**6. MRI evaluation of therapeutics against gliomas**

Clinically, therapeutic response to surgical resection of gliomas, followed by radiation and chemotherapy, can be assessed by dynamic contrast-enhanced morphological MRI, increases in ADC values detected by DWI (Waerzeggers *et al*., 2010), decreases in the fractional tumor volume with a corresponding low relative cerebral blood volume detected by perfusion imaging, and/or reduced choline levels detected by MRS (Waldman *et al*., 2009). DCE-MRI was used to establish reduced Gd enhancement consistent with decreased vascular permeability following i.v. bevacizumab and carboplatin therapy in a human glioma (UW28) nude rat model (Jahnke *et al*., 2009). DCE-MRI using a high molecular weight contrast agent, albumin-Gd-DTPA, showed significantly increased Ktrans at the rim of

Assessment of Rodent Glioma Models Using Magnetic Resonance Imaging Techniques 265

ADC values have been found to increase particularly in the early phase of anticancer therapies (Waerzeggers *et al*., 2010). Increases in ADC were found to be a time and dose sensitive marker of tumor (mouse xenografts) response to radiation therapy (Larocque *et al*., 2009). Contrast-enhanced MRI and DWI were used to characterize the vascular and cellular responses of GL261 and U87 gliomas to the tumor-vascular disrupting agent (VDA) 5,6 dimethylxanthenone-4-acetic acid (DMXAA), which indicated significantly increased ADC values and the accumulation of contrast agent in treated tumors (Seshadri and Ciesielski, 2009). ADC and 3D T2\*-weighted MRI measurements were used to validate ZD6474 (tyrosine kinase receptor inhibitor) inhibition on tumor growth and angiogenesis in

In a gene therapy-induced apoptosis (ganciclovir-treated herpes simplex thymidine kinase (HSV-tk) gene-transfected BT4C gliomas) study, combined DWI and 1H-MRS assessment was used to find interconnecting trends following therapeutic response in water diffusion and water-referenced concentrations of mobile lipids (Liimatainen *et al*., 2009). It is thought that apoptosis leads to an increase in 1H-MRS detectable mobile cholesterol compounds and unsaturated lipids resulting from the gene therapy-induced apoptosis (Hakumäki et al., 1999; Liimatainen *et al*., 2006, 2009). Amide proton transfer (APT) MRI was recently used to differentiate between different glioma models (SF188/V+ glioma and 9L gliosarcoma) and radiation-induced necrosis, where viable glioma tissue was hyperintense and radiation

Iron oxide-based nanoparticles have recently been used as cell tracking agents, or nontargeted or targeted drug delivery. Magnetically-labeled cytotoxic T-cells were used as cellular probes and tracked by T2- and T2\*-weighted MRI to differentiate glioma tissue from focal radiation necrosis in U-251 glioma-bearing rodents (Arbab *et al*., 2010). Focused ultrasound, which was used to permeabilize the blood-brain barrier and increase passive diffusion, was found to increase the delivery of drug (1,3-bis(2-chloroethyl)-1-nitrosourea and iron oxide nanoparticles that can be monitored with MRI, in a rat C6 glioma model (Chen *et al*., 2010). EGFRvIII antibody-conjugated iron oxide nanoparticles were used for convection-enhanced delivery and targeted therapy in glioblastoma mouse xenografts

(U87DeltaEGFRvIII), and assessed by T2-weighted MRI (Hadjipanayis *et al*., 2010).

There are several orthotopic rodent glioma models that have been used for several decades, and more recent transgenic, orthotopic xenograft neurosphere- or PDGFB-expressing virusinduced models that better reflect the genetic and stem-cell involvement in glia tumorigenesis. It is important that appropriate glioma models are used that best represent our current knowledge of malignant glioblastomas in humans. Ideally the more recent models should be used if possible, however if an orthotopic syngeneic model is required, then the rat F98 or RG2, and mouse GL26(1) models seem to have some characteristics that resemble aspects of human glioblastomas, such as vascular proliferation, and aggressive and infiltrative tumor growth. The human U87 MG glioma cell xenograft model in athymic rodents also has beneficial characteristics resembling some aspects of human GBM. Choosing an appropriate model is particularly important when evaluating new anti-glioma therapies, as these models need to consider recurrent gliomas, possibly derived from cancer stem cells, which are radiation- and chemotherapy-resistant, and currently reflect the poor

EGFRvIII-expressing GBM8 gliomas (Yiin *et al*., 2010).

necrosis was hypointense to isointense (Zhou *et al*., 2011).

**7. Conclusions** 

prognosis of high-grade gliomas in humans.

a VEGFR tyrosine kinase inhibitor (Vetanalib, PTK787) (anti-angiogenic) treated U251 gliomas in rats (Ali *et al*., 2010). Low-molecular-weight (Gd-DOTA; gadoterate meglumine) and macromolecular (P846, 3.5 kDa) MR contrast-enhanced imaging was used to assess the therapeutic effect of an anti-angiogenic compound, sorafenic, and microbeam radiation therapy in a 9L gliosarcoma model, finding that anti-angiogenic therapy decreased tumor vessel permeability to the macromolecular contrast agent (Lemasson *et al*., 2010). Dynamic perfusion MRI using iron oxide nanoparticles (ferumoxytol) was used to assess the vascular effects of an anti-angiogenic agent versus corticosteroid (dexamethasone) treatment in a U87MG human glioma model in athymic rats, which found that bevacizumab significantly decreased the tumor blood volume and decreased permeability as determined by an increased time-to-peak enhancement (Varallyay *et al*., 2009).

Morphological MRI, MR angiography and perfusion imaging were used to assess the therapeutic efficacy of nitrone-based compounds as anti-glioma agents in a rat C6 glioma model. It was demonstrated that the nitrone, α-phenyl-*tert*-butyl nitrone (PBN) was able to prevent and/or decrease tumor volumes (by ~60-fold, with significance, p<0.001), increase animal survival (>90%), and decrease total tumor blood volumes (by ~20%), in comparison to non-treated rats bearing C6 gliomas (Doblas *et al*., 2008). Another cohort of rats were intracerebrally implanted with C6 glioma cells, monitored for tumor growth, and when tumors reached a volume of ~50 mm3 (approximately at 15 days post-intracerebral implantation of C6 glioma cells), PBN was administered (drinking water, 0.065% w/v) for a period of 25 days (Doblas *et al*., 2008). In the post-tumor treatment group, PBN was found to increase survival (40% of the treated rats, p<0.05), and decrease tumor volumes by ~2 fold, but was found to be non-significant (Doblas *et al*., 2008). Regarding post-tumor treatment, PBN was also found to not significantly affect blood tumor volumes, compared to non-treated rats (Doblas *et al*., 2008). It was concluded from these studies that PBN, when administered prophylactically, may have an effect on angiogenesis (Doblas *et al*., 2008).

Conversely, rats post-tumor treated with a PBN-derivative, OKN-007, were found to have significantly decreased tumor volumes (~3-fold, p<0.05), decreased the apparent diffusion coefficients (ADC) (~20%, p<0.05), and increased tissue perfusion rates (~60%, p<0.05) in tumors, compared to non-treated rats (Garteiser et al., 2010). OKN-007 was administered in the drinking water at 10 mg/kg/day starting when tumors had reached ~50 mm3 in volume (about day 15 following intracerebral implantation of rat C6 glioma cells), and continued for a total of 10 days (Garteiser *et al*., 2010). One group of rats was euthanized after the 10 day treatment period, and a second group was monitored for an additional 25 days following the treatment period (Garteiser *et al*., 2010). In the cohort of animals that were treated for 10 days and then euthanized, percent survival was 100% (p<0.0001), whereas for the rats that were monitored for an additional 25 days the percent survival was greater than 80% (p<0.001) (Garteiser *et al*., 2010). Morphological MRI was used to calculate tumor volumes; diffusion-weighted imaging (DWI) was used to measure ADC, which assesses changes in water diffusion due to tissue structural alterations; and perfusion-weighted MRI (pMRI) was used to characterize tissue perfusion rates, which can provide information on alterations in the vascular capillary bed. Currently, the known pharmacological effects of the nitrones are primarily anti-inflammatory in nature. The parent nitrone compound, PBN, is known to inhibit (1) cyclooxygenase-2 (COX-2), (2) inducible nitric oxide synthase (iNOS), and (3) nuclear factor kappaB (NF-κB) (Floyd *et al*., 2008).

ADC values have been found to increase particularly in the early phase of anticancer therapies (Waerzeggers *et al*., 2010). Increases in ADC were found to be a time and dose sensitive marker of tumor (mouse xenografts) response to radiation therapy (Larocque *et al*., 2009). Contrast-enhanced MRI and DWI were used to characterize the vascular and cellular responses of GL261 and U87 gliomas to the tumor-vascular disrupting agent (VDA) 5,6 dimethylxanthenone-4-acetic acid (DMXAA), which indicated significantly increased ADC values and the accumulation of contrast agent in treated tumors (Seshadri and Ciesielski, 2009). ADC and 3D T2\*-weighted MRI measurements were used to validate ZD6474 (tyrosine kinase receptor inhibitor) inhibition on tumor growth and angiogenesis in EGFRvIII-expressing GBM8 gliomas (Yiin *et al*., 2010).

In a gene therapy-induced apoptosis (ganciclovir-treated herpes simplex thymidine kinase (HSV-tk) gene-transfected BT4C gliomas) study, combined DWI and 1H-MRS assessment was used to find interconnecting trends following therapeutic response in water diffusion and water-referenced concentrations of mobile lipids (Liimatainen *et al*., 2009). It is thought that apoptosis leads to an increase in 1H-MRS detectable mobile cholesterol compounds and unsaturated lipids resulting from the gene therapy-induced apoptosis (Hakumäki et al., 1999; Liimatainen *et al*., 2006, 2009). Amide proton transfer (APT) MRI was recently used to differentiate between different glioma models (SF188/V+ glioma and 9L gliosarcoma) and radiation-induced necrosis, where viable glioma tissue was hyperintense and radiation necrosis was hypointense to isointense (Zhou *et al*., 2011).

Iron oxide-based nanoparticles have recently been used as cell tracking agents, or nontargeted or targeted drug delivery. Magnetically-labeled cytotoxic T-cells were used as cellular probes and tracked by T2- and T2\*-weighted MRI to differentiate glioma tissue from focal radiation necrosis in U-251 glioma-bearing rodents (Arbab *et al*., 2010). Focused ultrasound, which was used to permeabilize the blood-brain barrier and increase passive diffusion, was found to increase the delivery of drug (1,3-bis(2-chloroethyl)-1-nitrosourea and iron oxide nanoparticles that can be monitored with MRI, in a rat C6 glioma model (Chen *et al*., 2010). EGFRvIII antibody-conjugated iron oxide nanoparticles were used for convection-enhanced delivery and targeted therapy in glioblastoma mouse xenografts (U87DeltaEGFRvIII), and assessed by T2-weighted MRI (Hadjipanayis *et al*., 2010).
