**2. Histological classification of gliomas. Implications in MRI diagnosis**

Glioma is a non-specific term, broadly suggesting that the observed tumor originates from glial cells. These tumors account for more than 85% of primary brain neoplasms classified histologically, immunohistochemically and structurally as astrocytomas, ologodendrogliomas or oligoastrocytomas, depending on their morphological, immunochemical or molecular features. Gliomas are assigned a grade depending on their histolopathological properties (figure 1). The grade provides insight into the growth rate of the tumor. According to the classification of the World Health Organization (WHO), tumors in grade I depict the lowest malignancy, their cells look nearly normal and the tumor grows slowly. Grade II tumors are also slow growing, but show a slightly abnormal microscopic appearance. Grade III tumors are malignant by definition. Their cells look very different from the original neural cells and are actively proliferating. Grade IV is the most malignant and aggressive, defined by galloping genetic instability, complete morphological and metabolic reprogramming, fast and uncontrolled proliferation, intense resistance to apoptosis, diffuse infiltration, robust angiogenesis and propensity to necrosis. From this group Glioblastoma multiforme (GBM) is the most frequent, aggressive and lethal intracranial tumor, reaching approximately 50% of all astrocytomas (Furnari, Fenton et al. 2007). Despite decades of intensive research, high-grade gliomas (WHO grade III and IV) are currently estimated incurable with a poor or very poor survival, up to one year in glioblastoma and from two to three years in anaplastic astrocytomas (Reardon, Rich et al. 2006).

The tumor grade determines the most appropriate chemotherapeutic or surgical treatment, but histology, requiring a tumor biopsy, remains currently the only method able to yield

in brain oncology, particularly in neovascular imaging (Calli, Kitis et al. 2006). The diffusion approaches investigate the thermally induced random molecular motions of water molecules in tissues. The perfusion phenomenon describes and quantifies the microvascular blood flow which feeds a volume element of an organ or tissue. At a first glance these two phenomena seem to be very different, however a closer look identifies they both concerned with underlying molecular motions. Moreover, the random distribution of capillaries in tissues such as brain, provides tissue perfusion with some of the characteristic features of diffusive

This chapter will summarize the arsenal of MRI approaches providing functional information on gliomas based on diffusion and perfusion MRI methods. We begin with a description of the pathological classification of gliomas and its implications for MRI diagnosis, continuing with the applications of diffusion weighted imaging (DWI), diffusion tensor imaging (DTI) and tractography to the characterization of these brain tumors. We conclude showing the MRI methodologies available for the evaluation of cerebral and tumoral perfusion, addressing the use of Dynamic Susceptibility Contrast (DSC), Dynamic Contrast Enhanced (DCE) MR images and Arterial Spin Labelling (ASL). Our review focuses mainly in the clinical applications of these methodologies, providing a brief introduction into the physical principles of each one. For interested readers, the following review articles provide more detailed descriptions of the corresponding physical principles of diffusion weighted (DW) MRI (Mori and Barker 1999), DTI (Basser and Jones 2002) and Tractography (Mori and van Zijl 2002) or the different approaches to investigate microvascular perfusion

(Ostergaard 2005; Detre, Wang et al. 2009; Larsson, Courivaud et al. 2009).

astrocytomas (Reardon, Rich et al. 2006).

**2. Histological classification of gliomas. Implications in MRI diagnosis** 

Glioma is a non-specific term, broadly suggesting that the observed tumor originates from glial cells. These tumors account for more than 85% of primary brain neoplasms classified histologically, immunohistochemically and structurally as astrocytomas, ologodendrogliomas or oligoastrocytomas, depending on their morphological, immunochemical or molecular features. Gliomas are assigned a grade depending on their histolopathological properties (figure 1). The grade provides insight into the growth rate of the tumor. According to the classification of the World Health Organization (WHO), tumors in grade I depict the lowest malignancy, their cells look nearly normal and the tumor grows slowly. Grade II tumors are also slow growing, but show a slightly abnormal microscopic appearance. Grade III tumors are malignant by definition. Their cells look very different from the original neural cells and are actively proliferating. Grade IV is the most malignant and aggressive, defined by galloping genetic instability, complete morphological and metabolic reprogramming, fast and uncontrolled proliferation, intense resistance to apoptosis, diffuse infiltration, robust angiogenesis and propensity to necrosis. From this group Glioblastoma multiforme (GBM) is the most frequent, aggressive and lethal intracranial tumor, reaching approximately 50% of all astrocytomas (Furnari, Fenton et al. 2007). Despite decades of intensive research, high-grade gliomas (WHO grade III and IV) are currently estimated incurable with a poor or very poor survival, up to one year in glioblastoma and from two to three years in anaplastic

The tumor grade determines the most appropriate chemotherapeutic or surgical treatment, but histology, requiring a tumor biopsy, remains currently the only method able to yield

motion.

unambiguously this information. In spite of this, many of the histopathological features of gliomas are amenable to MRI explorations. In particular, the growth rate and the appearance of necrotic zones may be followed by sequential morphological MRI examinations, the cellularity and tumor microstucture investigated by diffusion weighted MRI, the compression effects of the tumor on surrounding neuronal tracts of the white matter may be examined by MRI tractography from DTI approaches and the development of the neoangiogenic vasculature characterized by perfusion imaging. These properties entail the MRI method with unprecedented capabilities to asses non invasively glioma grade and its potential repercussions. In the following sections we shall provide a description of main physical principles underlying these methodologies and provide illustrative examples on their applications.

Fig. 1. Histopathological features from gliomas of increasing grade as revealed by haematoxylin/eosin staning. (a) WHO grade II astrocitoma , (b) WHO grade III anaplastic astrocytoma and (c) WHO grade IV glioblastoma multiforme
