The assessment of brain tumor angiogenesis using susceptibility contrast agent-based functional magnetic resonance imaging
Abstract
In the last few years, antiangiogenic drugs have been one of the most exciting developments for the management and treatment of cancer. These early clinical trials have highlighted the need to develop techniques to assess the efficacy of these new therapies in patients. The current standard for assessing the degree of angiogenesis is to measure the microvessel density (MVD). Although the prognostic importance of MVD in breast cancer has been well established with results indicating a correlation between areas of most intense neovascularization and metastases, studies in other cancers failed to find any such correlation. These inconsistencies have largely been attributed to differences in methodology i.e. different counting techniques, selection of microvessels etc. Given these results, it is questionable whether MVD is in fact the most appropriate correlate for angiogenesis and/or validation of MRI techniques. Further, since it is not clinically feasible to regularly obtain MVD samples invasively from the patient throughout his/her treatment, a non-invasive method of evaluation seems to be the optimal avenue for the management and planning of antiangiogenic therapy. In lieu of the above, it seems that to obtain more specific information about the tumor vasculature it may be better to measure blood volume using magnetic resonance imaging (MRI). The rationale for this approach is twofold: (i) with MRI we measure the fractional blood volume and not the MVD; however, this blood volume takes into account both vessel diameter and density, which is especially important since the diameter of tumor capillaries can be as much as 3-4 times that of normal capillaries, and (ii) MRI is a non-invasive technique. Thus, tumor blood volume assays determined with contrast-enhanced MRI for assessing tumor angiogenesis hold great promise. Though MRI techniques are an indirect indicator of angiogenesis (complicated by the associated susceptibility physics), we have demonstrated in a 9L gliosarcoma rat brain tumor model, that the MR-derived CBV correlates directly with the histologically determined fractional cerebral blood volume, inversely with the MVD and exhibits no correlation with the fractional area of vessels. In addition, since MRI can be made sensitive to vessel caliber as well, we also showed that the MR-derived vessel diameter seemed to track tumor aggressiveness in both the rat tumor model as well as in patients. Finally, we developed an approach for simulating susceptibility-based contrast mechanisms for arbitrary microvascular geometries. These initial demonstrations of the ability of MRI in assessing the vascular remodeling that accompanies tumor angiogenesis underscore the fact that MRI probes of angiogenesis, sensitive to blood volume and vessel caliber have the potential to be new non-invasive in vivo prognostic indicators of tumor angiogenesis.
This paper has been withdrawn.