---

Malignant glioma is one of the most virulent neoplastic diseases in humans. The median survival remains in the 10 to 12 month range despite aggressive surgery, radiation, and chemotherapy. Although there are improvements in molecular characterization of high-grade glioma, in targeted therapeutic approaches and in refined radiological methods, the average life span of an afflicted individual has not been significantly prolonged, due to recurrent refractory disease. The tumor stimulates the formation of new blood vessels which are structurally and functionally abnormal, impairs effective delivery of therapeutic agents and hence reduces the effectiveness of radiation and chemotherapy.

Angiogenic therapies: The tight correlation between angiogenic processes and prognosis have lead, during recent years, to propose new strategies which utilize combinations of antiangiogenic and chemo therapies. Such drugs, when used judiciously, have the potential to normalize structurally and functionally abnormal tumor vasculature, reduce the risk of hemorrhage, enhance the penetration of concurrently administered chemotherapeutic and improve the efficacy of cytotoxic drugs and radiation by alleviating hypoxia. Moreover, such regimens decrease microvessel density, new vessel evolution and interstitial pressure. Yet, improvement is limited and the exact patho-physiologic processes of the therapeutic effects are not fully understood. The first aim of this study is to improve understanding of the patho-physiological processes of glioma in relation to the clinical course and to pathological parameters, prior to and following therapeutic intervention. This knowledge regarding the tumor response at the system level will be achieved using animal models.

MRI is the method-of-choice for noninvasive whole brain assessment of gliomas, and has an essential role in classification and grading, preoperative evaluation, follow-up and therapeutic management. MRI can provide structural, biochemical and functional information regarding the tumor and its surrounding parenchyma. With the development of advanced MR methods, an increase in sensitivity along with improvement of reliability and specificity in the diagnosis of tumors has been achieved. However, more sensitive tissue differentiation, such as compartmentalization of the tumor into several regions differing in vessel density, diameter and in necrosis is needed. In addition, correlation to pathology, i.e. the micro characterization of the tumor like blood vessel density and morphology, is desired. One major problem is the reliability of the existing diagnostic criteria for assessing treatment response, which is questionable and can sometimes be misleading.

Advanced imaging methods provide additional information regarding vascularization which determines one of the pathophysiological characteristics of the tumor and its potential response to therapy. Density, permeability and blood brain barrier (BBB) breakdown are some aspects of the tumor vasculature that can provide valuable information for tumor prognosis and response to therapy. However, noninvasive methods for quantification and detecting vascularization are currently limited. We have previously presented a novel fMRI method using hyperoxia and hypercapnia (hemodynamic response imaging – HRI) for the detection of vascular functionality and maturation in somatic tumors. This method has been shown to facilitate detection of mature vessels resistant to anti-angiogenic therapy in animal models. Recently, we obtained preliminary results showing the high sensitivity of this method in detecting changes occurring in response to combination of Avastin and Temozolomide, as well as in response to radiation. Using this method we were able to differentiate between various tissue types within the tumor and its surrounding tissue, according to its response both to O₂ and CO₂. We were also able to detect changes in the response of the tissue to both hyperoxia and hypercapnia under therapy. High correlation to pathology was obtained both to vascular density, vascular endothelial proliferation and neogenesis as detected using various immuno-histochemical staining such as anti CD34 antibody, anti CD31 antibody, smooth muscle actin (SMA) staining and more. Another novel method, susceptibility weighted imaging (SWI), has been shown to have high sensitivity for hemorrhagic changes but has not been extensively studied in the context of hemorrhagic changes occurring during tumor progression and therapy response. Our preliminary results show also high sensitivity of this method to monitor treatment changes, indicating a reduction in the hemorrhage both in response to Avastin and chemotherapeutic drugs as Temozolomide as well as to radiation therapy. These methods, in combination with other advanced imaging methods seem to improve diagnostic sensitivity and specificity, and might help better understanding of treatment response. The second aim of this study is to provide a comprehensive imaging diagnostic tool that will take into account all aspects of structure, metabolism, functional and angiogenic features of the tumor tissue and surrounding.

Pathology: Although MRI, in many cases, is the sole non-invasive method in evaluation of gliomas, histo-pathology is the gold standard for tumor tissue classification and grading. Histopathological tissue diagnosis is based primarily on characterization of cell morphology and protein expression levels of specific tissue elements, and more recently on detection of molecular and genetic features of the tumor tissue. In recent years, imaging and image analysis methods have become integrated into the pathologists' common practice. Though they are not yet routinely used by all, these new utilities overcome a few of the drawbacks of the pathologist's diagnostic process such as the lack of objectivity, reproducibility, reliable quantification and standardization. Nonetheless, pathologists are still in the forefront of the diagnostic process, though they are now being aided by the computerized analysis of the specimen slides at hand. These new development allow, for the first time, the establishment of unified statistical samples that can be analyzed internally and correlated externally with other diagnostic methods such as MRI. More accurate diagnosis, prognosis and possibly personalized therapy, all derived by the utilization of the micro (pathology)-macro (MRI) connection, should make use of this combination.

We plan to achieve these aims by applying conventional and novel MR techniques, correlating the data to clinical evaluation and pathology and to develop software based on multi-scale multi-domain diagnostic strategy. Specifically, we will focus on patients who are candidates for radiation, anti-angiogenic therapy or surgery. A full MR protocol that includes conventional and advanced MR methods will be applied and correlated to clinical evaluation. In addition, patients who will be referred to surgery/biopsy will have their imaging parameters obtained from the tumor site and its surrounding correlated to advanced histopathology probes. Finally, in order to further address the mechanism of anti-angiogenic treatment (such as vessel cooption), we will use an intracranial glioma mouse model receiving similar therapy and undergoing the same MR protocol as the human subjects. Data analysis will be done in a multi-model analysis strategy, combining information obtained from the various imaging methods. We hope that such a multi-scale multi-domain approach to tumor diagnosis will improve the sensitivity and specificity of MRI based diagnosis of brain tumors and characterizing of their clinical course. In addition we intend to use advanced methods of signal processing for image analysis, which are becoming easier to apply in relative rapid formats. Bridging between pathology, advanced imaging and animal models is strongly needed, leading to more accurate diagnosis (sensitive and specific) and better prediction of prognosis.

In summary: In this study we intend to deal with several core open issues in clinical aspects of brain tumors such as diagnosis, prognostic evaluation and therapeutic planning and response. For that we will focus on the angiogenic process in tumors and the available agents to counteract it in animal models and humans. Improving the sensitivity and specificity of non-invasive diagnostic tools will help optimize a treatment approach based on the individual's tumor-profile, and might help to develop new strategies of therapy and new drugs. We hope that such a comprehensive approach will, in a long term, extend the limited life expectancy of patients with brain gliomas.