Image-based Systems Biology of Vascular Co-option in Brain Tumors

点击翻译按钮获取中文摘要

ABSTRACT: Recent clinical and preclinical evidence has shown that gliomas can initially grow, invade, evade antiangiogenic therapies and eventually recur by hijacking or “co-opting” the brain’s preexisting blood vessels. Vascular co-option is a nonangiogenic glioma growth mechanism in which, co-opting tumor cells cause astrocytes to lose intimate contact with blood vessels, i.e. cause gliovascular uncoupling (GVU), and alter cerebral hemodynamics. Additionally, in aggressive high- grade gliomas (e.g. glioblastoma or GBM), vessel co-option facilitates the migration and invasion of cancer cells into healthy brain tissue. Yet, the evolution of vessel co-option over the glioma’s life-time, resultant GVU and hemodynamic changes remain poorly understood due to a lack of microvascular-resolution lifecycle and multimodality/multiscale imaging approaches. Moreover, co-optive glioma growth is radiologically undetectable due to an absence of contrast enhancement and the lack of specificity of conventional MRI (e.g. T2/FLAIR) approaches. Therefore, our goal is to use an image-based systems biology approach to elucidate the hemodynamics of co-optive glioma over its lifecycle and develop an fMRI biomarker of co-option induced GVU. Guided by compelling preliminary data, we will pursue three Specific Aims: (1) Characterize vessel co-option in a patient-derived glioma xenograft (PDX) over its lifecycle with multiscale imaging; (2) Develop an image-based model of brain-wide hemodynamic changes induced by vessel co-option in glioma; and (3) Determine if rs-fMRI can detect vessel co-option induced GVU in a patient-derived glioma xenograft. Under Aim1, we propose a paradigm-shifting approach that employs a miniscope for microvessel resolution (~5 µm) multicontrast in vivo imaging of co-option induced hemodynamic changes over the lifecycle of a patient-derived glioma xenograft. We will complement these microvascular-scale measurements with multimodality/multiscale whole-brain data from ex vivo CT/MRI/light sheet microscopy (LSM) in the same animal to corelate structural/functional/cellular changes in the vascular microenvironment (VME). Under Aim2, we employ these data in a model of co-option induced hemodynamic dysregulation to simulate brain-wide changes that could be exploited as fMRI biomarkers of co-optive glioma. Under Aim3, we will determine if resting-state fMRI (rs-fMRI) can detect GVU in a co-optive PDX and differentiate it from non-co-optive glioma growth. Our approach is innovative because it blends cutting-edge advances in miniaturized microscopy, multiscale/multimodality imaging and image-based systems biology. The proposed research is significant because these studies will establish: (i) freely downloadable, co-registered multiscale data for cancer systems biology investigators; (ii) a hemodynamic model for co-optive glioma; (iii) a novel biomarker of glioma co-option with the potential to transform patient management and stimulate the development of therapies to thwart antiangiogenic resistance. We also expect this approach to be adaptable to other CNS diseases dependent on vessel co-option (e.g. brain metastases).