Tumor hypoxia as a mechanism of resistance to bevacizumab in a murine model.

Abstract

e13111 Background: Despite the promise of preclinical and early phase clinical studies, the efficacy of bevacizumab in solid tumors is more limited than expected. One of the presumed reasons is the induction of tumor hypoxia by the anti-angiogenic effects of bevacizumab, leading to therapy resistance. The aim of this study was to assess the effect of bevacizumab on tumor hypoxia in vivo in a colorectal cancer model, using functional imaging techniques. Methods: Nude mice with s.c. LS174T colon carcinoma xenografts (0.05 - 0.3 cm3) were treated with bevacizumab (5 mg/kg; 2/wk, i.p.) or saline as a control. To assess tumor hypoxia in vivo 18F-MISO-PET microPET or T2*-MRI images were acquired of separate groups of mice (n=5) before treatment and at day 2, 6 and 10 days after start of treatment. Tumors were harvested directly after imaging to microscopically assess the hypoxic fraction (pimonidazole staining) and vascular density (9F1 staining). Results: Linear regression analyses showed that FMISO uptake increased significantly more over time in the control group than in the bevacizumab group (beta -0.44, p=0.02), indicating that bevacizumab reduced the inherent increase in tumor hypoxia over time. T2* time increased significantly less in the bevacizumab group (beta -0.45, p=0.01), indicating a higher deoxyhemoglobine concentration, which might indicate a higher perfusion of the tumor and thus less hypoxia. The hypoxic fraction did not change over time and no difference was observed between the tumors in the treated and the control group. Vessel density significantly decreased over time in the bevacizumab group (beta -0.25, p=0.05), while the hypoxic fraction remained unchanged in the control group. Conclusions: The bevacizumab-induced changes in the tumor were most prominent at 10 days after treatment initiation, implying a build-up effect of repeated bevacizumab administration. The treatment induced changes could be detected with both T2*-MRI as well as with FMISO microPET. Bevacizumab did not induce tumor hypoxia, despite the observed decrease in vascular density. Therefore, induction of tumor hypoxia as a resistance mechanism to bevacizumab treatment seems unlikely.