weiner, 2008).

expressed as an effect of the genetic alterations which drive tumorigenicity (Reya et al., 2001; Gilbertson, 2006; Das et al., 2008). Basically, the CSCs, which have also been described as tumor initiating cells or tumor propagating cells, are tumor cells that self-renew and propagate tumors phenotypically similar to the parental tumor (Li et al., 2009). Furthermore, recent studies have suggested that CSC cause tumor recurrence based on their resistance to radiotherapy and chemotherapy (Inoue et al, 2010).

Although considerable controversy still surrounds the existence, behaviors and even the nomenclature of CSCs, there is no doubt that populations of cells with stem-like properties do exist inside several solid and non-solid tumors, including brain cancers. So, despite the fact that CSCs in solid tumors have not yet been precisely identified, the "CSC hypothesis" opens a new paradigm in understanding the biology of cancers. For this reason, the search for the tumor stem cells that may originate and perpetuate the tumor growth has been receiving great attention in the literature (Sanchez-Martin, 2008), but the available knowledge on this issue with regards to the gliomas is scant. Particularly, the exact identity and cell(s) of origin of the so-called glioma stem cell remains elusive (Park & Rich, 2009). Vescovi (2006) offered a functional definition of brain tumor stem cells, namely: brain tumor cells should qualify as stem cells if they show cancer-initiating ability upon orthotopic implantation, extensive self-renewal ability demonstrated either ex vivo or in vivo, karyotypic or genetic alterations, aberrant differentiation properties, capacity to generate non-tumorigenic end cells, and multilineage differentiation capacity. Furthermore, parallels between normal neurogenesis and brain tumorigenesis have been proposed (Singh et al., 2004). It has been more recently confirmed that cancer stem cells from glioblastomas share some characteristics with normal neural stem cells including the expression of neural stem cell markers, the capacity for self-renewal and long term proliferation, the formation of neurospheres, and the ability to differentiate into multiple nervous system lineages (astrocytic, oligodendrocytic and/or neuronal differentiation) (Li et al., 2009).

Among the so far evaluated stem cell markers, the transmembrane protein CD133 has been widely used to isolate putative CSC populations in several cancer types. In fact, CD133 is currently one of the best markers to characterize CSCs (Singh et al., 2004). In both human glioblastomas (GBMs) and medulloblastomas, the expression of the neural stem cell marker CD133 (also known as prominin 1) has been associated with both tumor initiation capacity and radioresistance (Pérez Castillo et al., 2008). Extensive computational comparisons with a compendium of published gene expression profiles revealed that the CD133 gene signature transcriptionally resembles human embryonic stem cells and in vitro cultured GBMs stem cells (GSC), and this signature successfully distinguishes GBMs from lower-grade gliomas. Moreover, the CD133 gene signature identifies an aggressive subtype of GBMs seen in younger patients with a shorter survival (Yan et al., 2011), confirming previous observations that Glioma stem cells are more aggressive in recurrent tumors (Huang et al., 2008). Nevertheless, it must be pointed out that the use of CD133 as a unique glioma stem cell marker is probably not sufficient to tag the whole self-renewing tumor cell reservoir (Clément et al., 2009).

Holmberg et al (2011) have recently characterized human gliomas in various malignancy grades according to the expression of stem cell regulatory proteins. These authors have shown that cells in high grade glioma co-express an array of markers defining neural stem cells (NSCs) and that these proteins can fulfill similar functions in tumor cells as in NSCs. In contrast to NSCs, the glioma cells co-express neural proteins together with pluripotent stem

expressed as an effect of the genetic alterations which drive tumorigenicity (Reya et al., 2001; Gilbertson, 2006; Das et al., 2008). Basically, the CSCs, which have also been described as tumor initiating cells or tumor propagating cells, are tumor cells that self-renew and propagate tumors phenotypically similar to the parental tumor (Li et al., 2009). Furthermore, recent studies have suggested that CSC cause tumor recurrence based on their resistance to

Although considerable controversy still surrounds the existence, behaviors and even the nomenclature of CSCs, there is no doubt that populations of cells with stem-like properties do exist inside several solid and non-solid tumors, including brain cancers. So, despite the fact that CSCs in solid tumors have not yet been precisely identified, the "CSC hypothesis" opens a new paradigm in understanding the biology of cancers. For this reason, the search for the tumor stem cells that may originate and perpetuate the tumor growth has been receiving great attention in the literature (Sanchez-Martin, 2008), but the available knowledge on this issue with regards to the gliomas is scant. Particularly, the exact identity and cell(s) of origin of the so-called glioma stem cell remains elusive (Park & Rich, 2009). Vescovi (2006) offered a functional definition of brain tumor stem cells, namely: brain tumor cells should qualify as stem cells if they show cancer-initiating ability upon orthotopic implantation, extensive self-renewal ability demonstrated either ex vivo or in vivo, karyotypic or genetic alterations, aberrant differentiation properties, capacity to generate non-tumorigenic end cells, and multilineage differentiation capacity. Furthermore, parallels between normal neurogenesis and brain tumorigenesis have been proposed (Singh et al., 2004). It has been more recently confirmed that cancer stem cells from glioblastomas share some characteristics with normal neural stem cells including the expression of neural stem cell markers, the capacity for self-renewal and long term proliferation, the formation of neurospheres, and the ability to differentiate into multiple nervous system lineages

(astrocytic, oligodendrocytic and/or neuronal differentiation) (Li et al., 2009).

Among the so far evaluated stem cell markers, the transmembrane protein CD133 has been widely used to isolate putative CSC populations in several cancer types. In fact, CD133 is currently one of the best markers to characterize CSCs (Singh et al., 2004). In both human glioblastomas (GBMs) and medulloblastomas, the expression of the neural stem cell marker CD133 (also known as prominin 1) has been associated with both tumor initiation capacity and radioresistance (Pérez Castillo et al., 2008). Extensive computational comparisons with a compendium of published gene expression profiles revealed that the CD133 gene signature transcriptionally resembles human embryonic stem cells and in vitro cultured GBMs stem cells (GSC), and this signature successfully distinguishes GBMs from lower-grade gliomas. Moreover, the CD133 gene signature identifies an aggressive subtype of GBMs seen in younger patients with a shorter survival (Yan et al., 2011), confirming previous observations that Glioma stem cells are more aggressive in recurrent tumors (Huang et al., 2008). Nevertheless, it must be pointed out that the use of CD133 as a unique glioma stem cell marker is probably not sufficient to tag the whole self-renewing tumor cell reservoir

Holmberg et al (2011) have recently characterized human gliomas in various malignancy grades according to the expression of stem cell regulatory proteins. These authors have shown that cells in high grade glioma co-express an array of markers defining neural stem cells (NSCs) and that these proteins can fulfill similar functions in tumor cells as in NSCs. In contrast to NSCs, the glioma cells co-express neural proteins together with pluripotent stem

radiotherapy and chemotherapy (Inoue et al, 2010).

(Clément et al., 2009).

cell markers, including the transcription factors as Oct4, Sox2, Nanog and Klf4. In line with these findings, in high grade gliomas, mesodermal- and endodermal-specific transcription factors were detected together with neural proteins, a combination of lineage markers not normally present in the central nervous system. These findings demonstrate a general deregulated expression of neural and pluripotent stem cell traits in malignant human gliomas.

#### **3. Stem cells and the origin of gliomas**

Primarily because of the diffuse nature of gliomas, relatively little is known about the processes by which they develop (Hulleman & Helin, 2005). The concept of stem cells originating gliomas is gaining increased recognition in neuro-oncology (Richj & Eyler, 2008). Until recently, the paradigm of a tumor-initiating stem cell was confined to hematopoietic malignancies where the hierarchical lineages of stem progenitor cells are well established. Nevertheless, the demonstration of persistent stem cells and cycling progenitors in the adult brain is coupled with the expansion of the cancer stem cell concept to solid tumors, leading to the exploration of "stemness" within gliomas. Emerging data are highly suggestive of the subsistence of transformed multipotential cells within a glioma, with a subfraction of cells exhibiting increased efficiency at tumor initiation stage. However, data in support of the true glioma stem cells are inconclusive to date, particularly in respect to the functional characterization of these cells. (Panagiotakos & Tabar, 2007).

Thus, it may be considered that currently it is conceivable thought that malignant gliomas may arise from neural stem cells and appear to contain tumor stem cells. It is thought that normal stem cells live in protected pockets of the body called *niches*, where they divide infrequently to avoid accumulating damaging mutations. Upon injury or in response to normal stimuli, stem cells are mobilized to divide (Gilbertson, 2006). Hence, parallel to the role that normal stem cells play in organogenesis, stem cells are thought to be crucial for tumorigenesis.

The normal adult neural stem cells (NSCs) arise from radial glia (RG) within the central nervous system (Weiner**,** 2008). The RG progeny includes all the main lineages of the CNS: neurons, astrocytes, oligodendrocytes, ependymocytes and adult neural stem cells (Malatesta et al. 2003). By comparing the gene expression profiles of ependymomas with those of cells in the normal developing nervous system, it was possible to identify the RG as candidate stem cells of this brain tumor (Gilbertson, 2006). Furthermore, RG cells produce neurons in addition to glia during central nervous system development in all vertebrates and are also involved in reparative process (Weiner, 2008).

Until recently, it was thought that ependymomas originated from neuroepithelial cells, glioblastomas from abnormal astrocytes, and medulloblastomas from primitive cells in the external granular layer, but there is now evidence that all tumors can originate from a special type of stem cell called "radial glial cell" (RGC). It is interesting to note that in the human brain, most of stem cells are located in the subventricular zones (SVZ). Both supraand infratentorially and when stimulated with carcinogens, cells in the SVZ become tumorigenic faster than those located elsewhere. In the SVZ, stem cells exist in the form of RGCs, which remain quiescent until they receive transformational signals. It is not clear whether RGCs, after receiving transformation signals, return to their initial stem cell configuration and then become tumorigenic or they transform to tumor progenitor cells

The Role of Stem Cells in the Glioma Growth 193

The available information in regards to the existence of a field phenomenon in gliomas is scant. In malignant gliomas, the high recurrence rates, the characteristically heterogeneous features and frequent diffuse spread within the brain have raised the question of whether malignant gliomas arise monoclonally from a single precursor cell or polyclonally from multiple transformed cells forming confluent clones (Inoue et al., 2008). To address this issue, Kattar et al (1997) have evaluated the clonality of low-grade and malignant gliomas by using polymerase chain reaction (PCR)-based assay for nonrandom X chromosome inactivation using surgical and autopsy material. The same pattern of nonrandom X chromosome inactivation was present in all areas of fifteen of 19 tumors, which were considered as monoclonal, suggesting that low-grade and malignant gliomas are, at least, usually monoclonal tumors, and extensively infiltrating tumors must result from migration

*Gliomatosis cerebri* may shed some light in this issue. It is a rare condition in which the brain is infiltrated by an exceptionally diffusely growing of malignant glial cell population involving at least 2 lobes, though often more extensive, sometimes even affecting infratentorial regions. Kross et al (2002) have evaluated the existence of field cancerization in this affection, since *gliomatosis cerebri* may initiate as an oligoclonal process or result from collision of different gliomas. It was hypothesized that the presence of an identical set of genetic aberrations throughout the lesion would point to monoclonality of the process. In contrast, the finding of non-identical genetic changes in widely separated regions within the neoplasm would support the concept of collision of different mutated clones. For such, the authors used one autopsy case of *gliomatosis cerebri*, from which tissue samples were randomly taken from 24 locations throughout the brain and used for genetic investigation. With this aim, genome-wide screening for chromosomal aberrations was accomplished by comparative genomic hybridization (CGH). The authors found a wide distribution of particular sets of genetic aberrations, supporting the concept of monoclonal tumor proliferation (Kross et al., 2002). Nevertheless, it has been observed and well documented in one clinical case that on the long term, after initial treatment for *gliomatosis cerebri*, one glioblastoma multiforme has developed, and in a location separate from the initial lesion, suggesting that different clonal origin may had occurred (Inoue et al., 2008). More recently, Chen et al (2010) showed that the capacities for self-renewal and tumour initiation in GBM

need not be restricted to a uniform population of stemlike cells.

**cells and field cancerization concepts** 

observed in clinical practice (Bulnes-Sesma, 2006).

**5. The contribution of studies in animal models: Unifying the cancer stem** 

Many genetic alterations have been identified in human gliomas, however, establishing unequivocal correlation between these genetic alterations and gliomagenesis requires accurate animal models for these cancers (Dai & Holland, 2001). Indeed, it is useful and necessary to have animal models for CNS tumors studies allowing to be carried out in different stages of tumor growth, especially in early stages, rare to be detected and

Experimental models of gliomagenesis most commonly used alkylating agents such as Nethyl N-nitrosourea (ENU), which has been considered as a suitable model to study malignant changes. These changes were reported to appear firstly as early neoplastic proliferation (ENP) center, which continues in following stages subsequently progressing to

''microtumors'' until a tumor in itself. (Koestner et al., 1971; Naito et al., 1984).

of tumor cells.

directly. In the cerebellum, depending upon the signals received, RGCs and stem cells may give origin to either ependymoma or medulloblastoma.

Tumours with the highest incidence in humans— medulloblastomas and glioblastomas both originate from abnormal brain stem cells. . Not surprising, both of these tumors are CD133-positive, containing great neuronal differentiation, which makes them prone to be diffuse and resistant to treatment (Castilo, 2010).

#### **4. Gliomas and the field cancerization concept**

It is universally accepted that tumors growth as a clonal evolution from a single cell (Nowell, 1976). The "field cancerization theory" was introduced more than fifty years ago by Slaugher et al (1953), when studying the presence of histologically abnormal tissue surrounding carcinomas. In a classic report on oral cancer, Slaughter called "field cancerization" – a process of repeated exposure of a region's entire tissue area to carcinogenic insult (e.g., tobacco and alcohol), which increases the tissue's risk for developing multiple independent premalignant and malignant foci. The field cancerization hypothesis states that multiple cells form independent tumors on one given tissue, since carcinogenic exposure affects multiple cells in the field (Slaughter *et al*., 1953), and predicts that second primary or synchronous tumours arise from independent genetic events (Garcia et al., 1999). The field cancerization theory may be explained by the concept that a given stem cell that acquires genetic alterations may form a "patch", a clonal unit of altered daughter cells. The proliferation of these patch cells forms expanding fields which gradually displace the normal tissue and, by clonal divergence, ultimately leads to the development of one or more tumors within a contiguous field of preneoplastic cells (Garcia et al., 1999). An important clinical implication is that fields often remain after surgery of the primary tumor and may lead to new cancers, designated presently by clinicians as "a second primary tumor" or "local recurrence," depending on the exact site and time interval (Braakhuis et al., 2003; Ryan, 2007). We had previously discussed how mutated clones from mutated stem cells may spread on tissues and that the field cancerization theory implies that the mutated genotype and molecular changes occur before the appearance of histopathological evidence of malignant cells (Garcia et al., 1999). Therefore, this "anomaly" might be due to changes that occur in a "premalignant" neoplastic condition that was histologically identified as "normal". In the clinical aspect, the field cancerization may have an etiologic role in a substantial number of recurrences. For example, a surgical resection margin that includes a genetically altered field can explain the occurrence of scar recurrence. This explanation suggests that molecular profiling of surgical margins will help reduce scar recurrences. Since multiple independent patches of cancer fields may be present in the same organ exposed to the same insults, clean molecular margins may not necessarily prevent recurrences in the residual organ (Dakubo et al., 2007). Similarly to gliomas, tumor recurrence is a major clinical concern for patients with urothelial carcinoma of the urinary bladder. Traditional morphological analysis is of limited utility for identifying cases in which recurrence will occur. However, recent studies have suggested that urothelial carcinogenesis occurs as a 'field effect' that can involve any number of sites in the bladder mucosa. Accumulating evidence supports the notion that resident urothelial stem cells in the affected field are transformed into cancer stem cells by acquiring genetic alterations that lead to tumor formation through clonal expansion (Cheng et al., 2009).

directly. In the cerebellum, depending upon the signals received, RGCs and stem cells may

Tumours with the highest incidence in humans— medulloblastomas and glioblastomas both originate from abnormal brain stem cells. . Not surprising, both of these tumors are CD133-positive, containing great neuronal differentiation, which makes them prone to be

It is universally accepted that tumors growth as a clonal evolution from a single cell (Nowell, 1976). The "field cancerization theory" was introduced more than fifty years ago by Slaugher et al (1953), when studying the presence of histologically abnormal tissue surrounding carcinomas. In a classic report on oral cancer, Slaughter called "field cancerization" – a process of repeated exposure of a region's entire tissue area to carcinogenic insult (e.g., tobacco and alcohol), which increases the tissue's risk for developing multiple independent premalignant and malignant foci. The field cancerization hypothesis states that multiple cells form independent tumors on one given tissue, since carcinogenic exposure affects multiple cells in the field (Slaughter *et al*., 1953), and predicts that second primary or synchronous tumours arise from independent genetic events (Garcia et al., 1999). The field cancerization theory may be explained by the concept that a given stem cell that acquires genetic alterations may form a "patch", a clonal unit of altered daughter cells. The proliferation of these patch cells forms expanding fields which gradually displace the normal tissue and, by clonal divergence, ultimately leads to the development of one or more tumors within a contiguous field of preneoplastic cells (Garcia et al., 1999). An important clinical implication is that fields often remain after surgery of the primary tumor and may lead to new cancers, designated presently by clinicians as "a second primary tumor" or "local recurrence," depending on the exact site and time interval (Braakhuis et al., 2003; Ryan, 2007). We had previously discussed how mutated clones from mutated stem cells may spread on tissues and that the field cancerization theory implies that the mutated genotype and molecular changes occur before the appearance of histopathological evidence of malignant cells (Garcia et al., 1999). Therefore, this "anomaly" might be due to changes that occur in a "premalignant" neoplastic condition that was histologically identified as "normal". In the clinical aspect, the field cancerization may have an etiologic role in a substantial number of recurrences. For example, a surgical resection margin that includes a genetically altered field can explain the occurrence of scar recurrence. This explanation suggests that molecular profiling of surgical margins will help reduce scar recurrences. Since multiple independent patches of cancer fields may be present in the same organ exposed to the same insults, clean molecular margins may not necessarily prevent recurrences in the residual organ (Dakubo et al., 2007). Similarly to gliomas, tumor recurrence is a major clinical concern for patients with urothelial carcinoma of the urinary bladder. Traditional morphological analysis is of limited utility for identifying cases in which recurrence will occur. However, recent studies have suggested that urothelial carcinogenesis occurs as a 'field effect' that can involve any number of sites in the bladder mucosa. Accumulating evidence supports the notion that resident urothelial stem cells in the affected field are transformed into cancer stem cells by acquiring genetic alterations that

give origin to either ependymoma or medulloblastoma.

**4. Gliomas and the field cancerization concept** 

lead to tumor formation through clonal expansion (Cheng et al., 2009).

diffuse and resistant to treatment (Castilo, 2010).

The available information in regards to the existence of a field phenomenon in gliomas is scant. In malignant gliomas, the high recurrence rates, the characteristically heterogeneous features and frequent diffuse spread within the brain have raised the question of whether malignant gliomas arise monoclonally from a single precursor cell or polyclonally from multiple transformed cells forming confluent clones (Inoue et al., 2008). To address this issue, Kattar et al (1997) have evaluated the clonality of low-grade and malignant gliomas by using polymerase chain reaction (PCR)-based assay for nonrandom X chromosome inactivation using surgical and autopsy material. The same pattern of nonrandom X chromosome inactivation was present in all areas of fifteen of 19 tumors, which were considered as monoclonal, suggesting that low-grade and malignant gliomas are, at least, usually monoclonal tumors, and extensively infiltrating tumors must result from migration of tumor cells.

*Gliomatosis cerebri* may shed some light in this issue. It is a rare condition in which the brain is infiltrated by an exceptionally diffusely growing of malignant glial cell population involving at least 2 lobes, though often more extensive, sometimes even affecting infratentorial regions. Kross et al (2002) have evaluated the existence of field cancerization in this affection, since *gliomatosis cerebri* may initiate as an oligoclonal process or result from collision of different gliomas. It was hypothesized that the presence of an identical set of genetic aberrations throughout the lesion would point to monoclonality of the process. In contrast, the finding of non-identical genetic changes in widely separated regions within the neoplasm would support the concept of collision of different mutated clones. For such, the authors used one autopsy case of *gliomatosis cerebri*, from which tissue samples were randomly taken from 24 locations throughout the brain and used for genetic investigation. With this aim, genome-wide screening for chromosomal aberrations was accomplished by comparative genomic hybridization (CGH). The authors found a wide distribution of particular sets of genetic aberrations, supporting the concept of monoclonal tumor proliferation (Kross et al., 2002). Nevertheless, it has been observed and well documented in one clinical case that on the long term, after initial treatment for *gliomatosis cerebri*, one glioblastoma multiforme has developed, and in a location separate from the initial lesion, suggesting that different clonal origin may had occurred (Inoue et al., 2008). More recently, Chen et al (2010) showed that the capacities for self-renewal and tumour initiation in GBM need not be restricted to a uniform population of stemlike cells.

#### **5. The contribution of studies in animal models: Unifying the cancer stem cells and field cancerization concepts**

Many genetic alterations have been identified in human gliomas, however, establishing unequivocal correlation between these genetic alterations and gliomagenesis requires accurate animal models for these cancers (Dai & Holland, 2001). Indeed, it is useful and necessary to have animal models for CNS tumors studies allowing to be carried out in different stages of tumor growth, especially in early stages, rare to be detected and observed in clinical practice (Bulnes-Sesma, 2006).

Experimental models of gliomagenesis most commonly used alkylating agents such as Nethyl N-nitrosourea (ENU), which has been considered as a suitable model to study malignant changes. These changes were reported to appear firstly as early neoplastic proliferation (ENP) center, which continues in following stages subsequently progressing to ''microtumors'' until a tumor in itself. (Koestner et al., 1971; Naito et al., 1984).

The Role of Stem Cells in the Glioma Growth 195

fate choices (Gilbertson, 2006). It is well-known that stem cells and their microenvironments may influence each other (Scadden, 2006). In fact, Cues within the niche, from cell–cell interactions to diffusible factors, are spatially and temporally coordinated to regulate

In ENU treated rats, we have observed the existence of a close morphological relationship between MT positive cells and blood vessels. What is the relationship between them? It is known that MT is involved in the regulation of the functions of endothelial cells as well as in their protection against cytotoxic agents (Kaji et al., 1993). MT knock-out (MT-KO) mice presented dramatically decreased IL-6-induced angiogenesis caused by cortical freeze injury, suggesting that the MT have major regulatory functions in the angiogenesis process (Penkowa et al., 2000). In fact, human CD133+ Glioma CSCs are capable of producing vascular endothelial growth factor (VEGF) and thus may play an important role in glioma

**7. The concept of stemness, modulation of csc and glioma treatment** 

Understanding the characteristics and function of CSCs has shed light on their roles in glioma progression, including the implications for prognosis and treatment resistance. The original use of the term stemness was derived from a number of articles aimed to look for genes that could be expressed in general stem cell populations. The *stemness hypothesis* states that all stem cells use common mechanisms to regulate self-renewal and multi-lineage potential. This hypothesis has been debated and so far no conclusive evidence for a set of genes expressed in all stem cells. Certainly, identifying genes regulating stem cell properties will greatly improve our understanding of the molecular mechanisms regulating stem cell functions, our ability to manipulate stem cell fate, and the roles of stem cells in cancer (Koeva et al., 2011). Interestingly, overexpression of the transcription factor NANOG in gliomas and its close relationship with the undifferentiated state of glioma cells in vivo and in vitro indicated that NANOG may contribute to the existence of brains CSCs and may be related to tumorigenesis of the cerebrum by maintaining the undifferentiated state of glioma cells (Niu et al., 2011). The new concept stemness is closely related to the observation that there are tissue environment factors that are able to influence or modulate CSCs. The main one is hypoxia, which activates the Hypoxia Induced Factor alpha number 1 (HIFα-1) alpha to enhance the self-renewal activity of CD133-positive cells and to inhibit their differentiation (Soeda et al., 2009). This and other signaling systems drive the transformation of normal stem cells, and perhaps of the bulk of tumor cells to cancer stem cells or to maintain the CSC phenotype (Katoh, 2011). For instance, the oxygen level of 7% has been observed to enhance the stem cell–like phenotype of CD133+ in GBM cells (McCord et al., 2009). Furthermore, it has been observed that human glioblastoma cells from tumor biopsies, which were engrafted intracerebrally into nude rats, that CD133 negative glioma cells were tumorgenic in nude rats, and that CD133 positive cells can be obtained from these tumors. Upon the passing of the cell tumors in vivo, CD133 expression is upregulated, coinciding with the onset of angiogenesis and a shorter patient survival (Wang et al., 2008). Furthermore, the bone morphogenic protein BMP4 effectively reduces proliferation of CD133 positive cells in vitro and the tumor growth in vivo. BMP4 may act as a key inhibitory regulator of cancer initiation and therefore may be used in combined stem cell-based therapy as a non-cytotoxic

proliferation and neurogenesis, ultimately (Riquelme et al., 2008).

angiogenesis (Yao et al.,2008).

therapeutic agent (Altaner, 2008).

By using the experimental model of gliomagenesis induced by the N-ethyl N-nitrosourea, we were able to detect putative tumor stem cells in early oncogenesis, yielding to analyze a field cancerization process and observe a close morphological relationship between metallothionein (MT) positive cells and blood vessels. With this aim, we have developed an experimental model to track putative mutated stem cells, using the ENU experimental model and metallothioneins (MT) immunostaining. MTs are metal binding proteins that take part in the homeostasis of the ions of the metals which are necessary for the proper metabolism of the organism (zinc, copper), disintoxication of metals and protect the tissues from the effects of free radicals, radiation and from mutagens (Thirumoorthy et al., 2007). MT expression is present in a significant portion of especially malignant brain tumors. In astrocytic tumors an acquired enhanced ability to produce MT has been observed as the malignant potential of a tumor increases (Hiura et al., 1998), and MT might be involved in poor response to antineoplastic drugs (Maier et al., 1997). In the murine colonic mucosa, the crypt restricted immunopositivity for MT has been shown to be reliable marker of stem cell mutation that may be induced early after mutagen treatment and that can be assayed in paraffin-fixed tissue sections (Cook et al., 2000). We have observed that 30 days after the treatment of rats with ENU, the main location of the MT positive cells have striking similarity to that of the RG cells and that the frequency of these cells (a) is strongly correlated with the increased appearing of ENP centers and new blood vessels, (b) is augmented at higher levels in long-term observation, i.e., 180 days after the carcinogen administration, (c) is related to a high staining intensity in both nucleus and cytoplasm, and (d) is very similar to the pattern of immunostaining that was observed in the nervous tissue surrounding gliomas, which were originated at an average of 321 days after the ENU administration (Fernandes-da-Silva et al., 2009). The mechanisms and reasons why MT is expressed in the preneoplastic and neoplastic lesions remain to be fully elucidated. It has been hypothesized that mutation-induced MT overexpression may interfere with the function of zinc finger DNA binding transcription factors (Zeng et al., 1991), which have been implicated in transcriptional control of various genes, including TP53, involved in cell proliferation and apoptosis. These MT-mediated effects on gene transcription are thought to confer a selective growth or survival advantage (or both) on the mutated cells (Bruewer, 2002).

#### **6. Glioma, stem cells niche and angiogenesis**

Recently in a review article, Gilbertson & Rich JN (2007) address a number of key questions which remain to be answered: do all cancer stem cells require the support of aberrant niches? Are cancer stem cell niches the primary drivers of tumor development, or are they recruited by pre-formed cancer stem cells? How do cancer stem cells and their niches subvert the tight regulatory conditions that characterize normal stem cell niches?

The stem cells of glioblastoma seem to be dependent on signals from aberrant vascular niches that mimic the normal neural stem cell niche (Gilbertson & Rich, 2007). Stem cells of various tissues are tightly regulated by the immediate microenvironment or stem cell niche (Moore & Lemischka, 2006), which is provided by capillaries in specific locations (Riquelme et al., 2008). This organization places the stem cells in close proximity to endothelial and other vascular cells, facilitating cross-talking among these cell types and affecting stem cell

By using the experimental model of gliomagenesis induced by the N-ethyl N-nitrosourea, we were able to detect putative tumor stem cells in early oncogenesis, yielding to analyze a field cancerization process and observe a close morphological relationship between metallothionein (MT) positive cells and blood vessels. With this aim, we have developed an experimental model to track putative mutated stem cells, using the ENU experimental model and metallothioneins (MT) immunostaining. MTs are metal binding proteins that take part in the homeostasis of the ions of the metals which are necessary for the proper metabolism of the organism (zinc, copper), disintoxication of metals and protect the tissues from the effects of free radicals, radiation and from mutagens (Thirumoorthy et al., 2007). MT expression is present in a significant portion of especially malignant brain tumors. In astrocytic tumors an acquired enhanced ability to produce MT has been observed as the malignant potential of a tumor increases (Hiura et al., 1998), and MT might be involved in poor response to antineoplastic drugs (Maier et al., 1997). In the murine colonic mucosa, the crypt restricted immunopositivity for MT has been shown to be reliable marker of stem cell mutation that may be induced early after mutagen treatment and that can be assayed in paraffin-fixed tissue sections (Cook et al., 2000). We have observed that 30 days after the treatment of rats with ENU, the main location of the MT positive cells have striking similarity to that of the RG cells and that the frequency of these cells (a) is strongly correlated with the increased appearing of ENP centers and new blood vessels, (b) is augmented at higher levels in long-term observation, i.e., 180 days after the carcinogen administration, (c) is related to a high staining intensity in both nucleus and cytoplasm, and (d) is very similar to the pattern of immunostaining that was observed in the nervous tissue surrounding gliomas, which were originated at an average of 321 days after the ENU administration (Fernandes-da-Silva et al., 2009). The mechanisms and reasons why MT is expressed in the preneoplastic and neoplastic lesions remain to be fully elucidated. It has been hypothesized that mutation-induced MT overexpression may interfere with the function of zinc finger DNA binding transcription factors (Zeng et al., 1991), which have been implicated in transcriptional control of various genes, including TP53, involved in cell proliferation and apoptosis. These MT-mediated effects on gene transcription are thought to confer a selective growth or survival advantage (or both) on the mutated cells (Bruewer,

2002).

**6. Glioma, stem cells niche and angiogenesis** 

Recently in a review article, Gilbertson & Rich JN (2007) address a number of key questions which remain to be answered: do all cancer stem cells require the support of aberrant niches? Are cancer stem cell niches the primary drivers of tumor development, or are they recruited by pre-formed cancer stem cells? How do cancer stem cells and their niches

The stem cells of glioblastoma seem to be dependent on signals from aberrant vascular niches that mimic the normal neural stem cell niche (Gilbertson & Rich, 2007). Stem cells of various tissues are tightly regulated by the immediate microenvironment or stem cell niche (Moore & Lemischka, 2006), which is provided by capillaries in specific locations (Riquelme et al., 2008). This organization places the stem cells in close proximity to endothelial and other vascular cells, facilitating cross-talking among these cell types and affecting stem cell

subvert the tight regulatory conditions that characterize normal stem cell niches?

fate choices (Gilbertson, 2006). It is well-known that stem cells and their microenvironments may influence each other (Scadden, 2006). In fact, Cues within the niche, from cell–cell interactions to diffusible factors, are spatially and temporally coordinated to regulate proliferation and neurogenesis, ultimately (Riquelme et al., 2008).

In ENU treated rats, we have observed the existence of a close morphological relationship between MT positive cells and blood vessels. What is the relationship between them? It is known that MT is involved in the regulation of the functions of endothelial cells as well as in their protection against cytotoxic agents (Kaji et al., 1993). MT knock-out (MT-KO) mice presented dramatically decreased IL-6-induced angiogenesis caused by cortical freeze injury, suggesting that the MT have major regulatory functions in the angiogenesis process (Penkowa et al., 2000). In fact, human CD133+ Glioma CSCs are capable of producing vascular endothelial growth factor (VEGF) and thus may play an important role in glioma angiogenesis (Yao et al.,2008).

#### **7. The concept of stemness, modulation of csc and glioma treatment**

Understanding the characteristics and function of CSCs has shed light on their roles in glioma progression, including the implications for prognosis and treatment resistance. The original use of the term stemness was derived from a number of articles aimed to look for genes that could be expressed in general stem cell populations. The *stemness hypothesis* states that all stem cells use common mechanisms to regulate self-renewal and multi-lineage potential. This hypothesis has been debated and so far no conclusive evidence for a set of genes expressed in all stem cells. Certainly, identifying genes regulating stem cell properties will greatly improve our understanding of the molecular mechanisms regulating stem cell functions, our ability to manipulate stem cell fate, and the roles of stem cells in cancer (Koeva et al., 2011). Interestingly, overexpression of the transcription factor NANOG in gliomas and its close relationship with the undifferentiated state of glioma cells in vivo and in vitro indicated that NANOG may contribute to the existence of brains CSCs and may be related to tumorigenesis of the cerebrum by maintaining the undifferentiated state of glioma cells (Niu et al., 2011). The new concept stemness is closely related to the observation that there are tissue environment factors that are able to influence or modulate CSCs. The main one is hypoxia, which activates the Hypoxia Induced Factor alpha number 1 (HIFα-1) alpha to enhance the self-renewal activity of CD133-positive cells and to inhibit their differentiation (Soeda et al., 2009). This and other signaling systems drive the transformation of normal stem cells, and perhaps of the bulk of tumor cells to cancer stem cells or to maintain the CSC phenotype (Katoh, 2011). For instance, the oxygen level of 7% has been observed to enhance the stem cell–like phenotype of CD133+ in GBM cells (McCord et al., 2009). Furthermore, it has been observed that human glioblastoma cells from tumor biopsies, which were engrafted intracerebrally into nude rats, that CD133 negative glioma cells were tumorgenic in nude rats, and that CD133 positive cells can be obtained from these tumors. Upon the passing of the cell tumors in vivo, CD133 expression is upregulated, coinciding with the onset of angiogenesis and a shorter patient survival (Wang et al., 2008). Furthermore, the bone morphogenic protein BMP4 effectively reduces proliferation of CD133 positive cells in vitro

and the tumor growth in vivo. BMP4 may act as a key inhibitory regulator of cancer initiation and therefore may be used in combined stem cell-based therapy as a non-cytotoxic therapeutic agent (Altaner, 2008).

The Role of Stem Cells in the Glioma Growth 197

Dai, C. & Holland, E.C. 2001. Glioma models. *Biochim Biophys Acta*, Vol.1551, No.1, pp.M19-

Das, S.; Srikanth, M.; Kessler, J.A. 2008. Cancer stem cells and glioma. *Nat Clin Pract* 

Fernandes da Silva, J.C.; Kanen, V.C.; Turatti, A.; Ribeiro-Silva, A.; Herrero, C.F.S.; Garcia,

Frosina, G. 2011. Frontiers in targeting glioma stem cells. *Eur J Cancer*, Vol.47, No.4, Mar,

Garcia SB, Park HS, Novelli M, Wright NA. 199. Field cancerization, clonality, and epithelial

Gilbertson, R.J. 2006. Brain tumors provide new clues to the source of cancer stem cells: does

Gilbertson, R.J. & Rich, J.N. 2007. Making a tumour's bed: glioblastoma stem cells and the

Hiura, T.; Khalid, H.; Yamashita, H.; Tokunaga, Y.; Yasunaga, A.; Shibata, S. 1998.

Holmberg J, He X, Peredo I, Orrego A, Hesselager G, Ericsson C, Hovatta O, Oba-Shinjo SM,

Huang, Q.; Zhang, Q.B.; Dong, J.; Wu, Y.Y.; Shen, Y.T.; Zhao, Y.D.; Zhu, Y.D.; Diao, Y.;

Hulleman, E. & Helin, K. Molecular mechanisms in gliomagenesis. 2005. *Adv Cancer* 

Inoue, A.; Takahashi, H.; Harada, H.; Kohno, S.; Ohue, S.; Kobayashi, K.; Yano, H.; Tanaka,

Inoue, T.; Kanamori, M.; Sonoda, Y.; Watanabe, M.; Sasajima, T.; Kamisato, N.; Kumabe, T.;

Kaji, T.; Yamamoto, C.; Tsubaki, S.; Sakamoto, M.; Sato, M.; Kozuka, H. 1993.

cultured vascular endothelial cells. *Life Sci*, Vol.53, No.15, pp.1185-1191.

maintained long-term in vitro*, BMC Cancer*, Vol.8, Oct, pp.304.

oncology recapitulate ontogeny? *Cell Cycle*, Vol.5, pp.135–137.

vascular niche. *Nat Rev Cancer*, Vol.7, No.10, pp.733-736.

S.B. 2009. Overexpression of metallothioneins, stem cell niches and field cancerization in experimental gliomagenesis. *COLUNA/COLUMNA*, Vol.8, No.4,

stem cells: the spread of mutated clones in epithelial sheets. *J Pathol*, Vol.187, pp.61-

Immunohistochemical analysis of metallothionein in astrocytic tumors in relation to tumor grade, proliferative potential, and survival. *Cancer*, Vol.83, No.11, pp.2361-

Marie SK, Nistér M, Muhr J. 2011. Activation of neural and pluripotent stem cell signatures correlates with increased malignancy in human glioma. *PLoS One*, Vol.6,

Wang, A.D.; Lan, Q. 2008. Glioma stem cells are more aggressive in recurrent tumors with malignant progression than in the primary tumor, and both can be

Ohnishi, T. 2010. Cancer stem-like cells of glioblastoma characteristically express MMP-13 and display highly invasive activity. *Int Journal of Oncology*, Vol.37, No.5,

Tominaga, T. 2008. Glioblastoma multiforme developing separately from the initial lesion 9 years after successful treatment for gliomatosis cerebri: a case report. *No* 

Metallothionein induction by cadmium, cytokines, thrombin and endothelin-1 in

27.

pp.428-433.

pp.496-507.

81.

2369.

No.3, pp.e18454.

*Research*, Vol.94, pp.1-27.

*Shinkei Geka*, Vol.36, No.8, Aug, pp.709-715.

Nov, pp.1121-1131.

*Neurology*, Vol.4, pp.427–435.

If one accepts that there is a subpopulation of cancer cells with stem cell properties, which is responsible for tumor maintenance and progression, and may contribute to the resistance to anticancer treatments, it is very reasonable to deduce that compounds that target cancer stem-like cells could be effective to impair or even to destroy a neoplasm and nas important therapeutic implications. Various compounds have been investigated as putative influencers of stemness and malignancies in glioma stem-like cells, leading the proposal that stem cell regulatory factors may provide significant targets for therapeutic strategies (Holmberg et al., 2011). Ongoing work aims the identification of unique pathways governing self-renewal of these putative stem cells and their validation as ultimate therapeutic targets (Panagiotakos & Tabar, 2007). Additionally, it is possible to conceive that epigenetic-based drugs that modulate gene expression in CSC possibly constitute a promising alternative resource for target therapy in the treatment of these, thus far, incurable malignancy.

#### **8. References**

Altaner, C. 2008. Glioblastoma and stem cells. *Neoplasma,* Vol.55, No.5, pp.369-374.


If one accepts that there is a subpopulation of cancer cells with stem cell properties, which is responsible for tumor maintenance and progression, and may contribute to the resistance to anticancer treatments, it is very reasonable to deduce that compounds that target cancer stem-like cells could be effective to impair or even to destroy a neoplasm and nas important therapeutic implications. Various compounds have been investigated as putative influencers of stemness and malignancies in glioma stem-like cells, leading the proposal that stem cell regulatory factors may provide significant targets for therapeutic strategies (Holmberg et al., 2011). Ongoing work aims the identification of unique pathways governing self-renewal of these putative stem cells and their validation as ultimate therapeutic targets (Panagiotakos & Tabar, 2007). Additionally, it is possible to conceive that epigenetic-based drugs that modulate gene expression in CSC possibly constitute a promising alternative resource for

target therapy in the treatment of these, thus far, incurable malignancy.

implications. *Cancer Res*, Vol.63, No.8, pp.1727-1730.

*Journal Neuroradiology*, No.3, Mar, pp.389-390.

*Oncology*, Vol.5, No.6, pp. 843-857.

Vol.75 Suppl 2, pp.2-11.

colorectal carcinoma. *World Journal of Surgery*, 26,6,726-731.

glioblastoma. *Cancer Cell,* Vol.17, No.4, April, pp.362-375.

stem cell mutation. *J Pathology*, Vol.191, No.3, pp.306-312.

Altaner, C. 2008. Glioblastoma and stem cells. *Neoplasma,* Vol.55, No.5, pp.369-374.

Braakhuis BJM, Tabor MP, Kummer JA, Leemans CR, Brakenhoff RH. 2003. A genetic

Bruewer, M.; Schmid, K.W.; Krieglstein, C.F.; Senninger, N.; Schuermann, G. 2002.

Bulnes-Sesma S, Ullibarri-Ortiz de Zárate N, Lafuente-Sánchez JV. 2006. Tumour induction

Castillo, M. 2010. Stem cells, radial glial cells, and a unified origin of brain tumors. *American* 

Chen, R.; Nishimura, M.C.; Bumbaca, S.M.; Kharbanda, S.; Forrest, W.F.; Kasman, I.M.;

Cheng L, Zhang S, Davidson DD, MacLennan GT, Koch MO, Montironi R, Lopez-Beltran A.

Clément, V.; Dutoit, V.; Marino, D.; Dietrich, P.Y.; Radovanovic, I. 2009. Limits of CD133 as a marker of glioma self-renewing cells. *Int J Cancer*. Jul 1;125(1):244-8. Collins, V.P. Brain tumours: classification and genes. 2004. *J Neurol Neurosurg Psychiatry*,

Cook, A.H.; Williams, D.; Thomas, A.G. 2000. Crypt-restricted metallothionein

Dakubo, G.D.; Jakupciak, J.P.; Birch-Machin, M.A.; Parr, R.L.; 2007. Clinical implications and utility of field cancerization., *Cancer Cell International, Vol.*7, pp.2. ???

explanation of Slaughter's concept of field cancerization: evidence and clinical

Metallothionein: early marker in the carcinogenesis of ulcerative colitis-associated

by ethylnitrosourea in the central nervous system. *Rev Neurol*, Vol.43, No12, pp.733-

Greve, J.M.; Soriano, R.H.; Gilmour, L.L.; Rivers, C.S.; Modrusan, Z.; Nacu, S.; Guerrero, S.; Edgar, K.A.; Wallin, J.J.; Lamszus, K.; Westphal, M.; Heim, S.; James, C.D.; VandenBerg, S.R.; Costello, J.F.; Moorefield, S.; Cowdrey, C.J.; Prados, M.; Phillips, H.S. 2010. A hierarchy of self-renewing tumor-initiating cell types in

2009. Molecular determinants of tumor recurrence in the urinary bladder. *Future* 

immunopositivity in murine colon: validation of a model for studies of somatic

**8. References** 

738.


The Role of Stem Cells in the Glioma Growth 199

Penkowa, M.; Carrasco, J.; Giralt, M.; Molinero, A.; Hernández, J.; Campbell, I.L. 2000.

Pérez Castillo, A.; Aguilar-Morante, D.; Morales-García, J.A.; Dorado, J. 2008. Cancer stem cells and brain tumors. *Clin Transl Oncol*, Vol.10, No.5, May, pp.262-267. Reya, T.; Morrison, S.J.; Clarke, M.F.; Weissman, I.L. 2001. Stem cells, cancer, and cancer

Rich, J.N. & Eyler, C.E. 2008. Cancer stem cells in brain tumor biology. *Cold Spring Harb* 

Riquelme, P.A.; Drapeau, E.; Doetsch F. 2008. Brain micro-ecologies: neural stem cell niches

Ryan, L.P. 2007. Clinical implications and utility of field cancerization. *Cancer Cell* 

Sanchez-Martin, M. 2008. Brain tumor stem cells: implications for cancer therapy and regenerative medicine. *Curr Stem Cell Res Ther*, Vol.3, No.3, pp.197-207. Scadden, D. T. 2006. The stem-cell niche as an entity of action. *Nature*, Vol.441, pp.1075–

Singh, S.K.; Clarke, I.D.; Hide, T.; Dirks PB. 2004. Cancer stem cells in nervous system

Slaughter, D.P.; Southwick, H.W.; Smejkal, W. 1953. Field cancerization in oral stratified

Soeda, A.; Park, M.; Lee, D.; Mintz, A, Androutsellis-Theotokis A, McKay RD, Engh J,

Thirumoorthy, N.M.; Kumar, K.T.; Shyam, S.A.; Panayappan, L.; Chatterjee, M. 2007.

Vescovi, A.L.; Galli, R.; Reynolds, B.A. 2006. Brain tumor stem cells. *Nat Rev Cancer*, Vol.6,

Wang, J.; Sakariassen, P.O.; Tsinkalovsky, O.; Immervoll, H.; Bøe, S.O.; Svendsen, A.;

Weiner, L.P. 2008. Definitions and criteria for stem cells. *Methods Mol Biol*. Vol. 438,

Yan, X.; Ma, L.; Yi, D.; Yoon, J.G.; Diercks, A.; Foltz, G.; Price, N.D.; Hood, L.E.; Tian, Q.

CD133 positive cells. *Int J Cancer*, Vol.122, No.4, Feb, pp.761-768.

squamous epithelium; clinical implications of multicentric origin. *Cancer*. Vol.6,

Iwama T, Kunisada T, Kassam AB, Pollack IF, Park DM. 2009. Hypoxia promotes expansion of the CD133-positive glioma stem cells through activation of HIF-

Metallothionein: an overview. *World J Gastroenterology*, Vol.13, No.7, pp.993-

Prestegarden, L.; Røsland, G.; Thorsen, F.; Stuhr, L.; Molven, A.; Bjerkvig, R.; Enger, P.Ø. 2008. CD133 negative glioma cells form tumors in nude rats and give rise to

2011. A CD133-related gene expression signature identifies an aggressive glioblastoma subtype with excessive mutations. *Proc Natl Acad Science*, Vol.108,

tumors. *Oncogene*, Vol.23, No.43, Sep., pp.7267-7273.

No.5, Sep., pp.963-968. (PMID:13094644)

1alpha. *Oncogene*, Vol.28, No.45, pp.3949-3959.

in the adult mammalian brain. Philos Trans R Soc Lond B Biol Science, Vol.363,

No.8; pp.1174-1189.

No.1489, pp.123-137.

*International*, 7:2.

1079.

996.

pp.425–436

pp.3-8.

No.4, Jan, pp.1591-1596.

stem cells. *Nature.*;414:105–11.

*Symp Quant Biol*, Vol.73, pp.411-420.

Altered central nervous system cytokine-growth factor expression profiles and angiogenesis in metallothionein-I+II deficient mice. *J Cereb Blood Flow Metab*. Vol.20,


Katoh, M. 2011. Network of WNT and other regulatory signaling cascades in

Kattar, M.M.; Kupsky, W.J.; Shimoyama, R.K.; Vo, T.D.; Olson, M.W.; Bargar, G.R.; Sarkar,

Koestner, A.; Swenberg, J.A.; Wechsler, W. 1971. Transplacental production with

Koeva, M.; Forsberg, E.C.; Stuart, J.M. 2011 . Computational integration of homolog and

Kros, J.M.; Zheng, P.; Dinjens, W.N.; Alers, M; Janeke, C. 2002. Genetic Aberrations in

Li, Z.; Bao, S.; Wu, Q.; Wang, H.; Eyler, C.; Sathornsumetee, S.; Shi, Q.; Cao, Y.; Lathia, J.;

Li, Z. & Rich JN. 2010. Hypoxia and hypoxia inducible factors in cancer stem cell

Maier, H.; Jones, C.; Jasani, B.; Ofner, D.; Zelger, B.; Schmid, K.W.; Budka, H. 1997.

Malatesta, P.; Hack, M.A.; Hartfuss, E.; Kettenmann, H.; Klinkert, W.; Kirchhoff, F.; Götz, M.

McCord, A.M.; Jamal, M.; Shankavaram, U.T.; Lang, F.F.; Camphausen, K.; Tofilon, P.J. 2009.

Naito, M.; Naito, Y.; Ito, A.; Watanabe, H.; Kawashima, K. 1984. Spinal cord tumors induced

Niu, C.S.; Li, D.X.; Liu, Y.H.; Fu, X.M.; Tang, S.F.; Li, J. Expression of NANOG in human

Nowell, P.C. 1976. The clonal evolution of tumor cell populations. *Science*; Vol.194, pp. 23–

Panagiotakos, G. & Tabar, V. 2007. Brain tumor stem cells. *Current Neurology and* 

Park, D.M.; Rich, J.N. 2009. Biology of glioma cancer stem cells. *Mol Cells*, Vol.28, No.1, Jul,

May 13, doi: 10.3892/or.2011.1308. [Epub ahead of print]

Pathology, Vol.63, No.1, Apr, pp.37-56. ( PMID:4323476)

*& Experimental Neurology*, Vol.61, No.9, pp.806–814.

maintenance. *Curr Top Microbiol Immunol*, Vol.345, pp.21-30.

No.2, pp.160-170.

Vol.6, No.4, pp.e18968.

pp.501-513.

No.6, pp.599-604.

1885.

28.

pp.7-12.

Vol.37, No.5, pp.751-764.

*Cancer Institute*, Vol.2, No.3, pp.715-724.

*Neuroscience Reports*, Vol.7, No.3, pp.215-220.

1179.

pluripotent stem cells and cancer stem cells. *Curr Pharm Biotechnol*. Vol.12,

F.H. 1997. Clonal analysis of gliomas. *Human Pathology*, Vol.28, No.10, Oct, pp.1166-

ethylnitrosourea of neoplasms of the nervous system in Sprague-Dawley rats. Am J

pathway gene module expression reveals general stemness signatures. *PLoS One*.

Gliomatosis Cerebri Support Monoclonal Tumorigenesis. *Journal of Neuropathology* 

McLendon, R.E.; Hjelmeland, A.B.; Rich, J.N. 2009. Hypoxia-inducible factors regulate tumorigenic capacity of glioma stem cells. *Cancer Cell*, Vol.15, No.6, Jun,

Metallothionein overexpression in human brain tumours. *Acta Neuropathol*, Vol.94,

2003. Neuronal or glial progeny: regional differences in radial glia fate. *Neuron*,

Physiologic oxygen concentration enhances the stem-like properties of CD133+ human glioblastoma cells in vitro. *Mol Cancer Res*, Vol.7, No.4, Apr, pp.489-497. Moore, K. A. & Lemischka, I. R. 2006. Stem cells and their niches. *Science*, Vol.311, pp.1880–

by N-ethyl-N-nitrosourea in rats: presence of spinal subpial target cells. *J Natl* 

gliomas and its relationship with undifferentiated glioma cells. 2011. *Oncol Rep*ort,


**10** 

*USA* 

Jeffrey P. Greenfield et al.\* *Weill Cornell Medical College* 

**Bone Marrow-Derived Cells Support Malignant** 

Gliomas, the most common primary brain tumors, exist as a continuum between low-grade and high-grade states. Low grade gliomas are generally found in children and young adults. These tumors are characterized by well-differentiated cellularity which is mildly pleomorphic. These tumors lack mitotic figures and neovascularization and do not enhance on MRI. The average survival of patients after diagnosis is 7-10 years; the morbidity associated with these lesions is largely dependent on progression of these lesions to a higher grade state. High-grade gliomas, conversely, which exist on the other end of the glial neoplasm spectrum, are extremely malignant with poorly differentiated cells that are highly pleomorphic and display numerous mitotic figures. These tumors contain significant vascular proliferation, hemorrhage and necrosis. High grade gliomas enhance brightly on contrast MRI and often exhibit widespread invasion throughout the brain. Prognosis is poor for high grade gliomas, with a median survival of 18 months even with aggressive therapies. One of the key events in the transition from the low-grade to high-grade state has been referred to as the angiogenic switch. This is defined as the period during which the tumor undergoes a transition to an environment capable of rapid blood vessel formation supporting subsequent exponential tumor growth. It is theorized that in the low-grade state, tumor growth may be limited, at least in part, by a lack of blood supply limiting the tumor to linear growth. Once the tumor acquires the ability to recruit or form new blood vessels through this angiogenic switch, exponential growth may occur, which results in rapid clinical progression. It has been well-described in the literature that bone marrow-derived cells (BMDC) participate in the progression of cancer. BMDCs in the local tumor microenvironment have been proposed to be capable of breaking down normal structures thereby promoting vasculogenesis and invasiveness. This, in turn, provides an environment capable of sustaining and promoting tumor growth. The role of BMDC in metastatic disease has been well-documented and recent data suggests that BMDC participate in the growth and progression of brain tumors as well. This chapter will explore the role of BMDC in the transition from low-grade to high-grade gliomas particularly with respect to the angiogenic

\* William S. Cobb, Caitlin E. Hoffman, Xueying Chen, Prajwal Rajappa,

Chioma Ihunnah, Yujie Huang and David Lyden

*Weill Cornell Medical College* 

**1. Introduction** 

*USA* 

**Transformation of Low-Grade Glioma** 


## **Bone Marrow-Derived Cells Support Malignant Transformation of Low-Grade Glioma**

Jeffrey P. Greenfield et al.\* *Weill Cornell Medical College USA* 

#### **1. Introduction**

200 Glioma – Exploring Its Biology and Practical Relevance

Yao, X.H.; Ping, Y.F.; Chen, J.H.; Xu, C.P.; Chen, D.L.; Zhang, R.; Wang, J.M.; Bian, X.W.

Zeng, J.; Heuchel, R.; Schaffner, W.; Kägi, J.H. 1991. Thionein (apometallothionein) can

No.4, Aug, pp.369-376.

factor Sp1. *FEBS Lett*, Vol.279, No.2, pp.310-312.

2008. Glioblastoma stem cells produce vascular endothelial growth factor by activation of a G-protein coupled formylpeptide receptor FPR. *J Pathology*, Vol.215,

modulate DNA binding and transcription activation by zinc finger containing

Gliomas, the most common primary brain tumors, exist as a continuum between low-grade and high-grade states. Low grade gliomas are generally found in children and young adults. These tumors are characterized by well-differentiated cellularity which is mildly pleomorphic. These tumors lack mitotic figures and neovascularization and do not enhance on MRI. The average survival of patients after diagnosis is 7-10 years; the morbidity associated with these lesions is largely dependent on progression of these lesions to a higher grade state. High-grade gliomas, conversely, which exist on the other end of the glial neoplasm spectrum, are extremely malignant with poorly differentiated cells that are highly pleomorphic and display numerous mitotic figures. These tumors contain significant vascular proliferation, hemorrhage and necrosis. High grade gliomas enhance brightly on contrast MRI and often exhibit widespread invasion throughout the brain. Prognosis is poor for high grade gliomas, with a median survival of 18 months even with aggressive therapies. One of the key events in the transition from the low-grade to high-grade state has been referred to as the angiogenic switch. This is defined as the period during which the tumor undergoes a transition to an environment capable of rapid blood vessel formation supporting subsequent exponential tumor growth. It is theorized that in the low-grade state, tumor growth may be limited, at least in part, by a lack of blood supply limiting the tumor to linear growth. Once the tumor acquires the ability to recruit or form new blood vessels through this angiogenic switch, exponential growth may occur, which results in rapid clinical progression. It has been well-described in the literature that bone marrow-derived cells (BMDC) participate in the progression of cancer. BMDCs in the local tumor microenvironment have been proposed to be capable of breaking down normal structures thereby promoting vasculogenesis and invasiveness. This, in turn, provides an environment capable of sustaining and promoting tumor growth. The role of BMDC in metastatic disease has been well-documented and recent data suggests that BMDC participate in the growth and progression of brain tumors as well. This chapter will explore the role of BMDC in the transition from low-grade to high-grade gliomas particularly with respect to the angiogenic

<sup>\*</sup> William S. Cobb, Caitlin E. Hoffman, Xueying Chen, Prajwal Rajappa,

Chioma Ihunnah, Yujie Huang and David Lyden

*Weill Cornell Medical College* 

*USA* 

Bone Marrow-Derived Cells Support Malignant Transformation of Low-Grade Glioma 203

between 2,700 and 4,700 cases per year, comprising approximately 30% of all malignant gliomas (Schiff et al., 2007, Wessels et al., 2003). These tumors are most common in Caucasian males, and typically present in the second to fourth decades (Schomas et al., 2009, Wessels et al., 2003). Patients greater than 60 years of age carry a poorer prognosis with generally lower Karnofsky scores and larger tumor burden at diagnosis. In adults, the most common presenting symptom is seizure followed by incidental findings on imaging. Less common presentations include trauma trauma, sinus pathology, and pituitary disorder (Wessels et al., 2003). Thirty percent of patients present with neurological deficit, and only 10% present with symptoms of raised intracranial pressure ICP. Speech and language deficits have been reported in 10% of patients (Prabhu et al., 2010), however, focal deficits

In adults, LGGs are generally hemispheric, supratentorial, and typically occur in the frontal and temporal lobes. They may involve eloquent cortex, which limits the capacity for gross total resection due to significant risk of morbidity (Prabhu et al., 2010, Stieber, 2001). LGG are hypointense on T1 weighted magnetic resonance imaging (MRI), hyperintense on FLAIR and T2 sequences, and enhance in 30% of cases. There is often associated vasogenic edema

LGG are typically sporadic tumors, although they can occur in association with Li Fraumeni syndrome and Neurofibromatosis Types 1 and 2 (Prabhu et al., 2010, Wessels et al., 2003). Additional risk factors include previous irradiation and exposure to industrial chemicals (Prabhu et al., 2010, Wessels et al., 2003). Allergy has been reported to lower the risk for LGG, suggesting a possible role for immune surveillance in tumor pathogenesis (Prabhu et al., 2010). Survival is highly variable for LGG as median overall survival (OS) is reported to range from 3 to 40 years. Median progression free survival (PFS) is only 50% at 5 years and 17% at 15 years (Bauman et al., 1999, Berger et al., 1994, Jaeckle et al., Stieber, 2001). Median time to progression is 7.2 years (Schomas et al., 2009). In adults, the overall malignant transformation rate ranges from 35-89% with 74% in primary astrocytoma, 70% with mixed tumors, and 45% with primarily oligodendroglial histology (Jaeckle et al., 2010). Importantly, 50% of low risk adults, defined as patients less than 40 years of age with gross total resection (GTR), underwent transformation within 5 years

Multiple studies have found that age greater than 40 years, extent of resection, tumor diameter greater than 6cm, tumor crossing midline, neurological deficit at diagnosis, and astrocytic histology are risk factors for poor prognosis in LGG (Bauman et al., 1999, E. G. Shaw et al., 2008, Stieber, 2001, Jaeckle et al., 2010, Schiff et al., 2007, E. G. Shaw & Wisoff, 2003). The NCCTG found that astrocytomas carry a worse prognosis than oligodendroglioma. Other retrospective reports corroborate these finding and further specify gemistocytic astrocytoma as carrying a worse prognosis (Jaeckle et al., 2010, Schomas et al., 2009, Stieber, 2001, Wessels et al., 2003, E. G. Shaw et al., 2008). Contrast enhancement, Karnofsky score, mitotic activity, and genetics have also been identified as risk factors for progression (Schiff et al., 2007, Schomas et al., 2009, E. G. Shaw & Wisoff, 2003, Stieber, 2001). Additionally, a Ki67-MIB1 index greater than 4% is associated with a

The presentation and prognosis of LGG in children differs significantly from that in adults. Overall survival and rate of malignant transformation is significantly different in the

are less common (Schomas et al., 2009, Wessels et al., 2003).

(Prabhu et al., 2010, Wessels et al., 2003).

(Jaeckle et al., 2010, Schiff et al., 2007).

more rapid rate of transformation.

**3. Malignant transformation of pediatric low-grade glioma** 

switch. The possibility of this pathway as a potential therapeutic target will also be reviewed.

#### **2. Low-grade glioma transformation in adults**

Low-grade gliomas (LGG) are a heterogeneously diverse group of tumors with a generally benign histology and an associated variable outcome. This unpredictable course relies, in part, on the potential for malignant transformation to a higher grade. These tumors present a unique therapeutic challenge as they are typically associated with minimal symptoms and benign radiographic appearance. Initially, the majority of LGGs run an indolent clinical course but often ultimately progress into aggressive tumors with a poor prognosis. As a result, significant controversy exists as to appropriate treatment protocols for this disease. The natural history of LGG, and the risk factors for progression, have been one focus of glioma research due to the potential impact on treatment strategy. Many recent studies have helped clarify treatment recommendations including extent of resection, timing and efficacy of radiation therapy, and response to chemotherapy. Significant debate remains, however, regarding standardization of treatment for low-grade glioma given the tremendous diversity in tumor histology, biology, and outcome. While observation of low-grade gliomas was previously considered a valid treatment option to avoid the morbidity of surgery, chemotherapy, and radiation, early intervention has gradually become standard of care as the impact and incidence of malignant progression has become fully realized. Subjecting patients to the morbidity of aggressive treatment in an unpredictable tumor with variable outcome remains controversial, however. Currently, significant effort is focused on identification of risk factors and tumor characteristics that lead to progression. Better appreciation for the molecular and cellular mechanisms of malignant transformation carries the potential to create novel treatment regimens with less morbidity, thereby alleviating the use of radiation and chemotherapy which present significant toxicity to both children and adults. A review of the characteristics of low-grade gliomas, current treatment strategies, their transformation potential, and current efforts to define novel pathways involved in malignant transformation follows.

The term LGG includes World Health Organization Grade I and Grade II tumors, which are typically associated with indolent tumor growth and significantly better prognosis compared to high grade gliomas. Grade I gliomas include pilocytic astrocytoma, desmoplastic neuroectodermal tumors, subependymoma, ganglioglioma, myxopapillary ependymoma, and desmoplastic infantile tumors, which represent a spectrum of typically benign lesions. Within this class, pilocytic histology is the most common (Stieber, 2001). These pilocytic tumors are well-circumscribed, non-infiltrative, and do not generally transform to more malignant, higher grade lesions. While malignant transformation has been reported in WHOI tumors, the primary risk for malignant degeneration exists in Grade II tumors including low-grade or fibrillary astrocytoma, oligodendroglioma, or mixed oligoastrocytoma. Ependymoma, ganglioglioma, pleomorphic xanthoastrocytoma, and choroid gliomas of the third ventricle are also considered grade II. Fibrillary astrocytomas, which comprise the majority of grade II lesions (Stieber, 2001), have garnered significant attention due to the significant morbidity and mortality of patients with this diagnosis.

While WHO I gliomas are typically well-circumscribed tumors with benign histology, WHO II gliomas are diffuse, infiltrative and have malignant potential (Stieber, 2001). Both classes, however, are associated with slow tumor growth. The incidence of LGG is reported to be

switch. The possibility of this pathway as a potential therapeutic target will also be

Low-grade gliomas (LGG) are a heterogeneously diverse group of tumors with a generally benign histology and an associated variable outcome. This unpredictable course relies, in part, on the potential for malignant transformation to a higher grade. These tumors present a unique therapeutic challenge as they are typically associated with minimal symptoms and benign radiographic appearance. Initially, the majority of LGGs run an indolent clinical course but often ultimately progress into aggressive tumors with a poor prognosis. As a result, significant controversy exists as to appropriate treatment protocols for this disease. The natural history of LGG, and the risk factors for progression, have been one focus of glioma research due to the potential impact on treatment strategy. Many recent studies have helped clarify treatment recommendations including extent of resection, timing and efficacy of radiation therapy, and response to chemotherapy. Significant debate remains, however, regarding standardization of treatment for low-grade glioma given the tremendous diversity in tumor histology, biology, and outcome. While observation of low-grade gliomas was previously considered a valid treatment option to avoid the morbidity of surgery, chemotherapy, and radiation, early intervention has gradually become standard of care as the impact and incidence of malignant progression has become fully realized. Subjecting patients to the morbidity of aggressive treatment in an unpredictable tumor with variable outcome remains controversial, however. Currently, significant effort is focused on identification of risk factors and tumor characteristics that lead to progression. Better appreciation for the molecular and cellular mechanisms of malignant transformation carries the potential to create novel treatment regimens with less morbidity, thereby alleviating the use of radiation and chemotherapy which present significant toxicity to both children and adults. A review of the characteristics of low-grade gliomas, current treatment strategies, their transformation potential, and current efforts to define novel pathways involved in

The term LGG includes World Health Organization Grade I and Grade II tumors, which are typically associated with indolent tumor growth and significantly better prognosis compared to high grade gliomas. Grade I gliomas include pilocytic astrocytoma, desmoplastic neuroectodermal tumors, subependymoma, ganglioglioma, myxopapillary ependymoma, and desmoplastic infantile tumors, which represent a spectrum of typically benign lesions. Within this class, pilocytic histology is the most common (Stieber, 2001). These pilocytic tumors are well-circumscribed, non-infiltrative, and do not generally transform to more malignant, higher grade lesions. While malignant transformation has been reported in WHOI tumors, the primary risk for malignant degeneration exists in Grade II tumors including low-grade or fibrillary astrocytoma, oligodendroglioma, or mixed oligoastrocytoma. Ependymoma, ganglioglioma, pleomorphic xanthoastrocytoma, and choroid gliomas of the third ventricle are also considered grade II. Fibrillary astrocytomas, which comprise the majority of grade II lesions (Stieber, 2001), have garnered significant attention

While WHO I gliomas are typically well-circumscribed tumors with benign histology, WHO II gliomas are diffuse, infiltrative and have malignant potential (Stieber, 2001). Both classes, however, are associated with slow tumor growth. The incidence of LGG is reported to be

due to the significant morbidity and mortality of patients with this diagnosis.

**2. Low-grade glioma transformation in adults** 

malignant transformation follows.

reviewed.

between 2,700 and 4,700 cases per year, comprising approximately 30% of all malignant gliomas (Schiff et al., 2007, Wessels et al., 2003). These tumors are most common in Caucasian males, and typically present in the second to fourth decades (Schomas et al., 2009, Wessels et al., 2003). Patients greater than 60 years of age carry a poorer prognosis with generally lower Karnofsky scores and larger tumor burden at diagnosis. In adults, the most common presenting symptom is seizure followed by incidental findings on imaging. Less common presentations include trauma trauma, sinus pathology, and pituitary disorder (Wessels et al., 2003). Thirty percent of patients present with neurological deficit, and only 10% present with symptoms of raised intracranial pressure ICP. Speech and language deficits have been reported in 10% of patients (Prabhu et al., 2010), however, focal deficits are less common (Schomas et al., 2009, Wessels et al., 2003).

In adults, LGGs are generally hemispheric, supratentorial, and typically occur in the frontal and temporal lobes. They may involve eloquent cortex, which limits the capacity for gross total resection due to significant risk of morbidity (Prabhu et al., 2010, Stieber, 2001). LGG are hypointense on T1 weighted magnetic resonance imaging (MRI), hyperintense on FLAIR and T2 sequences, and enhance in 30% of cases. There is often associated vasogenic edema (Prabhu et al., 2010, Wessels et al., 2003).

LGG are typically sporadic tumors, although they can occur in association with Li Fraumeni syndrome and Neurofibromatosis Types 1 and 2 (Prabhu et al., 2010, Wessels et al., 2003). Additional risk factors include previous irradiation and exposure to industrial chemicals (Prabhu et al., 2010, Wessels et al., 2003). Allergy has been reported to lower the risk for LGG, suggesting a possible role for immune surveillance in tumor pathogenesis (Prabhu et al., 2010). Survival is highly variable for LGG as median overall survival (OS) is reported to range from 3 to 40 years. Median progression free survival (PFS) is only 50% at 5 years and 17% at 15 years (Bauman et al., 1999, Berger et al., 1994, Jaeckle et al., Stieber, 2001). Median time to progression is 7.2 years (Schomas et al., 2009). In adults, the overall malignant transformation rate ranges from 35-89% with 74% in primary astrocytoma, 70% with mixed tumors, and 45% with primarily oligodendroglial histology (Jaeckle et al., 2010). Importantly, 50% of low risk adults, defined as patients less than 40 years of age with gross total resection (GTR), underwent transformation within 5 years (Jaeckle et al., 2010, Schiff et al., 2007).

Multiple studies have found that age greater than 40 years, extent of resection, tumor diameter greater than 6cm, tumor crossing midline, neurological deficit at diagnosis, and astrocytic histology are risk factors for poor prognosis in LGG (Bauman et al., 1999, E. G. Shaw et al., 2008, Stieber, 2001, Jaeckle et al., 2010, Schiff et al., 2007, E. G. Shaw & Wisoff, 2003). The NCCTG found that astrocytomas carry a worse prognosis than oligodendroglioma. Other retrospective reports corroborate these finding and further specify gemistocytic astrocytoma as carrying a worse prognosis (Jaeckle et al., 2010, Schomas et al., 2009, Stieber, 2001, Wessels et al., 2003, E. G. Shaw et al., 2008). Contrast enhancement, Karnofsky score, mitotic activity, and genetics have also been identified as risk factors for progression (Schiff et al., 2007, Schomas et al., 2009, E. G. Shaw & Wisoff, 2003, Stieber, 2001). Additionally, a Ki67-MIB1 index greater than 4% is associated with a more rapid rate of transformation.

#### **3. Malignant transformation of pediatric low-grade glioma**

The presentation and prognosis of LGG in children differs significantly from that in adults. Overall survival and rate of malignant transformation is significantly different in the

Bone Marrow-Derived Cells Support Malignant Transformation of Low-Grade Glioma 205

for malignant transformation (defined as age > 40yrs, astrocytic histology, crossing midline, diameter > 6cm, or intractable seizures) or for control of disease at the time of progression (Prabhu et al., 2010). The study also recommended RT to all patients greater than 40 years of age irrespective of resection, as age was the most consistent prognostic factor for malignant transformation (Stieber, 2001). For patients aged 18 to 40, RT was recommended only for patients with incomplete resection. Regardless of these data, treatment protocols vary

Chemotherapy has also been used as an initial treatment in LGG, most commonly in the setting of unresectable disease, or in patients less than 3 years of age in which RT should be deferred (Prabhu et al., 2010). The response rates to available agents are highly variable, with favorable responses reported between 10 and 60%. Poor response is often associated with low grade tumors as they tend to have lower sensitivity to chemotherapeutic agents due their inherently slow growth and minimal mitosis (Prabhu et al., 2010). The Southwest Oncology Group (SWOG) investigated the use of CCNU in addition to RT following GTR and found no added benefit of CCNU (E. G. Shaw & Wisoff, 2003). The NCCTG found a favorable response using PCV in the treatment of primary disease. Currently, the RTOG is investigating the safety and efficacy of PCV in unfavorable patients following resection and RT. Temozolamide is also under investigation for use in LGG patients at high risk for transformation (Schomas et al., 2009). Clearly, the use of adjuvant therapy requires more investigation before formal recommendations can be defined. Until then, adjuvant therapies

Treatment of pediatric gliomas is subject to a different set of considerations and standards as toxicity of therapy has a greater impact on the developing nervous and skeletal system. Surgery with GTR is the primary mode of therapy as this has been shown to be the most effective method for cure (Fisher et al., 2008, Unal et al., 2008). While rare, malignant transformation does occur so observation is not recommended with lesions that are amenable to surgery. As an exception, optic and hypothalamic gliomas are treated initially with observation and chemotherapy. Due to their slow growth and associated morbidity with surgery or radiotherapy in these locations, conservative management is standard. Ultimately, these tumors are associated with a worse prognosis due to their location and difficulty of surgical intervention in the event of progression. Similarly, first line of therapy for brainstem lesions is observation and potential biopsy only for progression of symptoms or radiographic appearance (Fisher et al., 2008). Based on the 0% progression in the setting of GTR, RT has no role following complete resection in children, as compared to adults

Although standard dose RT (50.4-54 Gy) has been shown to be effective in the pediatric population, RT is deferred in children irrespective of residual tumor burden, recurrence or progression due to the risk of toxicity including endocrine dysfunction, cognitive impairment with decreased memory, lower IQ, attention deficit, cerebrovascular disease, and secondary neoplasms (Fisher et al., 2008, Pollack et al., 1995). Standard dose RT is associated with 34% cognitive dysfunction compared to 8.6% without RT, and 17% endocrine dysfunction compared to 2.9% without RT (Pollack et al., 1995). Overall, the rate of endocrine dysfunction was 10% and cognitive dysfunction was 21%. These findings support the use of repeat surgery and chemotherapy prior to the use of RT for recurrence in children. Chemotherapeutic agents possess significant toxicity as well. While carboplatin and vincristine showed good response rates with 68% 3 year PFS, 40% of patients demonstrated hypersensitivity reactions. CCNU, vincristine, and dibromodulcitol have all

widely and are often practitioner dependant.

will remain controversial and site dependant.

(Pollack et al., 1995).

pediatric population, leading to the hypothesis that tumor biology in children is inherently different from that in adults. For LGG in children, the overall rate of malignant transformation ranges from 4.3%-38%, which is much lower than in adults (Armstrong et al., 2011, Pollack et al., 1995). This difference may be accounted for in part by the higher rate of pilocytic astrocytomas that comprise the vast majority of pediatric LGG, a histological subset that rarely transforms (Tihan et al., 1999). While no prospective studies have been performed to identify reliable risk factors for transformation in children, radiation therapy is reported to be a possible causative agent (Dirks et al., 1994). Mean time to transformation is relatively short at approximately 6.4 years (Dirks et al., 1994). While the overall rate of progression is certainly lower in children, the risk of transformation in this population is still significant and warrants active and expectant observation.

Despite this risk for malignant degeneration, overall prognosis for children with LGG is significantly better than that for adults. Overall survival in children with LGG 65-90%, however, OS is 51% when pilocytic pathology is excluded (Armstrong et al., 2011, Fisher et al., 2008, Pollack et al., 1995). Following gross total resection, survival is 90-100% with 0% progression, in comparison to the adult transformation rate of 50% even in low risk, young patients with complete resection (Pollack et al., 1995). Progression free survival is between approximately 50% at 10 years, and 53% at 15 years (Armstrong et al., 2011). Gross total resection has been the only factor currently identified to have an impact on progression free survival in children with 0% progression with GTR and 17% progression with near total resection (Pollack et al., 1995). Due to the infiltrative nature of non-pilocytic grade II astrocytomas, this histology in children is more comparable to the adult population and is associated with poorer prognosis (Pollack et al., 1995).

#### **4. Effect of resection and adjuvant therapy on malignant transformation**

Currently, initial treatment consists of pharmacologic seizure control if patients present with seizures and steroids for vasogenic edema (Prabhu et al., 2010, Stieber, 2001). For patients with lesions amenable to surgery, the goal is gross total resection as many studies have found overall survival to correlate with extent of initial resection irrespective of adjuvant therapy. At 5 years, OS was 63% with GTR versus 27% OS with STR (Prabhu et al., 2010). Recurrence is also higher with STR (Prabhu et al., 2010). Berger et al. (1994) reported no recurrences within 54 months with GTR, 14.8% recurrence with residual tumor less than 10cm3, and 46.2% recurrence with residual greater than 10cm3 (Berger et al., 1994, Stieber, 2001). In some cases, tumor location within or near eloquent cortex limits the extent of resection, therefore, newer methods including functional MRI, fiber tracking with diffuser tensor imaging (DTI), intra-operative stimulation and mapping, or intra-operative MRI have helped reduce morbidity and allow more aggressive surgery. As survival decreases with lower Karnofsky score, while the surgical goal remains complete resection, equally important is the avoidance of new neurological deficit (Gil-Robles & Duffau, 2010, Schomas et al., 2009).

The role of adjuvant therapy following surgical resection remains controversial. Although LGG are fairly slow growing tumors with low or absent mitotic activity, their infiltrative behavior and high rate of recurrence and malignant transformation has caused most centers to institute adjuvant therapy regardless of the extent of resection. Recent prospective trials have addressed the role of radiation therapy (E. Shaw et al., 2002). RT was found to improve PFS but not OS (Stieber, 2001). As a result, early RT is administered to patients at high risk

pediatric population, leading to the hypothesis that tumor biology in children is inherently different from that in adults. For LGG in children, the overall rate of malignant transformation ranges from 4.3%-38%, which is much lower than in adults (Armstrong et al., 2011, Pollack et al., 1995). This difference may be accounted for in part by the higher rate of pilocytic astrocytomas that comprise the vast majority of pediatric LGG, a histological subset that rarely transforms (Tihan et al., 1999). While no prospective studies have been performed to identify reliable risk factors for transformation in children, radiation therapy is reported to be a possible causative agent (Dirks et al., 1994). Mean time to transformation is relatively short at approximately 6.4 years (Dirks et al., 1994). While the overall rate of progression is certainly lower in children, the risk of transformation in this population is still

Despite this risk for malignant degeneration, overall prognosis for children with LGG is significantly better than that for adults. Overall survival in children with LGG 65-90%, however, OS is 51% when pilocytic pathology is excluded (Armstrong et al., 2011, Fisher et al., 2008, Pollack et al., 1995). Following gross total resection, survival is 90-100% with 0% progression, in comparison to the adult transformation rate of 50% even in low risk, young patients with complete resection (Pollack et al., 1995). Progression free survival is between approximately 50% at 10 years, and 53% at 15 years (Armstrong et al., 2011). Gross total resection has been the only factor currently identified to have an impact on progression free survival in children with 0% progression with GTR and 17% progression with near total resection (Pollack et al., 1995). Due to the infiltrative nature of non-pilocytic grade II astrocytomas, this histology in children is more comparable to the adult population and is

**4. Effect of resection and adjuvant therapy on malignant transformation** 

Currently, initial treatment consists of pharmacologic seizure control if patients present with seizures and steroids for vasogenic edema (Prabhu et al., 2010, Stieber, 2001). For patients with lesions amenable to surgery, the goal is gross total resection as many studies have found overall survival to correlate with extent of initial resection irrespective of adjuvant therapy. At 5 years, OS was 63% with GTR versus 27% OS with STR (Prabhu et al., 2010). Recurrence is also higher with STR (Prabhu et al., 2010). Berger et al. (1994) reported no recurrences within 54 months with GTR, 14.8% recurrence with residual tumor less than 10cm3, and 46.2% recurrence with residual greater than 10cm3 (Berger et al., 1994, Stieber, 2001). In some cases, tumor location within or near eloquent cortex limits the extent of resection, therefore, newer methods including functional MRI, fiber tracking with diffuser tensor imaging (DTI), intra-operative stimulation and mapping, or intra-operative MRI have helped reduce morbidity and allow more aggressive surgery. As survival decreases with lower Karnofsky score, while the surgical goal remains complete resection, equally important is the avoidance of new neurological deficit (Gil-Robles & Duffau, 2010, Schomas

The role of adjuvant therapy following surgical resection remains controversial. Although LGG are fairly slow growing tumors with low or absent mitotic activity, their infiltrative behavior and high rate of recurrence and malignant transformation has caused most centers to institute adjuvant therapy regardless of the extent of resection. Recent prospective trials have addressed the role of radiation therapy (E. Shaw et al., 2002). RT was found to improve PFS but not OS (Stieber, 2001). As a result, early RT is administered to patients at high risk

significant and warrants active and expectant observation.

associated with poorer prognosis (Pollack et al., 1995).

et al., 2009).

for malignant transformation (defined as age > 40yrs, astrocytic histology, crossing midline, diameter > 6cm, or intractable seizures) or for control of disease at the time of progression (Prabhu et al., 2010). The study also recommended RT to all patients greater than 40 years of age irrespective of resection, as age was the most consistent prognostic factor for malignant transformation (Stieber, 2001). For patients aged 18 to 40, RT was recommended only for patients with incomplete resection. Regardless of these data, treatment protocols vary widely and are often practitioner dependant.

Chemotherapy has also been used as an initial treatment in LGG, most commonly in the setting of unresectable disease, or in patients less than 3 years of age in which RT should be deferred (Prabhu et al., 2010). The response rates to available agents are highly variable, with favorable responses reported between 10 and 60%. Poor response is often associated with low grade tumors as they tend to have lower sensitivity to chemotherapeutic agents due their inherently slow growth and minimal mitosis (Prabhu et al., 2010). The Southwest Oncology Group (SWOG) investigated the use of CCNU in addition to RT following GTR and found no added benefit of CCNU (E. G. Shaw & Wisoff, 2003). The NCCTG found a favorable response using PCV in the treatment of primary disease. Currently, the RTOG is investigating the safety and efficacy of PCV in unfavorable patients following resection and RT. Temozolamide is also under investigation for use in LGG patients at high risk for transformation (Schomas et al., 2009). Clearly, the use of adjuvant therapy requires more investigation before formal recommendations can be defined. Until then, adjuvant therapies will remain controversial and site dependant.

Treatment of pediatric gliomas is subject to a different set of considerations and standards as toxicity of therapy has a greater impact on the developing nervous and skeletal system. Surgery with GTR is the primary mode of therapy as this has been shown to be the most effective method for cure (Fisher et al., 2008, Unal et al., 2008). While rare, malignant transformation does occur so observation is not recommended with lesions that are amenable to surgery. As an exception, optic and hypothalamic gliomas are treated initially with observation and chemotherapy. Due to their slow growth and associated morbidity with surgery or radiotherapy in these locations, conservative management is standard. Ultimately, these tumors are associated with a worse prognosis due to their location and difficulty of surgical intervention in the event of progression. Similarly, first line of therapy for brainstem lesions is observation and potential biopsy only for progression of symptoms or radiographic appearance (Fisher et al., 2008). Based on the 0% progression in the setting of GTR, RT has no role following complete resection in children, as compared to adults (Pollack et al., 1995).

Although standard dose RT (50.4-54 Gy) has been shown to be effective in the pediatric population, RT is deferred in children irrespective of residual tumor burden, recurrence or progression due to the risk of toxicity including endocrine dysfunction, cognitive impairment with decreased memory, lower IQ, attention deficit, cerebrovascular disease, and secondary neoplasms (Fisher et al., 2008, Pollack et al., 1995). Standard dose RT is associated with 34% cognitive dysfunction compared to 8.6% without RT, and 17% endocrine dysfunction compared to 2.9% without RT (Pollack et al., 1995). Overall, the rate of endocrine dysfunction was 10% and cognitive dysfunction was 21%. These findings support the use of repeat surgery and chemotherapy prior to the use of RT for recurrence in children. Chemotherapeutic agents possess significant toxicity as well. While carboplatin and vincristine showed good response rates with 68% 3 year PFS, 40% of patients demonstrated hypersensitivity reactions. CCNU, vincristine, and dibromodulcitol have all

Bone Marrow-Derived Cells Support Malignant Transformation of Low-Grade Glioma 207

and show resistance to current therapeutic regimens (Furnari et al., 2007, Hatanpaa et al., Johns et al., 2007, Pelloski et al., 2007). While PTEN mutations occur more frequently in primary glioblastoma in adults, PTEN mutations exist in high frequency in pediatric

> **Incidence in Primary GBM**

gliomas that have undergone malignant transformation (Broniscer et al., 2007).

**Incidence in Grade II Astrocytoma and Secondary GBM**

**p53 (TP53)** ↑↑↑ ↑

**EGFR** ↑ ↑↑↑

**PTEN** ↑↑ ↑↑↑

**IDH 1&2** ↑↑↑ ↑↑

**PDGF** ↑↑↑ ↑↑

**BRAF** ↑ --

Table 1. Various genetic mutations associated with gliomas. EGFR - epidermal growth factor receptor, PTEN - phosphatase and tensin homolog, IDH - isocitrate deyhdrogenase, PDGFR

In contrast, secondary GBMs often have p53 mutations and overexpress PDGF. Mutations of p53 frequently are associated with low-grade gliomas occurring in 53% of astrocytoma, 44% of oligoastrocytoma, and 13% of oligodendroglioma (Okamoto et al., 2004). Therefore, p53 may be an important molecular event involved in the malignant progression of low-grade gliomas (Louis et al., 2007). Interestingly, in children, the rate of p53 mutations is reported as only 10% in progressive pediatric LGG. While this alteration may seem to possibly explain the improved survival in pediatric gliomas, the 1p19q deletion, an indicator of a favorable response to specific chemotherapies in adults, is not found in pediatric gliomas

Other molecular changes have also been identified (Ichimura et al., 2009, Watanabe et al., 2009). IDH1 abnormalities exist in 59-88% of diffuse astrocytomas, 68-82% of oligodendrogliomas, 50-78% of anaplastic astrocytomas, 49-75% of anaplastic oligodendrogliomas, and 50-88% of secondary glioblastomas and often co-exist with p53 mutated lesions or 1p19q co-deleted tumors (Hartmann et al., 2009, Ichimura et al., 2009, Parsons et al., 2008, Sanson et al., 2009, Watanabe et al., 2009, Yan et al., 2009). While the presence of IDH mutations in low-grade tumors and secondary GBMs suggests a role for IDH in malignant progression , the literature suggests that the presence of IDH1 or IDH2 mutations correlates with better outcomes in patients (De Carli et al., 2009, Yan et al., 2009). PDGFR and the p16ink4a /RB1 pathway have also been implicated in gliomagenesis as hypermethylation of the RB1 gene may result in uncontrolled cell cycle progression, which may then drive tumor formation (Sathornsumetee et al., 2007). Both primary and secondary GBMs express PDGF, but increased RB1 gene promoter methylation appears to occur more frequently in secondary GBMs (43%) than primary GBMs (14%) (Nakamura et al., 2001). The expression of MGMT, a DNA repair enzyme, has also been implicated in glioblastoma and low-grade gliomas (Bourne & Schiff, 2010). Of particular interest is the methylation

**Genetic Mutation** 

– platelet derived growth factor receptor.

(Fisher et al., 2008).

been associated with significant hypersensitivity reactions (Fisher et al., 2008). As a result, TPCV is now being tested for efficacy and safety in a prospective pediatric trial (Fisher et al., 2008).

### **5. Histology of malignant transformation**

As mentioned previously, low grade gliomas comprise a histologically diverse group of tumors. The current WHO classification describes four categories for astrocytomas (Kleihues et al., 1995, Louis et al., 2007). While it is theorized that the majority of grade IV glioblastomas (GBM) occur *de novo* (primary GBM), a significant number of lesions result from progression of a low-grade tumor (secondary GBM). Excluding Grade I pilocytic astrocytomas as they rarely progress, low grade (II) and high grade (III and IV) astrocytomas can be viewed to exist along a continuum based on the histological analysis of tumor tissue. Grade II lesions are defined by low or absent mitotic activity and, unlike Grade I gliomas, are infiltrative and invasive and should not be considered benign. Cellular density is low to moderate, and well-differentiated, mildly pleomorphic tumor cells are present. One important feature of low grade astrocytomas is the absence of neovascularization.

This is in distinct comparison to high grade gliomas, grade III anaplastic tumors and grade IV GBMs, which are poorly differentiated, widely infiltrative and display prominent mitotic activity and neovascularization. Both confer a poor prognosis. High-grade lesions display increased cellularity, marked pleomorphism and nuclear atypia and may include multinucleated giant cells. Necrosis is the defining feature of GBM and these areas are typically surrounded by pseudopalisading cells. Most importantly, extensive irregular vascular proliferation is present in GBM as these tumors have adopted the capability of undergoing the angiogenic switch to produce their own vasculature, allowing for exponential tumor growth.

While morbidity is associated with low-grade astrocytomas themselves, it is hypothesized that the majority of morbidity is caused by progression to high-grade tumor. One of the key factors in this progression is the angiogenic switch whereby the tumor adopts the ability to acquire its own vascular supply. This enables explosive growth and precipitates rapid clinical deterioration. While an increased understanding of LGG biology and behavior has led to a more aggressive approach to these tumors, clinical outcome measures still remain poor. This is due mostly in part to our inability to prevent or detect malignant degeneration. A significant amount of research is now focused on understanding the factors involved in the angiogenic switch, which is likely to lead to additional treatment targets and potentially better outcomes. This will be further discussed in the sections to follow.

#### **6. Molecular biology of malignant transformation**

While histological characteristics currently determine tumor grade in astrocytoma, important molecular differences also exist between low grade and high-grade gliomas (Table 1) (Godard et al., 2003). These molecular differences are likely to be an important factor in initiating or promoting the angiogenic switch (Wen & Kesari, 2008). Both primary and secondary GBM exhibit elevated VEGF expression and loss of heterozygosity at 10q. The majority of primary GBM show overexpression of EGFR and PTEN mutations. In particular, glioblastomas that express the EGFRvIII genetic variant have a worse prognosis

been associated with significant hypersensitivity reactions (Fisher et al., 2008). As a result, TPCV is now being tested for efficacy and safety in a prospective pediatric trial (Fisher et al.,

As mentioned previously, low grade gliomas comprise a histologically diverse group of tumors. The current WHO classification describes four categories for astrocytomas (Kleihues et al., 1995, Louis et al., 2007). While it is theorized that the majority of grade IV glioblastomas (GBM) occur *de novo* (primary GBM), a significant number of lesions result from progression of a low-grade tumor (secondary GBM). Excluding Grade I pilocytic astrocytomas as they rarely progress, low grade (II) and high grade (III and IV) astrocytomas can be viewed to exist along a continuum based on the histological analysis of tumor tissue. Grade II lesions are defined by low or absent mitotic activity and, unlike Grade I gliomas, are infiltrative and invasive and should not be considered benign. Cellular density is low to moderate, and well-differentiated, mildly pleomorphic tumor cells are present. One important feature of low grade astrocytomas is the absence of

This is in distinct comparison to high grade gliomas, grade III anaplastic tumors and grade IV GBMs, which are poorly differentiated, widely infiltrative and display prominent mitotic activity and neovascularization. Both confer a poor prognosis. High-grade lesions display increased cellularity, marked pleomorphism and nuclear atypia and may include multinucleated giant cells. Necrosis is the defining feature of GBM and these areas are typically surrounded by pseudopalisading cells. Most importantly, extensive irregular vascular proliferation is present in GBM as these tumors have adopted the capability of undergoing the angiogenic switch to produce their own vasculature, allowing for

While morbidity is associated with low-grade astrocytomas themselves, it is hypothesized that the majority of morbidity is caused by progression to high-grade tumor. One of the key factors in this progression is the angiogenic switch whereby the tumor adopts the ability to acquire its own vascular supply. This enables explosive growth and precipitates rapid clinical deterioration. While an increased understanding of LGG biology and behavior has led to a more aggressive approach to these tumors, clinical outcome measures still remain poor. This is due mostly in part to our inability to prevent or detect malignant degeneration. A significant amount of research is now focused on understanding the factors involved in the angiogenic switch, which is likely to lead to additional treatment targets and potentially

While histological characteristics currently determine tumor grade in astrocytoma, important molecular differences also exist between low grade and high-grade gliomas (Table 1) (Godard et al., 2003). These molecular differences are likely to be an important factor in initiating or promoting the angiogenic switch (Wen & Kesari, 2008). Both primary and secondary GBM exhibit elevated VEGF expression and loss of heterozygosity at 10q. The majority of primary GBM show overexpression of EGFR and PTEN mutations. In particular, glioblastomas that express the EGFRvIII genetic variant have a worse prognosis

better outcomes. This will be further discussed in the sections to follow.

**6. Molecular biology of malignant transformation** 

2008).

neovascularization.

exponential tumor growth.

**5. Histology of malignant transformation** 

and show resistance to current therapeutic regimens (Furnari et al., 2007, Hatanpaa et al., Johns et al., 2007, Pelloski et al., 2007). While PTEN mutations occur more frequently in primary glioblastoma in adults, PTEN mutations exist in high frequency in pediatric gliomas that have undergone malignant transformation (Broniscer et al., 2007).


Table 1. Various genetic mutations associated with gliomas. EGFR - epidermal growth factor receptor, PTEN - phosphatase and tensin homolog, IDH - isocitrate deyhdrogenase, PDGFR – platelet derived growth factor receptor.

In contrast, secondary GBMs often have p53 mutations and overexpress PDGF. Mutations of p53 frequently are associated with low-grade gliomas occurring in 53% of astrocytoma, 44% of oligoastrocytoma, and 13% of oligodendroglioma (Okamoto et al., 2004). Therefore, p53 may be an important molecular event involved in the malignant progression of low-grade gliomas (Louis et al., 2007). Interestingly, in children, the rate of p53 mutations is reported as only 10% in progressive pediatric LGG. While this alteration may seem to possibly explain the improved survival in pediatric gliomas, the 1p19q deletion, an indicator of a favorable response to specific chemotherapies in adults, is not found in pediatric gliomas (Fisher et al., 2008).

Other molecular changes have also been identified (Ichimura et al., 2009, Watanabe et al., 2009). IDH1 abnormalities exist in 59-88% of diffuse astrocytomas, 68-82% of oligodendrogliomas, 50-78% of anaplastic astrocytomas, 49-75% of anaplastic oligodendrogliomas, and 50-88% of secondary glioblastomas and often co-exist with p53 mutated lesions or 1p19q co-deleted tumors (Hartmann et al., 2009, Ichimura et al., 2009, Parsons et al., 2008, Sanson et al., 2009, Watanabe et al., 2009, Yan et al., 2009). While the presence of IDH mutations in low-grade tumors and secondary GBMs suggests a role for IDH in malignant progression , the literature suggests that the presence of IDH1 or IDH2 mutations correlates with better outcomes in patients (De Carli et al., 2009, Yan et al., 2009). PDGFR and the p16ink4a /RB1 pathway have also been implicated in gliomagenesis as hypermethylation of the RB1 gene may result in uncontrolled cell cycle progression, which may then drive tumor formation (Sathornsumetee et al., 2007). Both primary and secondary GBMs express PDGF, but increased RB1 gene promoter methylation appears to occur more frequently in secondary GBMs (43%) than primary GBMs (14%) (Nakamura et al., 2001).

The expression of MGMT, a DNA repair enzyme, has also been implicated in glioblastoma and low-grade gliomas (Bourne & Schiff, 2010). Of particular interest is the methylation

Bone Marrow-Derived Cells Support Malignant Transformation of Low-Grade Glioma 209

between low- and high-grade gliomas using a threshold FA value of 0.188 (Inoue et al.,

Fig. 1. Serial imaging of malignant progression of glioma. A, T1-weighted MRI with contrast. Patient presented with headache. MRI revealed hypointense lesion in right hemisphere. Note edema and mass effect but lack of contrast enhancement. Pt underwent gross total surgical resection and pathology revealed grade II astrocytoma. B, T1-weighted MRI with contrast. Subsequent imaging revealed recurrent tumor seen as a contrast enhancing lesion in the previous resection cavity. Pathology revealed progression to grade

MR spectroscopy (MRS) also can potentially differentiate low-grade versus high-grade gliomas in the brain. All gliomas have an increased choline peak and a reduced N-acetyl aspartate peak (NAA) which are markers of membrane turnover and neuronal cell death respectively. Levels of lipid and lactate are markers of necrosis and hypoxia respectively and are decidedly elevated in high-grade compared to low-grade gliomas (McBride et al., 1995, Nafe et al., 2003, Negendank et al., 1996). Creatine (Cr), which serves as a marker of energy metabolism, is decreased in brain tumors (Meyerand et al., 1999, Moller-Hartmann et al., 2002), however, this reduction does not appear to correlate with tumor grade by itself (Moller-Hartmann et al., 2002). Using the choline/Cr ratio may be more effective, however, as low-grade gliomas tend to have a lower ratio of choline/Cr (McBride et al., 1995, Murphy et al., 2002, Sijens & Oudkerk, 2002), as well as an increase in NAA/Cr ratio (Law et al., 2003, McKnight et al., 2002, Murphy et al., 2002, Nafe et al., 2003, Negendank et al., 1996). MRS is a promising technique in differentiating low- from high-grade gliomas with sensitivity between 73% and 92% and specificity between 63% and 100% (Astrakas et al., 2004, Fayed & Modrego, 2005, Law et al., 2003, Nafe et al., 2003, Setzer et al., 2007). MRS may also be capable of identifying regions that have undergone malignant transformation within a given tumor that may not be identifiable by other imaging techniques although one such study attempting to detect malignant transformation within low-grade glioma yielded

Positron emission tomography (PET) imaging has been employed to examine gliomas in the brain by measuring the metabolic activity of tissue. Fluorinated glucose analogue 2-[18F] fluoro-2-deoxy-D-glucose (FDG), which is administered to patients intravascularly, has high sensitivity for identifying areas of increased tumor metabolism and has been used as an index to predict tumor aggressiveness. While low-grade gliomas tend to have the same or even lower uptake of FDG than normal brain matter, high-grade gliomas demonstrate increased uptake of FDG on PET imaging (Derlon et al., 1997, Tamura et al., 1998), Studies have shown that it is possible to differentiate low- from high-grade gliomas with a sensitivity of 94% and specificity of 77% using a tumor-to-white-matter ratio of greater than

1.5 and tumor-to-grey-matter ratio of greater than 0.6 (Delbeke et al., 1995).

2005).

IV astrocytoma (GBM).

a specificity of only 57.1% (Alimenti et al., 2007).

status of MGMT as it may correlate to resistance to alkylating therapy in some patients (Hegi et al., 2005).

Finally, chromosomal e 7 (7q34) gene BRAF mutations and overexpression of B-raf, which stimulates the mitogen-activated protein kinase (MAPK) pathway, is a major factor in tumorigenesis of pilocytic astrocytomas (Pfister et al., 2008). This mutation is also present in 23-38% of adult grade II astrocytomas. The role of BRAF mutation in progression to highgrade tumors, however, has yet to be elucidated.

Defining molecular differences amongst glioma subpopulations offers an exciting new dynamic in understanding the behaviors of this highly diverse tumor although much work is required before the variability observed is completely delineated. Already, studies are underway to target tumors at the molecular level in hopes of providing better treatment options (Johns et al., 2007). As it is apparent that the angiogenic switch is important in the progression of low-grade to high-grade glioma, defining the molecular changes that promote this event may offer additional treatment benefits. Animal studies have already shown that preventing the angiogenic switch in other solid tumors reduces tumor growth (Lyden et al., 2001). Therefore, further understanding of how specific molecular changes in tumor cells promote angiogenesis may offer promising new treatment options in gliomas.

#### **7. Advancing imaging of low-grade gliomas**

MRI is the initial imaging modality of choice in brain tumors. Low-grade gliomas usually appear as well defined lesions with little mass effect. They have low-signal on T1- and highsignal on T2-weighted imaging - particularly on fluid attenuated inversion recovery (FLAIR) sequences where low-grade gliomas are very hyperintense (Kates et al., 1996). Currently, the absence of gadolinium enhancement is used to differentiate low grade versus high grade glioma (Fig. 2) (Castillo, 1994), however, a significant portion of the low-grade gliomas defined by MRI were found to be high-grade after biopsy (Kondziolka et al., 1993). As a result, MRI is not sensitive enough to definitively diagnose low-grade gliomas as there are frequently small areas within the tumor that have already undergone malignant progression. Therefore, advanced imaging technologies, such as perfusion imaging, diffusion-weighted and diffusion tensor imaging, MR spectroscopy, and position emission tomography (PET), are currently being employed to more accurately identify low-grade versus high-grade gliomas. These modalities provide exciting insight into tumor vascularity, cellularity, metabolism, and proliferation and may prove more effective in differentiating low-grade from high-grade glioma particularly in regions within a given tumor.

Since the degree of vascularity correlates with tumor grade in gliomas, (Daumas-Duport et al., 1997) perfusion MRI and MRI with gradient echo differentiates low-grade versus highgrade gliomas based on relative cerebral blood volume (rCBV) (Boxerman et al., 2006, Law et al., 2003, Law et al., 2004, Shin et al., 2002, Sugahara et al., 1998, Sugahara et al., 2001). While promising, it has been difficult to establish a reliable threshold based on rCBV for low- versus high-grade state. Diffusion-weighted MRI has also been utilized based on the apparent diffusion coefficient, which inversely correlates with tumor cellularity (Gauvain et al., 2001, Kono et al., 2001, Sugahara et al., 1999). Again, it has been difficult to reliably predict tumor grade using diffusion MRI (Bulakbasi et al., 2003, Stieber, 2001). Diffusion tensor imaging (DTI) is a modification of diffusion-weighted imaging and measures fractional anisotropy (FA), which correlated with tumor cellularity and vascularity (Price, 2010). DTI is a promising new modality as one study reports the ability to distinguish

status of MGMT as it may correlate to resistance to alkylating therapy in some patients

Finally, chromosomal e 7 (7q34) gene BRAF mutations and overexpression of B-raf, which stimulates the mitogen-activated protein kinase (MAPK) pathway, is a major factor in tumorigenesis of pilocytic astrocytomas (Pfister et al., 2008). This mutation is also present in 23-38% of adult grade II astrocytomas. The role of BRAF mutation in progression to high-

Defining molecular differences amongst glioma subpopulations offers an exciting new dynamic in understanding the behaviors of this highly diverse tumor although much work is required before the variability observed is completely delineated. Already, studies are underway to target tumors at the molecular level in hopes of providing better treatment options (Johns et al., 2007). As it is apparent that the angiogenic switch is important in the progression of low-grade to high-grade glioma, defining the molecular changes that promote this event may offer additional treatment benefits. Animal studies have already shown that preventing the angiogenic switch in other solid tumors reduces tumor growth (Lyden et al., 2001). Therefore, further understanding of how specific molecular changes in tumor cells promote angiogenesis may offer promising new treatment options in gliomas.

MRI is the initial imaging modality of choice in brain tumors. Low-grade gliomas usually appear as well defined lesions with little mass effect. They have low-signal on T1- and highsignal on T2-weighted imaging - particularly on fluid attenuated inversion recovery (FLAIR) sequences where low-grade gliomas are very hyperintense (Kates et al., 1996). Currently, the absence of gadolinium enhancement is used to differentiate low grade versus high grade glioma (Fig. 2) (Castillo, 1994), however, a significant portion of the low-grade gliomas defined by MRI were found to be high-grade after biopsy (Kondziolka et al., 1993). As a result, MRI is not sensitive enough to definitively diagnose low-grade gliomas as there are frequently small areas within the tumor that have already undergone malignant progression. Therefore, advanced imaging technologies, such as perfusion imaging, diffusion-weighted and diffusion tensor imaging, MR spectroscopy, and position emission tomography (PET), are currently being employed to more accurately identify low-grade versus high-grade gliomas. These modalities provide exciting insight into tumor vascularity, cellularity, metabolism, and proliferation and may prove more effective in differentiating

low-grade from high-grade glioma particularly in regions within a given tumor.

Since the degree of vascularity correlates with tumor grade in gliomas, (Daumas-Duport et al., 1997) perfusion MRI and MRI with gradient echo differentiates low-grade versus highgrade gliomas based on relative cerebral blood volume (rCBV) (Boxerman et al., 2006, Law et al., 2003, Law et al., 2004, Shin et al., 2002, Sugahara et al., 1998, Sugahara et al., 2001). While promising, it has been difficult to establish a reliable threshold based on rCBV for low- versus high-grade state. Diffusion-weighted MRI has also been utilized based on the apparent diffusion coefficient, which inversely correlates with tumor cellularity (Gauvain et al., 2001, Kono et al., 2001, Sugahara et al., 1999). Again, it has been difficult to reliably predict tumor grade using diffusion MRI (Bulakbasi et al., 2003, Stieber, 2001). Diffusion tensor imaging (DTI) is a modification of diffusion-weighted imaging and measures fractional anisotropy (FA), which correlated with tumor cellularity and vascularity (Price, 2010). DTI is a promising new modality as one study reports the ability to distinguish

(Hegi et al., 2005).

grade tumors, however, has yet to be elucidated.

**7. Advancing imaging of low-grade gliomas** 

between low- and high-grade gliomas using a threshold FA value of 0.188 (Inoue et al., 2005).

Fig. 1. Serial imaging of malignant progression of glioma. A, T1-weighted MRI with contrast. Patient presented with headache. MRI revealed hypointense lesion in right hemisphere. Note edema and mass effect but lack of contrast enhancement. Pt underwent gross total surgical resection and pathology revealed grade II astrocytoma. B, T1-weighted MRI with contrast. Subsequent imaging revealed recurrent tumor seen as a contrast enhancing lesion in the previous resection cavity. Pathology revealed progression to grade IV astrocytoma (GBM).

MR spectroscopy (MRS) also can potentially differentiate low-grade versus high-grade gliomas in the brain. All gliomas have an increased choline peak and a reduced N-acetyl aspartate peak (NAA) which are markers of membrane turnover and neuronal cell death respectively. Levels of lipid and lactate are markers of necrosis and hypoxia respectively and are decidedly elevated in high-grade compared to low-grade gliomas (McBride et al., 1995, Nafe et al., 2003, Negendank et al., 1996). Creatine (Cr), which serves as a marker of energy metabolism, is decreased in brain tumors (Meyerand et al., 1999, Moller-Hartmann et al., 2002), however, this reduction does not appear to correlate with tumor grade by itself (Moller-Hartmann et al., 2002). Using the choline/Cr ratio may be more effective, however, as low-grade gliomas tend to have a lower ratio of choline/Cr (McBride et al., 1995, Murphy et al., 2002, Sijens & Oudkerk, 2002), as well as an increase in NAA/Cr ratio (Law et al., 2003, McKnight et al., 2002, Murphy et al., 2002, Nafe et al., 2003, Negendank et al., 1996). MRS is a promising technique in differentiating low- from high-grade gliomas with sensitivity between 73% and 92% and specificity between 63% and 100% (Astrakas et al., 2004, Fayed & Modrego, 2005, Law et al., 2003, Nafe et al., 2003, Setzer et al., 2007). MRS may also be capable of identifying regions that have undergone malignant transformation within a given tumor that may not be identifiable by other imaging techniques although one such study attempting to detect malignant transformation within low-grade glioma yielded a specificity of only 57.1% (Alimenti et al., 2007).

Positron emission tomography (PET) imaging has been employed to examine gliomas in the brain by measuring the metabolic activity of tissue. Fluorinated glucose analogue 2-[18F] fluoro-2-deoxy-D-glucose (FDG), which is administered to patients intravascularly, has high sensitivity for identifying areas of increased tumor metabolism and has been used as an index to predict tumor aggressiveness. While low-grade gliomas tend to have the same or even lower uptake of FDG than normal brain matter, high-grade gliomas demonstrate increased uptake of FDG on PET imaging (Derlon et al., 1997, Tamura et al., 1998), Studies have shown that it is possible to differentiate low- from high-grade gliomas with a sensitivity of 94% and specificity of 77% using a tumor-to-white-matter ratio of greater than 1.5 and tumor-to-grey-matter ratio of greater than 0.6 (Delbeke et al., 1995).

Bone Marrow-Derived Cells Support Malignant Transformation of Low-Grade Glioma 211

(vasculogenesis). These factors mobilize bone marrow precursor cells which then travel to the site of ischemia via the bloodstream. It is theorized that these cells facilitate vasculogenesis by breaking down existing structures and creating an environment that

Tumors are capable of adopting this machinery to increase growth and invasiveness by activating the angiogenic switch (Bergers & Benjamin, 2003, Rafii & Lyden, 2008). During early tumor development, neoplastic cells rely on existing blood flow and grow in a slow linear fashion (Mandonnet et al., 2003). Once the switch is initiated and neovascularization brings more oxygen and nutrients, tumor cells grow at a much faster rate and tumor size increases significantly (Rees et al., 2009). This initial process is thought to occur mostly by angiogenesis (Kioi, 2010). The release of proteases and proangiogenic factors causes pericytes to detach from existing vessels creating a defect in the extracellular matrix in the environment surrounding the vessel wall (Bergers & Benjamin, 2003). Endothelial cells proliferate locally and sprout outward into the tumor bed creating newly formed blood vessels feeding the tumor. While angiogenesis is an important factor in the angiogenic switch, vasculogenesis and the contribution of BMDC play a critical role as well. For example, when recruitment of BMDC is impaired in an animal model of lymphoma and lung carcinoma, tumor angiogenesis and growth is significantly decreased (Lyden et al., 2001) suggesting that BMDCs contribute significantly to neovascularization and growth in

In metastatic disease, the contribution of BMDC has been well described (Wels et al., 2008). Endothelial (EPCs) and hematopoietic precursor cells (HPC), mesenchymal stem cells (MSC), myeloid-derived suppressor cells (MDSCs), Tie-2 expressing monocytes (TEM) and tumor associated macrophages (TAM) all are mobilized from the bone marrow to future metastatic sites prior to tumor formation. It should be noted that these primitive cells are prominent during embryology and that a significant population of these cells is not present under normal conditions. While the exact role of each cell type has yet to be fully elucidated, their basic function is to break down normal structures and promote vasculogenesis and tumor invasiveness. The net result is a tumor friendly environment capable of sustaining tumor growth. This has been demonstrated experimentally in a murine model of metastatic disease by implanting m-cherry labeled melanoma cells into the flank of mice with GFPlabeled bone marrow and examining the lungs of these animals over time (Kaplan et al., 2005). It was observed that the first cells to arrive in future metastases were not tumor cells, but actual BMDC. This suggested, at least in metastatic disease, that the environment in future metastatic sites is primed by cells from the bone marrow before tumors can begin to grow in these distant areas (Rafii & Lyden, 2008). This also supports the hypothesis by Stephen Paget over 100 years ago that the tumor microenvironment may play as important a

In the brain, the role of BMDC has only recently garnered attention. One of the basic histological differences between low-grade and high-grade gliomas is a lack of neovascularization. Thus, activation of the angiogenic switch is a key element in the transformation of low-grade to high-grade glioma. Two elements are likely to contribute to this process. Genetic changes in tumor cells that occur during progression of disease activate pro-angiogenic factors. This has been observed in human tumor samples whereby genes involved in angiogenesis are upregulated in glioblastoma as compared to low grade astrocytoma (Godard et al., 2003). Kioi et al. also showed in their animal model that release of soluble factors by tumor cells or cells within the tumor microenvironment including

promotes new blood vessel growth.

role as the tumor cells themselves.

solid tumors.

In addition to FDG, other tracers have been utilized in attempts to further characterize these tumors such as carbon 11 and fluorine 18 (18F)-labeled amino acid (Isselbacher, 1972). Methionine PET appears to have a higher sensitivity than FDG PET in detecting low-grade versus high-grade gliomas (Derlon et al., 1997, Giammarile et al., 2004, Ogawa et al., 1993). In particular, methionine PET exhibits a heightened sensitivity in detecting radiation necrosis from recurrent tumors, as inflammatory cells in radiation necrosis have little uptake of methionine (Thiel et al., 2000). Perhaps the most promising technique for diagnosing lowgrade gliomas is 18F-FDOPA PET imaging. 18F-FDOPA PET is more accurate that FDG PET and has been shown to be highly predictive in determining tumor grade on initial diagnosis and may help differentiate tumor necrosis from recurrence (Chen et al., 2006, Fueger et al., Tripathi et al., 2009).

While there currently is no one imaging modality capable of definitively determining lowgrade from high-grade tumors on its own, advanced imaging technology continues to develop and complement standard MRI. As we come to understand the behavior and variability of these tumors, advance imaging techniques provide exciting new possibilities for more precise treatments. Given the variability within a given tumor, advanced imaging techniques may allow for more precise targets for biopsy, vigilant monitoring of malignant transformation, and improved prognostic power in the management of low- and high-grade gliomas.

#### **8. The Role of bone-marrow derived cells in malignant transformation**

The vast majority of brain tumor research, molecular profiling, histological characteristics, diagnostic imaging modalities and treatment targets have focused on the actual tumor cells themselves. As mentioned earlier, one of the key events in the transition from the low-grade to the high-grade state is the angiogenic switch. It is theorized that in the low-grade state, tumor growth may be limited, at least in part, by a lack of blood supply. In this state, the tumor is only capable of a steady-state or linear growth (Mandonnet et al., 2003). Once the tumor acquires the ability to recruit or form new blood vessels, exponential growth occurs (Rees et al., 2009) resulting in rapid clinical decline. While there is considerable evidence that tumor cells undergo continued molecular changes that increase their malignant potential, these changes also allow these cells to initiate the angiogenic switch. It must also be noted that while recent evidence suggests that tumor cells may be capable of directly forming new blood vessels (Ricci-Vitiani et al., Wang et al.), a considerable body of evidence suggests that tumor cells do not do this completely on their own. While the exact details of this process still remain to be fully elucidated, tumor cells acquire the ability to transition the local tumor niche to an environment capable of rapid blood vessel formation. A variety of growth factors, signaling pathways, and indigenous populations of cells is hypothesized to participate in this process. If this theory proves to be correct, this population of cells forms an additional therapeutic target that may be as important as the tumor cells themselves. As current therapies directed at neoplastic cells are limited in part to their toxicity, elucidating other potential treatment pathways may further benefit patient outcome.

Neovascularization is a normal process in tissues and the brain during ischemia. In low oxygen states, cells release signals such as VEGF, PDGF, PlGF and HIF-1 that recruit from local existing vessels within the tissue itself (angiogenesis.) In addition, this can activate distant processes that facilitate neovascularization and may even form *de novo* blood vessels

In addition to FDG, other tracers have been utilized in attempts to further characterize these tumors such as carbon 11 and fluorine 18 (18F)-labeled amino acid (Isselbacher, 1972). Methionine PET appears to have a higher sensitivity than FDG PET in detecting low-grade versus high-grade gliomas (Derlon et al., 1997, Giammarile et al., 2004, Ogawa et al., 1993). In particular, methionine PET exhibits a heightened sensitivity in detecting radiation necrosis from recurrent tumors, as inflammatory cells in radiation necrosis have little uptake of methionine (Thiel et al., 2000). Perhaps the most promising technique for diagnosing lowgrade gliomas is 18F-FDOPA PET imaging. 18F-FDOPA PET is more accurate that FDG PET and has been shown to be highly predictive in determining tumor grade on initial diagnosis and may help differentiate tumor necrosis from recurrence (Chen et al., 2006, Fueger et al.,

While there currently is no one imaging modality capable of definitively determining lowgrade from high-grade tumors on its own, advanced imaging technology continues to develop and complement standard MRI. As we come to understand the behavior and variability of these tumors, advance imaging techniques provide exciting new possibilities for more precise treatments. Given the variability within a given tumor, advanced imaging techniques may allow for more precise targets for biopsy, vigilant monitoring of malignant transformation, and improved prognostic power in the management of low- and high-grade

**8. The Role of bone-marrow derived cells in malignant transformation** 

The vast majority of brain tumor research, molecular profiling, histological characteristics, diagnostic imaging modalities and treatment targets have focused on the actual tumor cells themselves. As mentioned earlier, one of the key events in the transition from the low-grade to the high-grade state is the angiogenic switch. It is theorized that in the low-grade state, tumor growth may be limited, at least in part, by a lack of blood supply. In this state, the tumor is only capable of a steady-state or linear growth (Mandonnet et al., 2003). Once the tumor acquires the ability to recruit or form new blood vessels, exponential growth occurs (Rees et al., 2009) resulting in rapid clinical decline. While there is considerable evidence that tumor cells undergo continued molecular changes that increase their malignant potential, these changes also allow these cells to initiate the angiogenic switch. It must also be noted that while recent evidence suggests that tumor cells may be capable of directly forming new blood vessels (Ricci-Vitiani et al., Wang et al.), a considerable body of evidence suggests that tumor cells do not do this completely on their own. While the exact details of this process still remain to be fully elucidated, tumor cells acquire the ability to transition the local tumor niche to an environment capable of rapid blood vessel formation. A variety of growth factors, signaling pathways, and indigenous populations of cells is hypothesized to participate in this process. If this theory proves to be correct, this population of cells forms an additional therapeutic target that may be as important as the tumor cells themselves. As current therapies directed at neoplastic cells are limited in part to their toxicity, elucidating other potential treatment pathways may further benefit patient

Neovascularization is a normal process in tissues and the brain during ischemia. In low oxygen states, cells release signals such as VEGF, PDGF, PlGF and HIF-1 that recruit from local existing vessels within the tissue itself (angiogenesis.) In addition, this can activate distant processes that facilitate neovascularization and may even form *de novo* blood vessels

Tripathi et al., 2009).

gliomas.

outcome.

(vasculogenesis). These factors mobilize bone marrow precursor cells which then travel to the site of ischemia via the bloodstream. It is theorized that these cells facilitate vasculogenesis by breaking down existing structures and creating an environment that promotes new blood vessel growth.

Tumors are capable of adopting this machinery to increase growth and invasiveness by activating the angiogenic switch (Bergers & Benjamin, 2003, Rafii & Lyden, 2008). During early tumor development, neoplastic cells rely on existing blood flow and grow in a slow linear fashion (Mandonnet et al., 2003). Once the switch is initiated and neovascularization brings more oxygen and nutrients, tumor cells grow at a much faster rate and tumor size increases significantly (Rees et al., 2009). This initial process is thought to occur mostly by angiogenesis (Kioi, 2010). The release of proteases and proangiogenic factors causes pericytes to detach from existing vessels creating a defect in the extracellular matrix in the environment surrounding the vessel wall (Bergers & Benjamin, 2003). Endothelial cells proliferate locally and sprout outward into the tumor bed creating newly formed blood vessels feeding the tumor. While angiogenesis is an important factor in the angiogenic switch, vasculogenesis and the contribution of BMDC play a critical role as well. For example, when recruitment of BMDC is impaired in an animal model of lymphoma and lung carcinoma, tumor angiogenesis and growth is significantly decreased (Lyden et al., 2001) suggesting that BMDCs contribute significantly to neovascularization and growth in solid tumors.

In metastatic disease, the contribution of BMDC has been well described (Wels et al., 2008). Endothelial (EPCs) and hematopoietic precursor cells (HPC), mesenchymal stem cells (MSC), myeloid-derived suppressor cells (MDSCs), Tie-2 expressing monocytes (TEM) and tumor associated macrophages (TAM) all are mobilized from the bone marrow to future metastatic sites prior to tumor formation. It should be noted that these primitive cells are prominent during embryology and that a significant population of these cells is not present under normal conditions. While the exact role of each cell type has yet to be fully elucidated, their basic function is to break down normal structures and promote vasculogenesis and tumor invasiveness. The net result is a tumor friendly environment capable of sustaining tumor growth. This has been demonstrated experimentally in a murine model of metastatic disease by implanting m-cherry labeled melanoma cells into the flank of mice with GFPlabeled bone marrow and examining the lungs of these animals over time (Kaplan et al., 2005). It was observed that the first cells to arrive in future metastases were not tumor cells, but actual BMDC. This suggested, at least in metastatic disease, that the environment in future metastatic sites is primed by cells from the bone marrow before tumors can begin to grow in these distant areas (Rafii & Lyden, 2008). This also supports the hypothesis by Stephen Paget over 100 years ago that the tumor microenvironment may play as important a role as the tumor cells themselves.

In the brain, the role of BMDC has only recently garnered attention. One of the basic histological differences between low-grade and high-grade gliomas is a lack of neovascularization. Thus, activation of the angiogenic switch is a key element in the transformation of low-grade to high-grade glioma. Two elements are likely to contribute to this process. Genetic changes in tumor cells that occur during progression of disease activate pro-angiogenic factors. This has been observed in human tumor samples whereby genes involved in angiogenesis are upregulated in glioblastoma as compared to low grade astrocytoma (Godard et al., 2003). Kioi et al. also showed in their animal model that release of soluble factors by tumor cells or cells within the tumor microenvironment including

Bone Marrow-Derived Cells Support Malignant Transformation of Low-Grade Glioma 213

Mobilization of BMDC in peripheral blood samples has similarly been observed in patients with astrocytomas. Circulating CD133+ and VEGFR2+ EPC were measured in patients with different grade gliomas. This population of cells was significantly elevated in brain tumor patients versus controls, correlated with tumor grade, and predicted survival. In one patient, this population also predicted recurrence prior to detection by serial radiographic study. Currently, patients are followed with serial imaging in order to diagnose recurrence or malignant progression. While advances in imaging technology show promise in earlier more accurate diagnosis, the critical event has already occurred and prognosis worsens considerably. Therefore, the identification of a potential surrogate biomarker that measures tumor angiogenicity and aggressiveness may potentially serve as an index for ongoing

As histological and molecular differences between low-grade and high-grade gliomas are further defined and it becomes apparent that tumors cannot be loosely classified, specific treatments based on the particular characteristics of each individual tumor can potentially be designed. In addition, the presence of particular populations of BMDC in these tumors may also provide additional information on tumor behavior and serve as an additional treatment target along with tumor cells themselves. It has already been shown that the presence of BMDC in the blood correlates with tumor grade and initial animal studies suggest that BMDC are present in high-grade tumors only (Greenfield et al., 2009). In addition, TAM have been associated with poorer prognosis in metastatic lesions and other solid tumors (Wels et al., 2008). Thus histological stains aimed at identifying this population of cells may provide more accurate diagnosis and prognosis. Likewise, the molecular markers of this particular population of cells may offer an even more specific therapeutic target. Based on data collected in glioma patients, EPC can be identified by cell surface markers including CD133 and VEGFR2. Knocking down this population with specifically designed drug therapies has the potential for preventing recurrence by decreasing migration of these cells and reducing vasculogenesis within the tumor bed. Finally, one can also envision a role for advanced imaging technologies for improved diagnosis and treatment. For instance, PET has been used to specifically measure VEGF that has been labeled with copper in an orthotropic mouse model of GBM (Cai et al., 2006). If one could identify and label molecular targets that are specific to individual tumors subtypes and sensitive to new imaging techniques, this provides exciting non-invasive possibilities for tumor specific

In summary, one of the primary factors predicting outcome in patients with low-grade glioma is malignant progression to high-grade tumor and it is evident that the angiogenic switch is an important event in this process. Initial management often entails surgical resection while adjuvant therapy for low-grade gliomas remains a controversial topic. Tumor grade is determined by histological analysis of tumor specimens, but the molecular fingerprint of these tumors is now being analyzed more thoroughly and holds promise for more exciting targeted treatment options. In addition, distinct, but as yet undefined, populations of cells are recruited to the tumor site and participate in neovascularization and promote tumor growth and invasiveness. Therefore, this population may represent an important therapeutic target in combating these tumors. Since survival is directly correlated with tumor grade, preventing tumor progression is imperative. While BMDC certainly are

treatment effectiveness or recurrence.

identification and treatment for each individual patient.

**9. Conclusions** 

VEGF, FGF and EGF stimulates local angiogenesis (Kioi, 2010). Secondly, hypoxia is an additional critical event in triggering the switch (Kioi, 2010). As tumor size grows and metabolic demand exceeds local perfusion, hypoxic conditions occur. Release of hypoxia inducible factor-1 (HIF-1α) by tumor cells or cells within the hypoxic tumor environment, combined with stromal cell-derived factor-1 (SDF-1) and CXCR-4 receptor activation, mobilizes BMDCs to the tumor site and promotes vasculogenesis in gliomas (Du et al., 2008, Greenfield et al., 2010, Kioi, 2010).

In an attempt to further understand these processes in gliomas, Du et al. utilized an orthotopic model of GBM in mice to demonstrate recruitment of BMDC in gliomas (Du et al., 2008). Based on their results they theorize that hypoxia and the subsequent release of HIF-1α is the key event in tumor progression. Elevation of VEGF, and subsequent SDF-1 release and CXCR-4 receptor activation, mobilizes BMDC and recruits EPC and myeloid cells to the tumor. The net effect tips the balance to a pro-angiogenic state and neovascularization within the tumor bed. Kioi et al also further theorized that radiation treatment may exacerbate the vasculogenesis process and boost eventual tumor recurrence observed in current treatment regimens (Greenfield et al., 2010, Kioi, 2010). The endothelialmesenchymal transition and MSC have also been described in metastatic disease (Singh & Settleman). MSC exist within the brain and mobilize to the tumor site as well (Hata et al., Kang et al.). The exact roles of these particular BMDC remain elusive and require more study before they are fully delineated.

In our laboratory, we have begun to investigate the correlation of BMDC mobilization and tumor grade in gliomas (unpublished data.) We used a PDGF-driven mouse model of GBM within which tumors develop slowly from low-grade to high-grade. Low-grade tumors have a clear absence of neovascularization and BMDC are not present within these lesions. In high-grade tumors, however, we have observed a profound increase in larger, irregularly shaped, hemorrhagic vessels and a significant population of BMDC exists that is not observed in low grade tumors. In addition, these cells are located near newly forming blood vessels in the perivascular niche. We have also observed that BMDC are mobilized in the bone marrow and are elevated in the peripheral blood of tumor bearing animals versus controls. In addition, a significant difference in this population of cells in the blood exists between low-grade and high-grade animals. While much work is yet to be done before this process is fully elucidated, it appears that the presence of BMDC correlates with tumor grade and the process of neovascularization. Thus, BMDC have a potential role in the angiogenic switch as tumors progress from low-grade to high-grade tumors.

Fig. 2. Bone marrow-derived cells in human glioma. A, Immunofluorescence in grade II astrocytoma shows normal blood vessels (red, VE Cadherin) and a paucity of CD11b+ myeloid suppressor cells (Green). B, GBM shows abnormal vessel formation and an influx of CD11b+ cells (unpublished data.)

Mobilization of BMDC in peripheral blood samples has similarly been observed in patients with astrocytomas. Circulating CD133+ and VEGFR2+ EPC were measured in patients with different grade gliomas. This population of cells was significantly elevated in brain tumor patients versus controls, correlated with tumor grade, and predicted survival. In one patient, this population also predicted recurrence prior to detection by serial radiographic study. Currently, patients are followed with serial imaging in order to diagnose recurrence or malignant progression. While advances in imaging technology show promise in earlier more accurate diagnosis, the critical event has already occurred and prognosis worsens considerably. Therefore, the identification of a potential surrogate biomarker that measures tumor angiogenicity and aggressiveness may potentially serve as an index for ongoing treatment effectiveness or recurrence.

As histological and molecular differences between low-grade and high-grade gliomas are further defined and it becomes apparent that tumors cannot be loosely classified, specific treatments based on the particular characteristics of each individual tumor can potentially be designed. In addition, the presence of particular populations of BMDC in these tumors may also provide additional information on tumor behavior and serve as an additional treatment target along with tumor cells themselves. It has already been shown that the presence of BMDC in the blood correlates with tumor grade and initial animal studies suggest that BMDC are present in high-grade tumors only (Greenfield et al., 2009). In addition, TAM have been associated with poorer prognosis in metastatic lesions and other solid tumors (Wels et al., 2008). Thus histological stains aimed at identifying this population of cells may provide more accurate diagnosis and prognosis. Likewise, the molecular markers of this particular population of cells may offer an even more specific therapeutic target. Based on data collected in glioma patients, EPC can be identified by cell surface markers including CD133 and VEGFR2. Knocking down this population with specifically designed drug therapies has the potential for preventing recurrence by decreasing migration of these cells and reducing vasculogenesis within the tumor bed. Finally, one can also envision a role for advanced imaging technologies for improved diagnosis and treatment. For instance, PET has been used to specifically measure VEGF that has been labeled with copper in an orthotropic mouse model of GBM (Cai et al., 2006). If one could identify and label molecular targets that are specific to individual tumors subtypes and sensitive to new imaging techniques, this provides exciting non-invasive possibilities for tumor specific identification and treatment for each individual patient.

#### **9. Conclusions**

212 Glioma – Exploring Its Biology and Practical Relevance

VEGF, FGF and EGF stimulates local angiogenesis (Kioi, 2010). Secondly, hypoxia is an additional critical event in triggering the switch (Kioi, 2010). As tumor size grows and metabolic demand exceeds local perfusion, hypoxic conditions occur. Release of hypoxia inducible factor-1 (HIF-1α) by tumor cells or cells within the hypoxic tumor environment, combined with stromal cell-derived factor-1 (SDF-1) and CXCR-4 receptor activation, mobilizes BMDCs to the tumor site and promotes vasculogenesis in gliomas (Du et al., 2008,

In an attempt to further understand these processes in gliomas, Du et al. utilized an orthotopic model of GBM in mice to demonstrate recruitment of BMDC in gliomas (Du et al., 2008). Based on their results they theorize that hypoxia and the subsequent release of HIF-1α is the key event in tumor progression. Elevation of VEGF, and subsequent SDF-1 release and CXCR-4 receptor activation, mobilizes BMDC and recruits EPC and myeloid cells to the tumor. The net effect tips the balance to a pro-angiogenic state and neovascularization within the tumor bed. Kioi et al also further theorized that radiation treatment may exacerbate the vasculogenesis process and boost eventual tumor recurrence observed in current treatment regimens (Greenfield et al., 2010, Kioi, 2010). The endothelialmesenchymal transition and MSC have also been described in metastatic disease (Singh & Settleman). MSC exist within the brain and mobilize to the tumor site as well (Hata et al., Kang et al.). The exact roles of these particular BMDC remain elusive and require more

In our laboratory, we have begun to investigate the correlation of BMDC mobilization and tumor grade in gliomas (unpublished data.) We used a PDGF-driven mouse model of GBM within which tumors develop slowly from low-grade to high-grade. Low-grade tumors have a clear absence of neovascularization and BMDC are not present within these lesions. In high-grade tumors, however, we have observed a profound increase in larger, irregularly shaped, hemorrhagic vessels and a significant population of BMDC exists that is not observed in low grade tumors. In addition, these cells are located near newly forming blood vessels in the perivascular niche. We have also observed that BMDC are mobilized in the bone marrow and are elevated in the peripheral blood of tumor bearing animals versus controls. In addition, a significant difference in this population of cells in the blood exists between low-grade and high-grade animals. While much work is yet to be done before this process is fully elucidated, it appears that the presence of BMDC correlates with tumor grade and the process of neovascularization. Thus, BMDC have a potential role in the

angiogenic switch as tumors progress from low-grade to high-grade tumors.

Fig. 2. Bone marrow-derived cells in human glioma. A, Immunofluorescence in grade II astrocytoma shows normal blood vessels (red, VE Cadherin) and a paucity of CD11b+ myeloid suppressor cells (Green). B, GBM shows abnormal vessel formation and an influx of

Greenfield et al., 2010, Kioi, 2010).

study before they are fully delineated.

CD11b+ cells (unpublished data.)

In summary, one of the primary factors predicting outcome in patients with low-grade glioma is malignant progression to high-grade tumor and it is evident that the angiogenic switch is an important event in this process. Initial management often entails surgical resection while adjuvant therapy for low-grade gliomas remains a controversial topic. Tumor grade is determined by histological analysis of tumor specimens, but the molecular fingerprint of these tumors is now being analyzed more thoroughly and holds promise for more exciting targeted treatment options. In addition, distinct, but as yet undefined, populations of cells are recruited to the tumor site and participate in neovascularization and promote tumor growth and invasiveness. Therefore, this population may represent an important therapeutic target in combating these tumors. Since survival is directly correlated with tumor grade, preventing tumor progression is imperative. While BMDC certainly are

Bone Marrow-Derived Cells Support Malignant Transformation of Low-Grade Glioma 215

Cai, W., Chen, K., Mohamedali, K. A., Cao, Q., Gambhir, S. S., Rosenblum, M. G. & Chen, X.

Castillo, M. (1994). Contrast enhancement in primary tumors of the brain and spinal cord.

Chen, W., Silverman, D. H., Delaloye, S., Czernin, J., Kamdar, N., Pope, W., Satyamurthy,

Daumas-Duport, C., Tucker, M. L., Kolles, H., Cervera, P., Beuvon, F., Varlet, P., Udo, N.,

De Carli, E., Wang, X. & Puget, S. (2009). IDH1 and IDH2 mutations in gliomas. *N Engl J* 

Delbeke, D., Meyerowitz, C., Lapidus, R. L., Maciunas, R. J., Jennings, M. T., Moots, P. L. &

Derlon, J. M., Petit-Taboue, M. C., Chapon, F., Beaudouin, V., Noel, M. H., Creveuil, C.,

Dirks, P. B., Jay, V., Becker, L. E., Drake, J. M., Humphreys, R. P., Hoffman, H. J. & Rutka, J.

Du, R., Lu, K. V., Petritsch, C., Liu, P., Ganss, R., Passegue, E., Song, H., Vandenberg, S.,

Fayed, N. & Modrego, P. J. (2005). The contribution of magnetic resonance spectroscopy and

Fisher, P. G., Tihan, T., Goldthwaite, P. T., Wharam, M. D., Carson, B. S., Weingart, J. D.,

low-grade astrocytomas. *Pediatr Blood Cancer*, Vol. 51, No. 2, pp. (245-50) Fueger, B. J., Czernin, J., Cloughesy, T., Silverman, D. H., Geist, C. L., Walter, M. A.,

diagnosed and recurrent gliomas. *J Nucl Med*, Vol. 51, No. 10, pp. (1532-8) Furnari, F. B., Fenton, T., Bachoo, R. M., Mukasa, A., Stommel, J. M., Stegh, A., Hahn, W. C.,

(225-33)

Vol. 47, No. 12, pp. (2048-56)

*Med*, Vol. 47, No. 6, pp. (904-11)

No. 1, pp. (61-78)

Vol. 195, No. 1, pp. (47-52)

(276-87; discussion 287-8)

*Neuroimaging Clin N Am*, Vol. 4, No. 1, pp. (63-80)

*Med*, Vol. 360, No. 21, pp. (2248; author reply 2249)

childhood. *Neurosurgery*, Vol. 34, No. 1, pp. (68-78)

and invasion. *Cancer Cell*, Vol. 13, No. 3, pp. (206-20)

*Neurooncol*, Vol. 72, No. 3, pp. (261-5)

evaluation of common brain tumors. *AJNR Am J Neuroradiol*, Vol. 24, No. 2, pp.

(2006). PET of vascular endothelial growth factor receptor expression. *J Nucl Med*,

N., Schiepers, C. & Cloughesy, T. (2006). 18F-FDOPA PET imaging of brain tumors: comparison study with 18F-FDG PET and evaluation of diagnostic accuracy. *J Nucl* 

Koziak, M. & Chodkiewicz, J. P. (1997). Oligodendrogliomas. Part II: A new grading system based on morphological and imaging criteria. *J Neurooncol*, Vol. 34,

Kessler, R. M. (1995). Optimal cutoff levels of F-18 fluorodeoxyglucose uptake in the differentiation of low-grade from high-grade brain tumors with PET. *Radiology*,

Courtheoux, P. & Houtteville, J. P. (1997). The in vivo metabolic pattern of lowgrade brain gliomas: a positron emission tomographic study using 18Ffluorodeoxyglucose and 11C-L-methylmethionine. *Neurosurgery*, Vol. 40, No. 2, pp.

T. (1994). Development of anaplastic changes in low-grade astrocytomas of

Johnson, R. S., Werb, Z. & Bergers, G. (2008). HIF1alpha induces the recruitment of bone marrow-derived vascular modulatory cells to regulate tumor angiogenesis

echoplanar perfusion-weighted MRI in the initial assessment of brain tumours. *J* 

Repka, M. X., Cohen, K. J. & Burger, P. C. (2008). Outcome analysis of childhood

Schiepers, C., Nghiemphu, P., Lai, A., Phelps, M. E. & Chen, W. Correlation of 6- 18F-fluoro-L-dopa PET uptake with proliferation and tumor grade in newly

Ligon, K. L., Louis, D. N., Brennan, C., Chin, L., DePinho, R. A. & Cavenee, W. K.

not the only factor in progression of disease and neovascularization, blocking recruitment of these cells to tumors has been shown to reduce growth in animal models of other tumor types. Therefore, a greater understanding of this process may define a role for targeting this population of cells. In addition, BMDC exist in the periphery, and make for an easier therapeutic target than tumor cells within the blood brain barrier. Lastly, current management of tumor recurrence relies on serial imaging studies. Therefore, an effective and accurate biomarker capable of predicting progression of disease may allow for earlier detection and better treatment outcomes. This makes the case for monitoring BMDC in the periphery in addition to therapy aimed at this population of cells as potential adjuvant therapy in glioma.

#### **10. References**


not the only factor in progression of disease and neovascularization, blocking recruitment of these cells to tumors has been shown to reduce growth in animal models of other tumor types. Therefore, a greater understanding of this process may define a role for targeting this population of cells. In addition, BMDC exist in the periphery, and make for an easier therapeutic target than tumor cells within the blood brain barrier. Lastly, current management of tumor recurrence relies on serial imaging studies. Therefore, an effective and accurate biomarker capable of predicting progression of disease may allow for earlier detection and better treatment outcomes. This makes the case for monitoring BMDC in the periphery in addition to therapy aimed at this population of cells as potential adjuvant

Alimenti, A., Delavelle, J., Lazeyras, F., Yilmaz, H., Dietrich, P. Y., de Tribolet, N. & Lovblad,

Armstrong, G. T., Conklin, H. M., Huang, S., Srivastava, D., Sanford, R., Ellison, D. W.,

Astrakas, L. G., Zurakowski, D., Tzika, A. A., Zarifi, M. K., Anthony, D. C., De Girolami, U.,

Bauman, G., Lote, K., Larson, D., Stalpers, L., Leighton, C., Fisher, B., Wara, W., MacDonald,

Berger, M. S., Deliganis, A. V., Dobbins, J. & Keles, G. E. (1994). The effect of extent of

Bergers, G. & Benjamin, L. E. (2003). Tumorigenesis and the angiogenic switch. *Nat Rev* 

Bourne, T. D. & Schiff, D. (2010). Update on molecular findings, management and outcome

Boxerman, J. L., Schmainda, K. M. & Weisskoff, R. M. (2006). Relative cerebral blood volume

Broniscer, A., Baker, S. J., West, A. N., Fraser, M. M., Proko, E., Kocak, M., Dalton, J.,

Bulakbasi, N., Kocaoglu, M., Ors, F., Tayfun, C. & Ucoz, T. (2003). Combination of single-

in low-grade gliomas. *Nat Rev Neurol*, Vol. 6, No. 12, pp. (695-701)

grade glioma in children. *J Clin Oncol*, Vol. 25, No. 6, pp. (682-9)

gliomas. *Eur Neurol*, Vol. 58, No. 4, pp. (198-209)

glioma. *Neuro Oncol*, Vol. 13, No. 2, pp. (223-34)

*Cancer Res*, Vol. 10, No. 24, pp. (8220-8)

*Oncol Biol Phys*, Vol. 45, No. 4, pp. (923-9)

*Cancer*, Vol. 74, No. 6, pp. (1784-91)

*Cancer*, Vol. 3, No. 6, pp. (401-10)

No. 4, pp. (859-67)

K. O. (2007). Monovoxel 1H magnetic resonance spectroscopy in the progression of

Merchant, T. E., Hudson, M. M., Hoehn, M. E., Robison, L. L., Gajjar, A. & Morris, E. B. (2011). Survival and long-term health and cognitive outcomes after low-grade

Tarbell, N. J. & Black, P. M. (2004). Noninvasive magnetic resonance spectroscopic imaging biomarkers to predict the clinical grade of pediatric brain tumors. *Clin* 

D., Stitt, L. & Cairncross, J. G. (1999). Pretreatment factors predict overall survival for patients with low-grade glioma: a recursive partitioning analysis. *Int J Radiat* 

resection on recurrence in patients with low grade cerebral hemisphere gliomas.

maps corrected for contrast agent extravasation significantly correlate with glioma tumor grade, whereas uncorrected maps do not. *AJNR Am J Neuroradiol*, Vol. 27,

Zambetti, G. P., Ellison, D. W., Kun, L. E., Gajjar, A., Gilbertson, R. J. & Fuller, C. E. (2007). Clinical and molecular characteristics of malignant transformation of low-

voxel proton MR spectroscopy and apparent diffusion coefficient calculation in the

therapy in glioma.

**10. References** 

evaluation of common brain tumors. *AJNR Am J Neuroradiol*, Vol. 24, No. 2, pp. (225-33)


Bone Marrow-Derived Cells Support Malignant Transformation of Low-Grade Glioma 217

Ichimura, K., Pearson, D. M., Kocialkowski, S., Backlund, L. M., Chan, R., Jones, D. T. &

Isselbacher, K. J. (1972). Sugar and amino acid transport by cells in culture--differences between normal and malignant cells. *N Engl J Med*, Vol. 286, No. 17, pp. (929-33) Jaeckle, K. A., Decker, P. A., Ballman, K. V., Flynn, P. J., Giannini, C., Scheithauer, B. W.,

status, and heterodimerization. *Clin Cancer Res*, Vol. 13, No. 6, pp. (1911-25) Kang, S. G., Shinojima, N., Hossain, A., Gumin, J., Yong, R. L., Colman, H., Marini, F.,

stem cells from mouse brain. *Neurosurgery*, Vol. 67, No. 3, pp. (711-20) Kaplan, R. N., Riba, R. D., Zacharoulis, S., Bramley, A. H., Vincent, L., Costa, C., MacDonald,

initiate the pre-metastatic niche. *Nature*, Vol. 438, No. 7069, pp. (820-7) Kates, R., Atkinson, D. & Brant-Zawadzki, M. (1996). Fluid-attenuated inversion recovery

Kioi, M., et al. (2010). Inhibition of vasculogenesis, but not angiogenesis, prevents the

Kleihues, P., Soylemezoglu, F., Schauble, B., Scheithauer, B. W. & Burger, P. C. (1995).

Kondziolka, D., Lunsford, L. D. & Martinez, A. J. (1993). Unreliability of contemporary

Kono, K., Inoue, Y., Nakayama, K., Shakudo, M., Morino, M., Ohata, K., Wakasa, K. &

Law, M., Yang, S., Wang, H., Babb, J. S., Johnson, G., Cha, S., Knopp, E. A. & Zagzag, D.

Law, M., Yang, S., Babb, J. S., Knopp, E. A., Golfinos, J. G., Zagzag, D. & Johnson, G. (2004).

astrocytoma. *J Neurosurg*, Vol. 79, No. 4, pp. (533-6)

tumors. *AJNR Am J Neuroradiol*, Vol. 22, No. 6, pp. (1081-8)

MR imaging. *AJNR Am J Neuroradiol*, Vol. 24, No. 10, pp. (1989-98)

*Neurosurg*, Vol. 107, No. 3, pp. (174-80)

*Imaging*, Vol. 8, No. 6, pp. (389-96)

pp. (XXX-YYY)

21)

Collins, V. P. (2009). IDH1 mutations are present in the majority of common adult gliomas but rare in primary glioblastomas. *Neuro Oncol*, Vol. 11, No. 4, pp. (341-7) Inoue, T., Ogasawara, K., Beppu, T., Ogawa, A. & Kabasawa, H. (2005). Diffusion tensor

imaging for preoperative evaluation of tumor grade in gliomas. *Clin Neurol* 

Jenkins, R. B. & Buckner, J. C. (2010). Transformation of low grade glioma and correlation with outcome: an NCCTG database analysis. *J Neurooncol*, Vol. No. pp. Johns, T. G., Perera, R. M., Vernes, S. C., Vitali, A. A., Cao, D. X., Cavenee, W. K., Scott, A.

M. & Furnari, F. B. (2007). The efficacy of epidermal growth factor receptor-specific antibodies against glioma xenografts is influenced by receptor levels, activation

Andreeff, M. & Lang, F. F. Isolation and perivascular localization of mesenchymal

D. D., Jin, D. K., Shido, K., Kerns, S. A., Zhu, Z., Hicklin, D., Wu, Y., Port, J. L., Altorki, N., Port, E. R., Ruggero, D., Shmelkov, S. V., Jensen, K. K., Rafii, S. & Lyden, D. (2005). VEGFR1-positive haematopoietic bone marrow progenitors

(FLAIR): clinical prospectus of current and future applications. *Top Magn Reson* 

recurrencec of glioblastoma following irradiation in mice. *J Clin Invest*, Vol. 120, No.

Histopathology, classification, and grading of gliomas. *Glia*, Vol. 15, No. 3, pp. (211-

neurodiagnostic imaging in evaluating suspected adult supratentorial (low-grade)

Yamada, R. (2001). The role of diffusion-weighted imaging in patients with brain

(2003). Glioma grading: sensitivity, specificity, and predictive values of perfusion MR imaging and proton MR spectroscopic imaging compared with conventional

Comparison of cerebral blood volume and vascular permeability from dynamic

(2007). Malignant astrocytic glioma: genetics, biology, and paths to treatment. *Genes Dev*, Vol. 21, No. 21, pp. (2683-710)


Gauvain, K. M., McKinstry, R. C., Mukherjee, P., Perry, A., Neil, J. J., Kaufman, B. A. &

Giammarile, F., Cinotti, L. E., Jouvet, A., Ramackers, J. M., Saint Pierre, G., Thiesse, P.,

Gil-Robles, S. & Duffau, H. (2010). Surgical management of World Health Organization

Godard, S., Getz, G., Delorenzi, M., Farmer, P., Kobayashi, H., Desbaillets, I., Nozaki, M.,

Greenfield, J. P., Jin, D. K., Young, L. M., Christos, P. J., Abrey, L., Rafii, S. & Gutin, P. H.

Greenfield, J. P., Cobb, W. S. & Lyden, D. (2010). Resisting arrest: a switch from angiogenesis

Hartmann, C., Meyer, J., Balss, J., Capper, D., Mueller, W., Christians, A., Felsberg, J.,

Hata, N., Shinojima, N., Gumin, J., Yong, R., Marini, F., Andreeff, M. & Lang, F. F. Platelet-

Hegi, M. E., Diserens, A. C., Gorlia, T., Hamou, M. F., de Tribolet, N., Weller, M., Kros, J. M.,

tensor imaging. *AJR Am J Roentgenol*, Vol. 177, No. 2, pp. (449-54)

functional structures. *Neurosurg Focus*, Vol. 28, No. 2, pp. (E8)

*Dev*, Vol. 21, No. 21, pp. (2683-710)

*Neurooncol*, Vol. 68, No. 3, pp. (263-74)

*Neuropathol*, Vol. 118, No. 4, pp. (469-74)

*Neoplasia*, Vol. 12, No. 9, pp. (675-84)

No. 20, pp. (6613-25)

826-7)

(663-7)

(997-1003)

(2007). Malignant astrocytic glioma: genetics, biology, and paths to treatment. *Genes* 

Hayashi, R. J. (2001). Evaluating pediatric brain tumor cellularity with diffusion-

Jouanneau, E., Guyotat, J., Pelissou-Guyotat, I., Setiey, A., Honnorat, J., Le Bars, D. & Frappaz, D. (2004). High and low grade oligodendrogliomas (ODG): correlation of amino-acid and glucose uptakes using PET and histological classifications. *J* 

Grade II gliomas in eloquent areas: the necessity of preserving a margin around

Diserens, A. C., Hamou, M. F., Dietrich, P. Y., Regli, L., Janzer, R. C., Bucher, P., Stupp, R., de Tribolet, N., Domany, E. & Hegi, M. E. (2003). Classification of human astrocytic gliomas on the basis of gene expression: a correlated group of genes with angiogenic activity emerges as a strong predictor of subtypes. *Cancer Res*, Vol. 63,

(2009). Surrogate markers predict angiogenic potential and survival in patients with glioblastoma multiforme. *Neurosurgery*, Vol. 64, No. 5, pp. (819-26; discussion

to vasculogenesis in recurrent malignant gliomas. *J Clin Invest*, Vol. 120, No. 3, pp.

Wolter, M., Mawrin, C., Wick, W., Weller, M., Herold-Mende, C., Unterberg, A., Jeuken, J. W., Wesseling, P., Reifenberger, G. & von Deimling, A. (2009). Type and frequency of IDH1 and IDH2 mutations are related to astrocytic and oligodendroglial differentiation and age: a study of 1,010 diffuse gliomas. *Acta* 

derived growth factor BB mediates the tropism of human mesenchymal stem cells for malignant gliomas. *Neurosurgery*, Vol. 66, No. 1, pp. (144-56; discussion 156-7) Hatanpaa, K. J., Burma, S., Zhao, D. & Habib, A. A. Epidermal growth factor receptor in

glioma: signal transduction, neuropathology, imaging, and radioresistance.

Hainfellner, J. A., Mason, W., Mariani, L., Bromberg, J. E., Hau, P., Mirimanoff, R. O., Cairncross, J. G., Janzer, R. C. & Stupp, R. (2005). MGMT gene silencing and benefit from temozolomide in glioblastoma. *N Engl J Med*, Vol. 352, No. 10, pp.


Bone Marrow-Derived Cells Support Malignant Transformation of Low-Grade Glioma 219

Ogawa, T., Shishido, F., Kanno, I., Inugami, A., Fujita, H., Murakami, M., Shimosegawa, E.,

Okamoto, Y., Di Patre, P. L., Burkhard, C., Horstmann, S., Jourde, B., Fahey, M., Schuler, D.,

Parsons, D. W., Jones, S., Zhang, X., Lin, J. C., Leary, R. J., Angenendt, P., Mankoo, P., Carter,

Pelloski, C. E., Ballman, K. V., Furth, A. F., Zhang, L., Lin, E., Sulman, E. P., Bhat, K.,

Pfister, S., Janzarik, W. G., Remke, M., Ernst, A., Werft, W., Becker, N., Toedt, G., Wittmann,

Pollack, I. F., Claassen, D., al-Shboul, Q., Janosky, J. E. & Deutsch, M. (1995). Low-grade

Prabhu, V. C., Khaldi, A., Barton, K. P., Melian, E., Schneck, M. J., Primeau, M. J. & Lee, J. M.

Price, S. J. (2010). Advances in imaging low-grade gliomas. *Adv Tech Stand Neurosurg*, Vol.

Rafii, S. & Lyden, D. (2008). Cancer. A few to flip the angiogenic switch. *Science*, Vol. 319,

Rees, J., Watt, H., Jager, H. R., Benton, C., Tozer, D., Tofts, P. & Waldman, A. (2009).

early malignant transformation. *Eur J Radiol*, Vol. 72, No. 1, pp. (54-64) Ricci-Vitiani, L., Pallini, R., Biffoni, M., Todaro, M., Invernici, G., Cenci, T., Maira, G., Parati,

glioblastoma multiforme. *Science*, Vol. 321, No. 5897, pp. (1807-12)

astrocytomas. *J Clin Invest*, Vol. 118, No. 5, pp. (1739-49)

Vol. 82, No. 4, pp. (536-47)

4, pp. (1037-59)

35, No. pp. (1-34)

7325, pp. (824-8)

No. 5860, pp. (163-4)

methionine PET. *Radiology*, Vol. 186, No. 1, pp. (45-53)

*Neuropathol*, Vol. 108, No. 1, pp. (49-56)

94)

Ito, H., Hatazawa, J., Okudera, T. & et al. (1993). Cerebral glioma: evaluation with

Probst-Hensch, N. M., Yasargil, M. G., Yonekawa, Y., Lutolf, U. M., Kleihues, P. & Ohgaki, H. (2004). Population-based study on incidence, survival rates, and genetic alterations of low-grade diffuse astrocytomas and oligodendrogliomas. *Acta* 

H., Siu, I. M., Gallia, G. L., Olivi, A., McLendon, R., Rasheed, B. A., Keir, S., Nikolskaya, T., Nikolsky, Y., Busam, D. A., Tekleab, H., Diaz, L. A., Jr., Hartigan, J., Smith, D. R., Strausberg, R. L., Marie, S. K., Shinjo, S. M., Yan, H., Riggins, G. J., Bigner, D. D., Karchin, R., Papadopoulos, N., Parmigiani, G., Vogelstein, B., Velculescu, V. E. & Kinzler, K. W. (2008). An integrated genomic analysis of human

McDonald, J. M., Yung, W. K., Colman, H., Woo, S. Y., Heimberger, A. B., Suki, D., Prados, M. D., Chang, S. M., Barker, F. G., 2nd, Buckner, J. C., James, C. D. & Aldape, K. (2007). Epidermal growth factor receptor variant III status defines clinically distinct subtypes of glioblastoma. *J Clin Oncol*, Vol. 25, No. 16, pp. (2288-

A., Kratz, C., Olbrich, H., Ahmadi, R., Thieme, B., Joos, S., Radlwimmer, B., Kulozik, A., Pietsch, T., Herold-Mende, C., Gnekow, A., Reifenberger, G., Korshunov, A., Scheurlen, W., Omran, H. & Lichter, P. (2008). BRAF gene duplication constitutes a mechanism of MAPK pathway activation in low-grade

gliomas of the cerebral hemispheres in children: an analysis of 71 cases. *J Neurosurg*,

(2010). Management of diffuse low-grade cerebral gliomas. *Neurol Clin*, Vol. 28, No.

Volumes and growth rates of untreated adult low-grade gliomas indicate risk of

E. A., Stassi, G., Larocca, L. M. & De Maria, R. Tumour vascularization via endothelial differentiation of glioblastoma stem-like cells. *Nature*, Vol. 468, No.

susceptibility contrast-enhanced perfusion MR imaging with glioma grade. *AJNR Am J Neuroradiol*, Vol. 25, No. 5, pp. (746-55)


Louis, D. N., Ohgaki, H., Wiestler, O. D., Cavenee, W. K., Burger, P. C., Jouvet, A.,

the central nervous system. *Acta Neuropathol*, Vol. 114, No. 2, pp. (97-109) Lyden, D., Hattori, K., Dias, S., Costa, C., Blaikie, P., Butros, L., Chadburn, A., Heissig, B.,

*Am J Neuroradiol*, Vol. 25, No. 5, pp. (746-55)

No. 4, pp. (524-8)

(794-802)

(449-58)

No. 5, pp. (371-81)

Vol. 16, No. 4, pp. (329-34)

*Neurooncol*, Vol. 63, No. 3, pp. (233-45)

*Neurol*, Vol. 44, No. 2, pp. (137-44)

susceptibility contrast-enhanced perfusion MR imaging with glioma grade. *AJNR* 

Scheithauer, B. W. & Kleihues, P. (2007). The 2007 WHO classification of tumours of

Marks, W., Witte, L., Wu, Y., Hicklin, D., Zhu, Z., Hackett, N. R., Crystal, R. G., Moore, M. A., Hajjar, K. A., Manova, K., Benezra, R. & Rafii, S. (2001). Impaired recruitment of bone-marrow-derived endothelial and hematopoietic precursor cells blocks tumor angiogenesis and growth. *Nat Med*, Vol. 7, No. 11, pp. (1194-201) Mandonnet, E., Delattre, J. Y., Tanguy, M. L., Swanson, K. R., Carpentier, A. F., Duffau, H.,

Cornu, P., Van Effenterre, R., Alvord, E. C., Jr. & Capelle, L. (2003). Continuous growth of mean tumor diameter in a subset of grade II gliomas. *Ann Neurol*, Vol. 53,

(1995). Analysis of brain tumors using 1H magnetic resonance spectroscopy. *Surg* 

M. W., Dillon, W. P., Graves, E. E., Pirzkall, A. & Nelson, S. J. (2002). Histopathological validation of a three-dimensional magnetic resonance spectroscopy index as a predictor of tumor presence. *J Neurosurg*, Vol. 97, No. 4, pp.

Classification of biopsy-confirmed brain tumors using single-voxel MR

U. & Zanella, F. E. (2002). Clinical application of proton magnetic resonance spectroscopy in the diagnosis of intracranial mass lesions. *Neuroradiology*, Vol. 44,

P. R., Griffiths, J. R. & Bell, B. A. (2002). The contribution of proton magnetic resonance spectroscopy (1HMRS) to clinical brain tumour diagnosis. *Br J Neurosurg*,

Zanella, F. & Lanfermann, H. (2003). Preoperative proton-MR spectroscopy of gliomas--correlation with quantitative nuclear morphology in surgical specimen. *J* 

Heerschap, A., Kamada, K., Lee, B. C., Mengeot, M. M., Moser, E., Padavic-Shaller, K. A., Sanders, J. A., Spraggins, T. A., Stillman, A. E., Terwey, B., Vogl, T. J., Wicklow, K. & Zimmerman, R. A. (1996). Proton magnetic resonance spectroscopy in patients with glial tumors: a multicenter study. *J Neurosurg*, Vol. 84, No. 3, pp.

McBride, D. Q., Miller, B. L., Nikas, D. L., Buchthal, S., Chang, L., Chiang, F. & Booth, R. A.

McKnight, T. R., von dem Bussche, M. H., Vigneron, D. B., Lu, Y., Berger, M. S., McDermott,

Meyerand, M. E., Pipas, J. M., Mamourian, A., Tosteson, T. D. & Dunn, J. F. (1999).

Moller-Hartmann, W., Herminghaus, S., Krings, T., Marquardt, G., Lanfermann, H., Pilatus,

Murphy, M., Loosemore, A., Clifton, A. G., Howe, F. A., Tate, A. R., Cudlip, S. A., Wilkins,

Nafe, R., Herminghaus, S., Raab, P., Wagner, S., Pilatus, U., Schneider, B., Schlote, W.,

Nakamura, M., Yonekawa, Y., Kleihues, P. & Ohgaki, H. (2001). Promoter hypermethylation of the RB1 gene in glioblastomas. *Lab Invest*, Vol. 81, No. 1, pp. (77-82) Negendank, W. G., Sauter, R., Brown, T. R., Evelhoch, J. L., Falini, A., Gotsis, E. D.,

spectroscopy. *AJNR Am J Neuroradiol*, Vol. 20, No. 1, pp. (117-23)


Bone Marrow-Derived Cells Support Malignant Transformation of Low-Grade Glioma 221

vascularity of gliomas. *AJR Am J Roentgenol*, Vol. 171, No. 6, pp. (1479-86) Sugahara, T., Korogi, Y., Kochi, M., Ikushima, I., Shigematu, Y., Hirai, T., Okuda, T., Liang,

Sugahara, T., Korogi, Y., Kochi, M., Ushio, Y. & Takahashi, M. (2001). Perfusion-sensitive

Thiel, A., Pietrzyk, U., Sturm, V., Herholz, K., Hovels, M. & Schroder, R. (2000). Enhanced

Tihan, T., Fisher, P. G., Kepner, J. L., Godfraind, C., McComb, R. D., Goldthwaite, P. T. &

Tripathi, M., Sharma, R., D'Souza, M., Jaimini, A., Panwar, P., Varshney, R., Datta, A.,

Unal, E., Koksal, Y., Cimen, O., Paksoy, Y. & Tavli, L. (2008). Malignant glioblastomatous

Wang, R., Chadalavada, K., Wilshire, J., Kowalik, U., Hovinga, K. E., Geber, A., Fligelman,

Watanabe, T., Nobusawa, S., Kleihues, P. & Ohgaki, H. (2009). IDH1 mutations are early

Wels, J., Kaplan, R. N., Rafii, S. & Lyden, D. (2008). Migratory neighbors and distant invaders: tumor-associated niche cells. *Genes Dev*, Vol. 22, No. 5, pp. (559-74) Wen, P. Y. & Kesari, S. (2008). Malignant gliomas in adults. *N Engl J Med*, Vol. 359, No. 5, pp.

Wessels, P. H., Weber, W. E., Raven, G., Ramaekers, F. C., Hopman, A. H. & Twijnstra, A.

Yan, H., Parsons, D. W., Jin, G., McLendon, R., Rasheed, B. A., Yuan, W., Kos, I., Batinic-

tumour endothelium. *Nature*, Vol. 468, No. 7325, pp. (829-33)

gliomas. *J Magn Reson Imaging*, Vol. 9, No. 1, pp. (53-60)

4, pp. (210-5)

(1061-8)

83)

pp. (1385-9)

(492-507)

Vol. 174, No. 4, pp. (1149-53)

*Lancet Neurol*, Vol. 2, No. 7, pp. (395-403)

*Neurosurgery*, Vol. 46, No. 1, pp. (232-4)

cerebral blood volume maps with histologic and angiographic determination of

L., Ge, Y., Komohara, Y., Ushio, Y. & Takahashi, M. (1999). Usefulness of diffusionweighted MRI with echo-planar technique in the evaluation of cellularity in

MR imaging of gliomas: comparison between gradient-echo and spin-echo echoplanar imaging techniques. *AJNR Am J Neuroradiol*, Vol. 22, No. 7, pp. (1306-15) Tamura, M., Shibasaki, T., Zama, A., Kurihara, H., Horikoshi, S., Ono, N., Oriuchi, N. &

Hirano, T. (1998). Assessment of malignancy of glioma by positron emission tomography with 18F-fluorodeoxyglucose and single photon emission computed tomography with thallium-201 chloride. *Neuroradiology*, Vol. 40, No.

accuracy in differential diagnosis of radiation necrosis by positron emission tomography-magnetic resonance imaging coregistration: technical case report.

Burger, P. C. (1999). Pediatric astrocytomas with monomorphous pilomyxoid features and a less favorable outcome. *J Neuropathol Exp Neurol*, Vol. 58, No. 10, pp.

Kumar, N., Garg, G., Singh, D., Grover, R. K., Mishra, A. K. & Mondal, A. (2009). Comparative evaluation of F-18 FDOPA, F-18 FDG, and F-18 FLT-PET/CT for metabolic imaging of low grade gliomas. *Clin Nucl Med*, Vol. 34, No. 12, pp. (878-

transformation of a low-grade glioma in a child. *Childs Nerv Syst*, Vol. 24, No. 12,

B., Leversha, M., Brennan, C. & Tabar, V. Glioblastoma stem-like cells give rise to

events in the development of astrocytomas and oligodendrogliomas. *Am J Pathol*,

(2003). Supratentorial grade II astrocytoma: biological features and clinical course.

Haberle, I., Jones, S., Riggins, G. J., Friedman, H., Friedman, A., Reardon, D.,


Sanson, M., Marie, Y., Paris, S., Idbaih, A., Laffaire, J., Ducray, F., El Hallani, S., Boisselier, B.,

Sathornsumetee, S., Reardon, D. A., Desjardins, A., Quinn, J. A., Vredenburgh, J. J. & Rich, J.

Schiff, D., Brown, P. D. & Giannini, C. (2007). Outcome in adult low-grade glioma:

Schomas, D. A., Laack, N. N., Rao, R. D., Meyer, F. B., Shaw, E. G., O'Neill, B. P., Giannini,

Setzer, M., Herminghaus, S., Marquardt, G., Tews, D. S., Pilatus, U., Seifert, V., Zanella, F. &

Shaw, E., Arusell, R., Scheithauer, B., O'Fallon, J., O'Neill, B., Dinapoli, R., Nelson, D.,

Shaw, E. G. & Wisoff, J. H. (2003). Prospective clinical trials of intracranial low-grade glioma

Shaw, E. G., Berkey, B., Coons, S. W., Bullard, D., Brachman, D., Buckner, J. C., Stelzer, K. J.,

Shin, J. H., Lee, H. K., Kwun, B. D., Kim, J. S., Kang, W., Choi, C. G. & Suh, D. C. (2002).

Sijens, P. E. & Oudkerk, M. (2002). 1H chemical shift imaging characterization of human

Singh, A. & Settleman, J. EMT, cancer stem cells and drug resistance: an emerging axis of

Stieber, V. W. (2001). Low-grade gliomas. *Curr Treat Options Oncol*, Vol. 2, No. 6, pp. (495-

Sugahara, T., Korogi, Y., Kochi, M., Ikushima, I., Hirai, T., Okuda, T., Shigematsu, Y., Liang,

L., Ge, Y., Ushio, Y. & Takahashi, M. (1998). Correlation of MR imaging-determined

brain tumor and edema. *Eur Radiol*, Vol. 12, No. 8, pp. (2056-61)

evil in the war on cancer. *Oncogene*, Vol. 29, No. 34, pp. (4741-51)

Barger, G. R., Brown, P. D., Gilbert, M. R. & Mehta, M. (2008). Recurrence following neurosurgeon-determined gross-total resection of adult supratentorial low-grade glioma: results of a prospective clinical trial. *J Neurosurg*, Vol. 109, No. 5, pp. (835-

Using relative cerebral blood flow and volume to evaluate the histopathologic grade of cerebral gliomas: preliminary results. *AJR Am J Roentgenol*, Vol. 179, No. 3,

in adults and children. *Neuro Oncol*, Vol. 5, No. 3, pp. (153-60)

Vol. 27, No. 25, pp. (4150-4)

1, pp. (13-24)

(1366-73)

pp. (437-45)

pp. (2267-76)

86)

41)

506)

pp. (783-9)

Mokhtari, K., Hoang-Xuan, K. & Delattre, J. Y. (2009). Isocitrate dehydrogenase 1 codon 132 mutation is an important prognostic biomarker in gliomas. *J Clin Oncol*,

N. (2007). Molecularly targeted therapy for malignant glioma. *Cancer*, Vol. 110, No.

the impact of prognostic factors and treatment. *Neurology*, Vol. 69, No. 13, pp.

C. & Brown, P. D. (2009). Intracranial low-grade gliomas in adults: 30-year experience with long-term follow-up at Mayo Clinic. *Neuro Oncol*, Vol. 11, No. 4,

Lanfermann, H. (2007). Diagnostic impact of proton MR-spectroscopy versus image-guided stereotactic biopsy. *Acta Neurochir (Wien)*, Vol. 149, No. 4, pp. (379-

Earle, J., Jones, C., Cascino, T., Nichols, D., Ivnik, R., Hellman, R., Curran, W. & Abrams, R. (2002). Prospective randomized trial of low- versus high-dose radiation therapy in adults with supratentorial low-grade glioma: initial report of a North Central Cancer Treatment Group/Radiation Therapy Oncology Group/Eastern Cooperative Oncology Group study. *J Clin Oncol*, Vol. 20, No. 9, cerebral blood volume maps with histologic and angiographic determination of vascularity of gliomas. *AJR Am J Roentgenol*, Vol. 171, No. 6, pp. (1479-86)


**Part 3** 

**Glioma Progression** 

Herndon, J., Kinzler, K. W., Velculescu, V. E., Vogelstein, B. & Bigner, D. D. (2009). IDH1 and IDH2 mutations in gliomas. *N Engl J Med*, Vol. 360, No. 8, pp. (765-73)
