1Introduction

Routine diagnostics and treatment monitoring of brain tumors is usually based on magnetic resonance imaging (MRI), but the capacity of conventional MRI to differentiate tumor tissue from nonspecific tissue changes may be limited especially after therapeutic interventions such as neurosurgical resection, radiotherapy, and chemotherapy. Molecular imaging using positron-emission tomography (PET) may provide relevant additional information on tumor metabolism, which allows for more accurate diagnostics especially in clinically equivocal situations. In the last decades, a variety of molecular targets have been addressed by specific PET tracers, but only a few have achieved relevance in routine clinical practice. This chapter is focussed on PET tracers that appear to be especially helpful in clinical decision-making with regard to a better delineation of brain tumors, prognosis, and grading, improved differentiation of tumor recurrence from nonspecific posttherapeutic changes, and treatment monitoring.

2Methods

Positron-emission tomography (PET) is based on the use of positron-emitting radionuclides that are incorporated either into substrates normally used by the human organism, such as glucose or amino acids, or into molecules that bind to receptors or participate in specific metabolic pathways. More than 70 different PET tracers have been explored in brain tumors in the last decades. Since it is beyond the scope of this chapter to address all of these tracers, this overview is focussed on those PET tracers or tracer groups that are promising in providing valuable clinical information on the basis of the current literature.

2.1 18F-2-Fluoro-2-Deoxy-D-Glucose

18F-2-fluoro-2-deoxy-D-glucose (FDG), which measures cellular glucose metabolism as a function of the enzyme hexokinase, is the most common clinically utilized PET tracer due to its high potential to detect tumors in the body based on increased energy demand of malignant tumors. In the brain, FDG exhibits high uptake in normal gray matter, reflecting the metabolic demands of neurons and glia. Regional cerebral glucose consumption can be calculated by measuring quantitative FDG uptake in brain, blood glucose concentration, an arterial input of the tracer, and a “lumped constant” which accounts for differences in enzyme affinity between FDG and glucose (Herholz et al. 2012). In clinical practice, however, FDG uptake is usually quantified by standard uptake values (SUV), which reflect regional FDG

uptake normalized to body weight and injected dose. In brain neoplasms, tumor-to-brain ratios using the mean or maximal tracer uptake in the tumor region divided by the mean uptake value in the contralateral brain is the preferred method. The high and regionally variable FDG uptake in normal brain parenchyma often makes the delineation of brain tumors difficult. Thus, the localization of brain tumors with FDG PET is difficult and only co-registration of FDG uptake images with MRI may allow the rating of glucose metabolism in specific areas of a tumor. There have been studies suggesting that additional delayed imaging at 180 min or later after tracer injection may increase the contrast between malignant tumors with high FDG uptake and normal brain, but the sensitivity to detect the extent of low-grade gliomas remains poor (Spence et al. 2004; Prieto et al. 2011). Another problem of FDG is the high tracer uptake in inflammatory cells. FDG accumulates in malignant tissue but also at the sites of infection and inflammation and in autoimmune and granulomatous diseases by the overexpression of distinct facultative glucose transporter (GLUT) isotypes (mainly GLUT-1 and GLUT-3) and by overproduction of glycolytic enzymes in cancer cells and inflammatory cells (Meller et al. 2007). Therefore, FDG PET is also used as a diagnostic method in fever of unknown origin (Meller et al. 2007).

2.2Radiolabeled Amino Acids

Besides FDG, radiolabeled amino acids are the most commonly used PET tracers for brain tumors. An advantage over FDG is the relatively low uptake of amino acids by normal brain tissue. Therefore, cerebral gliomas can be distinguished from the surrounding normal tissue with higher contrast compared with FDG. Many natural amino acids and their synthetic analogs have been labeled by positron-emit- ting isotopes and explored as tumor imaging agents (Jager et al. 2001; Crippa et al. 2012; Huang and McConathy 2013). Most PET studies of cerebral gliomas have been performed with the amino acid [11C]methyl-L-methionine (MET) (Singhal et al. 2008; Crippa et al. 2012), although the short half-life of 11C (20 min) limits the use of this tracer to the few centers that are equipped with an in-house cyclotron facility. The increasing use of 18F-labeled amino acids (half-life, 109 min) such as O-(2-18F-fluoroethyl)-L-tyrosine (FET) or 3,4-dihydroxy-6-[18F]fluoro-L-phenylalanine (FDOPA) will probably replace MET in the future (Becherer et al. 2003; Langen et al. 2006, 2008; Chen et al. 2008; Herholz et al. 2012; Walter et al. 2012).

The increased uptake of amino acids such as MET, FET, or FDOPA by cerebral glioma tissue appears to be caused almost entirely by increased transport via specific amino acid transporters, namely, transport system L for large neutral amino acids (Jager et al. 2001; Bergmann et al. 2004;

Langen and Broer 2004; Langen et al. 2006; Huang and McConathy 2013). MET also shows some incorporation into proteins and participation in other metabolic pathways (Singhal et al. 2008), but comparative studies between MET, FET, and FDOPA have shown that imaging of cerebral gliomas is very similar with these amino acids (Weber et al. 2000; Becherer et al. 2003; Langen et al. 2003; Grosu et al. 2011). Therefore, the participation of MET in other metabolic pathways than transport appears to be of minor importance and the clinical results obtained with the different tracers can be considered together. Since large neutral amino acids also enter normal brain tissue, a disruption of the blood-brain barrier (BBB), i.e., enhancement of contrast agent in CT or MRI scans, is not a prerequisite for intratumoral accumulation of these amino acids. Consequently, uptake of the tracers has been reported in many low-grade gliomas without BBB leakage (Herholz et al. 1998; Ribom et al. 2001; Floeth et al. 2007; Kunz et al. 2011; Smits and Baumert 2011; Rapp et al. 2013a). The sensitivity and specificity of PET using MET and FET to differentiate between gliomas and nonneoplastic lesions is within the range of 60–90 % (Herholz et al. 1998; Pichler et al. 2010; Dunet et al. 2012; Rapp et al. 2013b), and the possibility of nonspecific enhancement in inflammatory cells or reactive glial tissue must be kept in mind. There have been reports of perifocal MET and FET uptake around hematomas and areas of ischemia, as well as of rare cases of uptake in or around ring-enhanc- ing lesion like brain abscesses and acute inflammatory demyelination (Delbeke et al. 1995; Floeth et al. 2006; Singhal et al. 2008; Hutterer et al. 2013).

2.3Radiolabeled Nucleoside Analogs

Another approach in molecular imaging of brain tumors is the use of radiolabeled nucleoside analogs such as [18F]3′-deoxy- 3′-fluorothymidine (FLT) (Shields et al. 1998; Shields 2003). Once FLT is transported into the cell, it is phosphorylated by thymidine kinase (TK-1) and trapped inside the cell (Bading and Shields 2008). TK-1 is a cytosolic enzyme that is expressed during the DNA synthesis stage of the cell cycle. Compared to normal proliferating tissues, tumor cells have increased levels of TK-1, resulting in increased FLT uptake (Shields et al. 1998). A high rate of cellular proliferation is a key feature of malignant tumors, and proliferation markers (e.g., Ki-67) have shown a better correlation with the grade of malignancy and prognosis of cerebral gliomas than FDG uptake (Chen et al. 2005). Uptake of FLT, however, depends on BBB damage because transport across the normal BBB is slow (Chen et al. 2005; Jacobs et al. 2005). Therefore, this method does not delineate tumor parts with intact BBB (e.g., in low-grade gliomas) and is less suited to depict the full extent of cerebral gliomas.

2.4Imaging of Hypoxia

Furthermore, imaging of hypoxia is an interesting approach to explore the metabolic features in brain tumors. Hypoxia plays a critical role in tumor development and aggressiveness and is an important prognostic factor for resistance to antineoplastic treatments (Langen and Eschmann 2008). A number of hypoxia tracers are available for PET, of which 18F-fluoromisonidazole (FMISO) today is the most frequently studied tracer (Lee and Scott 2007). FMISO enters cells by passive diffusion, where it is reduced by nitroreductase enzymes to become trapped in cells with reduced tissue oxygen partial pressure. When oxygen is abundant in normally oxygenated cells, the parent compound is quickly regenerated by reoxidation and metabolites do not accumulate. However, in hypoxic cells, the low oxygen partial pressure prevents reoxidation of FMISO metabolites, resulting in tracer accumulation in hypoxic cells. Because FMISO only accumulates in hypoxic cells with functional nitroreductase enzymes, FMISO only accumulates in viable cells but not in dead necrotic cells (Lee and Scott 2007).

2.5Imaging Angiogenesis

Another target of growing interest in molecular imaging is angiogenesis. One target structure is the αvβ3-integrin receptor, which is highly expressed on activated endothelial cells during angiogenesis. Various ligands based on the tripeptide RGD (Arg-Gly-Asp), which binds with high affinity to the αvβ3-integrin receptor, have been developed for PET (Haubner et al. 2010). The glycosylated cyclic pentapeptide 18F-galacto-RGD has been successfully applied in patients with malignant gliomas, but studies on the clinical relevance of this approach for treatment planning are still scarce (Schnell et al. 2009).

2.6Somatostatin Receptors

Moreover, somatostatin receptors have been used as a target for molecular imaging of brain tumors, especially in meningiomas. Meningiomas demonstrate expression of a variety of receptors, including somatostatin receptor subtype 2 (SSTR2). The SSTR2 receptor ligand 68Ga-DOTATOC demonstrates high-resolution imaging and high tumorbackground contrast in meningiomas and may provide valuable additional information on the extent of meningiomas beneath osseous structures, especially at the skull base (Henze et al. 2005; Gehler et al. 2009; Nyuyki et al. 2010; Graf et al. 2013). Somatostatin receptors are also present in childhood tumors, especially in medulloblastomas.