Many patients with glioma harbor specific mutations in the isocitrate dehydrogenase gene IDH1 that associate with a relatively better prognosis. IDH1-mutated tumors produce the oncometabolite 2-hydroxyglutarate. Because IDH1 also regulates several pathways leading to lipid synthesis, we hypothesized that IDH1-mutant tumors have an altered phospholipid metabolite profile that would impinge on tumor pathobiology. To investigate this hypothesis, we performed 31 P-MRS imaging in mouse xenograft models of four human gliomas, one of which harbored the IDH1-R132H mutation.31 P-MR spectra from the IDH1-mutant tumor displayed a pattern distinct from that of the three IDH1 wild-type tumors, characterized by decreased levels of phosphoethanolamine and increased levels of glycerophosphocholine. This spectral profile was confirmed by ex vivo analysis of tumor extracts, and it was also observed in human surgical biopsies of IDH1-mutated tumors by 31 P highresolution magic angle spinning spectroscopy. The specificity of this profile for the IDH1-R132H mutation was established by in vitro 31 P-NMR of extracts of cells overexpressing IDH1 or IDH1-R132H. Overall, our results provide evidence that the IDH1-R132H mutation alters phospholipid metabolism in gliomas involving phosphoethanolamine and glycerophosphocholine. These new noninvasive biomarkers can assist in the identification of the mutation and in research toward novel treatments that target aberrant metabolism in IDH1-mutant glioma. Cancer Res; 74(17); 4898-907. Ó2014 AACR.

Proper delineation of gliomas using contrast-enhanced magnetic resonance imaging (CE-MRI) poses a problem in neuro-oncology. The blood brain barrier (BBB) in areas of diffuse-infiltrative growth may be intact, precluding extravasation and subsequent MR-based detection of the contrast agent gadolinium diethylenetriaminepenta-acetic acid (Gd-DTPA). Treatment with antiangiogenic compounds may further complicate tumor detection as such compounds can restore the BBB in angiogenic regions. The increasing number of clinical trials with antiangiogenic compounds for treatment of gliomas calls for the development of alternative imaging modalities. Here we investigated whether CE-MRI using ultrasmall particles of iron oxide (USPIO, Sinerem 1 ) as blood pool contrast agent has additional value for detection of glioma in the brain of nude mice. We compared conventional T1-weighted Gd-DTPA-enhanced MRI to T2*-weighted USPIOenhanced MRI in mice carrying orthotopic U87 glioma, which were either or not treated with the antiangiogenic compound vandetanib (ZD6474, ZACTIMA TM ). In untreated animals, vessel leakage within the tumor and a relatively high tumor blood volume resulted in good MRI visibility with Gd-DTPA-and USPIOenhanced MRI, respectively. Consistent with previous findings, vandetanib treatment restored the BBB in the tumor vasculature, resulting in loss of tumor detectability in Gd-DTPA MRI. However, due to decreased blood volume, treated tumors could be readily detected in USPIO-enhanced MRI scans. Our findings suggest that Gd-DTPA MRI results in overestimation of the effect of antiangiogenic therapy of glioma and that USPIO-MRI provides an important complementary diagnostic tool to evaluate response to antiangiogenic therapy of these tumors. ' 2007 Wiley-Liss, Inc.

Diffuse intrinsic pontine glioma (DIPG) is a fatal malignancy because of its diffuse infiltrative growth pattern. Translational research suffers from the lack of a representative DIPG animal model. Hence, human E98 glioma cells were stereotactically injected into the pons of nude mice. The E98 DIPG tumors presented a strikingly similar histhopathology to autopsy material of a DIPG patient, including diffuse and perivascular growth, brainstem- and supratentorial invasiveness and leptomeningeal growth. Magnetic resonance imaging (MRI) was effectively employed to image the E98 DIPG tumor. [(18) F] 3'-deoxy-3'-[(18) F]fluorothymidine (FLT) positron emission tomography (PET) imaging was applied to assess the subcutaneous (s.c.) E98 tumor proliferation status but no orthotopic DIPG activity could be visualized. Next, E98 cells were cultured in vitro and engineered to express firefly luciferase and mCherry (E98-Fluc-mCherry). These cultured E98-Fluc-mCherry cells developed focal pontine glioma when injected into the pons directly. However, the diffuse E98 DIPG infiltrative phenotype was restored when cells were injected into the pons immediately after an intermediate s.c. passage. The diffuse E98-Fluc-mCherry model was subsequently used to test escalating doses of irradiation, applying the bioluminescent Fluc signal to monitor tumor recurrence over time. Altogether, we here describe an accurate DIPG mouse model that can be of clinical relevance for testing experimental therapeutics in vivo.

Currently available compounds that interfere with VEGF-A signalling effectively inhibit angiogenesis in gliomas, but influence diffuse infiltrative growth to a much lesser extent. Development of a functional tumour vascular bed not only involves VEGF-A but also requires platelet-derived growth factor receptor-β (PDGFRβ), which induces maturation of tumour blood vessels. Therefore, we tested whether combined inhibition of VEGFR and PDGFRβ increases therapeutic benefit in the orthotopic glioma xenograft models E98 and E473, both displaying the diffuse infiltrative growth that is characteristically observed in most human gliomas. We used bevacizumab and vandetanib as VEGF(R) inhibitors, and sunitinib to additionally target PDGFRβ. We show that combination therapy of sunitinib and vandetanib does not improve therapeutic efficacy compared to treatment with sunitinib, vandetanib or bevacizumab alone. Furthermore, all compounds induced reduction of vessel leakage in compact E98 tumour areas, resulting in decreased detectability of these mostly infiltrative xenografts in Gd-DTPA-enhanced MRI scans. These data show that inhibition of VEGF signalling cannot be optimized by additional PDGFR inhibition and support the concept that diffuse infiltrative areas in gliomas are resistant to anti-angiogenic therapy.

Purpose: To develop a transgenic mouse model of glioma that can be conveniently used for testing therapy intervention strategies. High-grade glioma is a devastating and uniformly fatal disease for which better therapy is urgently needed. Typical for high-grade glioma is that glioma cells infiltrate extensively into surrounding pivotal brain structures, thereby rendering current treatments largely ineffective. Evaluation of novel therapies requires the availability of appropriate glioma mouse models.Experimental Design: High-grade gliomas were induced by stereotactic intracranial injection of lentiviral GFAP-Cre or CMV-Cre vectors into compound LoxP-conditional mice, resulting in K-Ras v12 expression and loss of p16Ink4a /p19 Arf with or without concomitant loss of p53 or Pten.Results: Tumors reproduced many of the features that are characteristic for human high-grade gliomas, including invasiveness and blood-brain barrier functionality. Especially, CMV-Cre injection into p53; Ink4a/Arf;K-Ras v12 mice resulted in high-grade glioma with a short tumor latency (2-3 weeks) and full penetrance. Early detection and follow-up was accomplished by noninvasive bioluminescence imaging, and the practical utility for therapy intervention was shown in a study with temozolomide. Conclusion:We have developed a realistic high-grade glioma model that can be used with almost the same convenience as traditional xenograft models, thus allowing its implementation at the forefront of preclinical evaluation of new treatments. Clin Cancer Res; 16(13); 3431-41. ©2010 AACR.