Avastin alone or combined to Campto® reduces local Blood Oxygen Saturation in an orthotopic human glioblastoma model (U87-MG) in nude rats

Introduction: Despite aggressive surgery, radiotherapy and chemotherapy, malignant gliomas remain fatal. As these tumors are highly angiogenic, therapies directed against tumor vasculature or preventing angiogenesis have been developed. Monitoring changes in structural and functional microvasculature should help to evaluate the efficiency of these therapies. Previous results suggest that MRI follow-up of early modifications of microvascular parameters (blood volume fraction (BVf), vessel size index (VSI) and blood brain barrier permeability to a contrast agent (BBBperm.)) permits monitoring the effects of anti-angiogenic treatment on gliomas (1). Recently, it has been shown that local Blood Oxygen Saturation (lSO2), a possible indication of tumor oxygenation status, could be measured by MRI (2). Antioangenic therapies should alter the tumor oxygenation status. In this study, structural (BVf and VSI) and functional (BBBperm. and lSO2) microvascular parameters were thus measured to further characterize the effect of an anti-angiogenic therapy (antibodies against Vascular Endothelial Growth Factor, Avastin®, Roche) used alone or combined to a chemotherapy (topo-isomerase 1 inhibitor, Campto®, Pfizer) on a orthotopic human glioma model (U87-MG) xenografted in nude rats. Material and methods: Sixteen nude rats were orthotopically injected at day 0 (D0) with 10 U87-MG glioma cells. At D16, T2-weighted images were acquired to measure tumor size. Rats were randomized at D17 in 4 groups (n=4/group) with similar tumor volume (7±4mm3, data not shown). Treatment started at D18. The first group (Control group) received no treatment. The second group (Av group) received 5 injections of 5 mg/kg of Avastin i.v. every 4 days (D18, D22, D26 and D30). The third group (Cam group) received 3 injections of 40mg/kg of Campto every 7 days (D18, D25 and D32). The fourth group (Av-Cam group) received both treatments (same schedule as for independent treatment). In the Av-Cam group, one rat died before the MRI session. BVf, VSI, apparent diffusion coefficient (ADC), BBB perm. to a P846 (Gd-based contrast agent, 3.5kDa, obtained from Dr P. Robert, Guerbet, France) and lSO2 imaging were performed, at 4.7T (Bruker Avance 3 system), at D34 (after the end of both treatments). Tumor volume was computed from T2-weighted images (T2w). ADC, BVf and VSI were mapped using diffusion-weighted and multiple gradient-echo/spin-echo MR sequences applied before and after intravenous injection of ferumoxtran-10 (Sinerem®/Combidex®, 200μmol Fe/kg, obtained from Dr P. Robert, Guerbet/AMAG Pharmaceuticals) (2). BBB perm. was assessed based on T1-weighted images acquired before and 5-min. after injection of P846 (50μmol Gd/kg). lSO2 was mapped using a modified vesion of the method proposed by He and Yablonskiy (3). Voxel size was 234x234x1000μm , except for lSO2 maps (468x468x1000μm). Data, averaged across rats, are presented for 2 regions of interest (whole tumor and contralateral striatum (Contra)) and each group. Student t-tests (after assessment of variance homogeneity) were used to assess differences (*:p<0.05, **:p<0.01, ***:p<0.001). Results: Tumor growth in Av and Av-Cam groups were significantly inhibited compared to the Control group (Table1). In the contralateral striatum, there were no significant differences in ADC, VSI and BVf between groups (Table1). Contralateral lSO2 in Control, Av and Cam groups was similar but was significantly reduced in Av-Cam group (Table1). Tumoral ADC, BVf, VSI and lSO2 did not differ between Control and Cam group and were higher than in contralateral striatum (Table1). Tumoral ADC, BVf and lSO2 in Av and Av-Cam groups were significantly lower than in control group while tumoral VSI was higher than in the Control group (Table1). In all groups, BBB remained permeable to P846, although in Av-Cam group vessel appeared less permeable to P846 (data not shown).