**5.1.1 Electron Paramagnetic Resonance (EPR) oximetry**

EPR is a MR method that detects only species containing unpaired electrons (Gallez & Swartz, 2004a). One of the numerous applications of EPR is *in vivo* oximetry. Molecular oxygen is a triplet radical that possesses two unpaired electrons which are responsible for its paramagnetism. However, EPR is not able to detect oxygen itself when dissolved in fluids near room temperature: in biological systems, the output signal lines are so broadened as to be undetectable. Indirect methods exist. Most of these methods rely on the paramagnetic properties of molecular oxygen, which acts as an efficient relaxer for other paramagnetic species (Gallez et al., 2004b). The enhancement of relaxation rates scales linearly with the concentration of oxygen over a wide range of oxygen tensions. The lack of detectable levels of endogenous paramagnetic species makes it necessary to use exogenous paramagnetic materials. Variations in pO2 of less than 1mmHg can be detected using particulate materials. While EPR spectroscopy provides local measurements, EPR imaging techniques provide spatially resolved measurements of these materials. The spatial distribution of free radicals can be performed utilizing magnetic field gradients in a manner similar to that of MRI. Spectral–spatial EPR imaging encodes both the spatial distribution of the spin probe and the spectral information, which allows the mapping of molecular oxygen (Kuppusamy et al., 2003). For this purpose, the use of soluble EPR materials such as trityl radicals is more convenient as they can diffuse in the whole tissue.

EPR oximetry was compared with other methods that provide direct or indirect measurements of tumor oxygenation: with polarographic electrodes, the distribution of nitroimidazoles, the BOLD effect in MRI, and pO2 recordings using OxyLite (reviewed in Gallez et al., 2004b). Two major challenges are now considered for moving this technology into the clinic: (i) assuring biocompatibility of the oxygen sensors in humans and (ii) modifying the instruments so that they can be used for humans instead of small animals (Swartz et al., 2004).

### **5.1.2 19F relaxometry**

19F NMR spectroscopy and imaging of perfluorocarbon (PFC) emulsions (hydrocarbons with protons having been replaced with fluorine nuclei) has been extensively exploited to measure the oxygen tension of biological systems in preclinical studies. The 19F MR signal of the PFC is sensitive to the pO2 of the surrounding tumor tissue, and acts as an oximeter. The principle behind 19F MR oximetry relies on the linear increase of the NMR spin-lattice relaxation rate R1 (=1/T1) of PFC emulsions with increasing oxygen tension (Mason et al.

Targeting Tumor Perfusion and Oxygenation Modulates

(Tatum et al., 2006).

**5.2 Tumor perfusion measurements** 

**5.2.1 Dynamic Contrast Enhanced MRI** 

alteration in blood flow or tissue H2O content (O'Connor et al., 2009).

**5.1.5 Hypoxia assessed by Positron Emission Tomography (PET)** 

Hypoxia and Cancer Sensitivity to Radiotherapy and Systemic Therapies 301

water containing oxygen, due to the paramagnetic properties of oxygen. The measured change in R1 is, in theory, proportional to the change in tissue oxygen concentration (O'Connor et al., 2007, 2009). Studies have indeed demonstrated that oxygen-enhanced MRI produces measurable signal changes in normal tissues in patients and is feasible on conventional clinical scanners. Therefore, oxygen-induced increase in R1 has the potential to provide noninvasive measurements of change in tumor oxygen concentration, distinct from BOLD imaging. Nevertheless, the technique still lacks in sensitivity and the measured delta R1 may include errors resulting from changes independent of tissue oxygen content, such as

 18F-labeled fluoromisonidazole (18F-MISO) is probably the most widely used PET imaging agent for hypoxia. 18F-MISO accumulates in tissues by binding to intracellular macromolecules when pO2 *<* 10 mm Hg. Retention within tissues is dependent on nitroreductase activity (that is, on the reduction status of a NO2 group on the imidazole ring) and accumulation in hypoxic tissues over a range of blood flows has been observed (Tatum et al., 2006). 18F-MISO is only sensitive to the presence of hypoxia in viable cells: 18F-MISO is not retained in necrosis because the electron transport chain that reduces the nitroimidazole to a bioreductive alkylating agent is no longer active (Padhani et al., 2007). Nevertheless, 18F-MISO PET is able to monitor the changing hypoxia status of lung tumors during radiotherapy (Koh et al., 1995). Studies in sarcoma (Rajendran et al., 2003) and head and neck cancer (Rajendran et al., 2004) have demonstrated a correlation of 18F-MISO uptake with poor outcome to radiation and chemotherapy. Other 18F labeled nitroimidazoles are currently evaluated, including EF3 and fluoroazomycin arabinoside (FAZA), for example

Microvascular parameters such as permeability and perfusion are of particular interest in the context of the abnormal tumor microvascular network. Useful imaging systems have been developed to monitor angiogenesis and the microvasculature *in vivo*, including DCE-MRI (Choyke, et al. 2003), PET and Single Photon Emission Computed Tomography (SPECT), CT, Doppler ultrasound, and optical imaging methods (see Jennings, et al., 2008).

DCE-MRI consists in the acquisition of serial MR images before, during, and after the administration of an intravenous contrast agent (CA) to produce time series images that enable pixel-by-pixel analysis of contrast kinetics within a tumor. Pharmacokinetic models provide a means of summarizing contrast enhancement data in terms of parameters that relate to the underlying vascular anatomy and physiology. As described by Tofts et al. (1999), the essential features of a variety of models are covered by the generalized kinetic model. Most methods of analyzing dynamic contrast-enhanced T1-weighted data acquired with low molecular contrast medium use a compartmental analysis to obtain some combination of the three principal parameters: the transfer constant Ktrans in min-1 (volume transfer constant between blood plasma and ESS), the rate constant Kep in min-1 (rate constant between blood plasma and extravascular extracellular space [ESS]) and the volume of ESS per unit volume of tissue space, Vp (no unit). DCE-MRI has evolved from an

1996). 19F MR oximetry provides a sensitive measure of tissue oxygen tension and is a powerful approach for monitoring tumor hypoxia. Several PFCs have been used for NMR oximetry, but hexafluorobenzene (HFB) is preferred (Mason et al., 1996; Zhao, et al., 2004) because it has a six-fold symmetry with a single 19F NMR resonance, and a low sensitivity to temperature. Its spin lattice relaxation rate is highly sensitive to pO2 and exhibits a linear relationship across the entire range of tissue oxygenation.

Mason and colleagues have been successfully developing fluorocarbon relaxometry using echo planar imaging for dynamic oxygen mapping (FREDOM) MRI following direct intratumoral injection of the oxygen reporter molecule HFB. Our group further developed an MRI fluorocarbon oximetry technique using snapshot inversion recovery with an improved temporal resolution of 1.5 minutes (vs. 6.5 minutes for FREDOM) (Jordan et al., 2009). Our method therefore provides a rapid way to map tumor oxygenation and is particularly suitable to monitor acute changes of pO2 in tumors, including spontaneous fluctuations (Jordan et al., 2009; Magat et al., 2010). The translation to the clinic is currently limited by the lack of development of coils in the clinical setting, and the lack of characterization of PFCs in humans.

#### **5.1.3 Blood Oxygen Level Dependent (BOLD) MRI**

Functional MRI (fMRI) was first developed as an indirect method of imaging brain activity at high temporal resolution (Ogawa et al., 1990). The relative decrease in deoxyhemoglobin concentration, which has a paramagnetic effect, can be detected by MRI as a weak transient rise in the T2\* weighted signal. This is the BOLD contrast principle. Apart from its very large application in neuroscience, the use of BOLD contrast in tumors brought with it new challenges of understanding and interpretation. Since then, BOLD MRI has become a useful tool for addressing important questions regarding the pathophysiology of tumors. However, it has both advantages and disadvantages. One advantage of BOLD MRI is that it is noninvasive and can be used to monitor real time changes of tumor oxygenation during pharmacological treatments or to monitor spontaneous fluctuations in experimental tumors (Baudelet & Gallez 2002; Baudelet et al., 2004). It does not require externally administered contrast medium or radioactive isotopes, it can be repeated as necessary, and flow dependence can be decoupled. BOLD MRI, in combination with hypercapnia and hyperoxia, is also an attractive method for assessing maturation and the functional state of tumor blood vessels (Baudelet et al., 2006). As for disadvantages, BOLD MRI is unfortunately a nonquantitative method for monitoring tumor pO2. This is the result of the extreme sensitivity of changes in R2\* to the basal state of tumor oxygenation and blood volume fraction. The intra and intertumoral distribution of these parameters may be greatly heterogeneous, making it very difficult to compare estimated pO2 changes between two regions or individuals. Even more problematic is the fact that the change in R2\* is not always indicative of the change in pO2. Concomitant changes in blood volume, blood pH and metabolic status can lead to smaller-than-expected or even negative changes in R2\* (Baudelet & Gallez, 2002). Similarly, changes in oxygen consumption rate has been described to result in a lack of change in R2\* even though absolute pO2 is increased (Jordan et al., 2006b).

#### **5.1.4 Oxygen enhanced longitudinal relaxation MRI**

An alternative MRI technique for evaluating change in tumor oxygenation using endogenous contrast relies in the increase of the proton longitudinal relaxation rate (R1) of

1996). 19F MR oximetry provides a sensitive measure of tissue oxygen tension and is a powerful approach for monitoring tumor hypoxia. Several PFCs have been used for NMR oximetry, but hexafluorobenzene (HFB) is preferred (Mason et al., 1996; Zhao, et al., 2004) because it has a six-fold symmetry with a single 19F NMR resonance, and a low sensitivity to temperature. Its spin lattice relaxation rate is highly sensitive to pO2 and exhibits a linear

Mason and colleagues have been successfully developing fluorocarbon relaxometry using echo planar imaging for dynamic oxygen mapping (FREDOM) MRI following direct intratumoral injection of the oxygen reporter molecule HFB. Our group further developed an MRI fluorocarbon oximetry technique using snapshot inversion recovery with an improved temporal resolution of 1.5 minutes (vs. 6.5 minutes for FREDOM) (Jordan et al., 2009). Our method therefore provides a rapid way to map tumor oxygenation and is particularly suitable to monitor acute changes of pO2 in tumors, including spontaneous fluctuations (Jordan et al., 2009; Magat et al., 2010). The translation to the clinic is currently limited by the lack of development of coils in the clinical setting, and the lack of

Functional MRI (fMRI) was first developed as an indirect method of imaging brain activity at high temporal resolution (Ogawa et al., 1990). The relative decrease in deoxyhemoglobin concentration, which has a paramagnetic effect, can be detected by MRI as a weak transient rise in the T2\* weighted signal. This is the BOLD contrast principle. Apart from its very large application in neuroscience, the use of BOLD contrast in tumors brought with it new challenges of understanding and interpretation. Since then, BOLD MRI has become a useful tool for addressing important questions regarding the pathophysiology of tumors. However, it has both advantages and disadvantages. One advantage of BOLD MRI is that it is noninvasive and can be used to monitor real time changes of tumor oxygenation during pharmacological treatments or to monitor spontaneous fluctuations in experimental tumors (Baudelet & Gallez 2002; Baudelet et al., 2004). It does not require externally administered contrast medium or radioactive isotopes, it can be repeated as necessary, and flow dependence can be decoupled. BOLD MRI, in combination with hypercapnia and hyperoxia, is also an attractive method for assessing maturation and the functional state of tumor blood vessels (Baudelet et al., 2006). As for disadvantages, BOLD MRI is unfortunately a nonquantitative method for monitoring tumor pO2. This is the result of the extreme sensitivity of changes in R2\* to the basal state of tumor oxygenation and blood volume fraction. The intra and intertumoral distribution of these parameters may be greatly heterogeneous, making it very difficult to compare estimated pO2 changes between two regions or individuals. Even more problematic is the fact that the change in R2\* is not always indicative of the change in pO2. Concomitant changes in blood volume, blood pH and metabolic status can lead to smaller-than-expected or even negative changes in R2\* (Baudelet & Gallez, 2002). Similarly, changes in oxygen consumption rate has been described to result in a lack of

change in R2\* even though absolute pO2 is increased (Jordan et al., 2006b).

An alternative MRI technique for evaluating change in tumor oxygenation using endogenous contrast relies in the increase of the proton longitudinal relaxation rate (R1) of

**5.1.4 Oxygen enhanced longitudinal relaxation MRI** 

relationship across the entire range of tissue oxygenation.

**5.1.3 Blood Oxygen Level Dependent (BOLD) MRI** 

characterization of PFCs in humans.

water containing oxygen, due to the paramagnetic properties of oxygen. The measured change in R1 is, in theory, proportional to the change in tissue oxygen concentration (O'Connor et al., 2007, 2009). Studies have indeed demonstrated that oxygen-enhanced MRI produces measurable signal changes in normal tissues in patients and is feasible on conventional clinical scanners. Therefore, oxygen-induced increase in R1 has the potential to provide noninvasive measurements of change in tumor oxygen concentration, distinct from BOLD imaging. Nevertheless, the technique still lacks in sensitivity and the measured delta R1 may include errors resulting from changes independent of tissue oxygen content, such as alteration in blood flow or tissue H2O content (O'Connor et al., 2009).

### **5.1.5 Hypoxia assessed by Positron Emission Tomography (PET)**

 18F-labeled fluoromisonidazole (18F-MISO) is probably the most widely used PET imaging agent for hypoxia. 18F-MISO accumulates in tissues by binding to intracellular macromolecules when pO2 *<* 10 mm Hg. Retention within tissues is dependent on nitroreductase activity (that is, on the reduction status of a NO2 group on the imidazole ring) and accumulation in hypoxic tissues over a range of blood flows has been observed (Tatum et al., 2006). 18F-MISO is only sensitive to the presence of hypoxia in viable cells: 18F-MISO is not retained in necrosis because the electron transport chain that reduces the nitroimidazole to a bioreductive alkylating agent is no longer active (Padhani et al., 2007). Nevertheless, 18F-MISO PET is able to monitor the changing hypoxia status of lung tumors during radiotherapy (Koh et al., 1995). Studies in sarcoma (Rajendran et al., 2003) and head and neck cancer (Rajendran et al., 2004) have demonstrated a correlation of 18F-MISO uptake with poor outcome to radiation and chemotherapy. Other 18F labeled nitroimidazoles are currently evaluated, including EF3 and fluoroazomycin arabinoside (FAZA), for example (Tatum et al., 2006).

## **5.2 Tumor perfusion measurements**

Microvascular parameters such as permeability and perfusion are of particular interest in the context of the abnormal tumor microvascular network. Useful imaging systems have been developed to monitor angiogenesis and the microvasculature *in vivo*, including DCE-MRI (Choyke, et al. 2003), PET and Single Photon Emission Computed Tomography (SPECT), CT, Doppler ultrasound, and optical imaging methods (see Jennings, et al., 2008).
