Peter C.M. van Zijl - US grants

It is believed that early intervention is the key to successful therapeutic outcome in stroke, which is the third most frequent cause of mortality in western society. Thus, the ability to diagnose ischemic brain tissue with a high degree of specificity and sensitivity is critical. Magnetic Resonance Imaging (MRI) and Spectroscopy (MRS) hold great promise for noninvasive assessment of brain damage. However, if MR is to become useful for prognosis, it is essential to determine whether reversible or irreversible damage has occured during and after ischemic periods. In addition to assessing large- and small-vessel perfusion using MR angiography (MRA) and dynamic contrast imaging, it is essential to have access to MR parameters that reflect reversible and irreversible tissue damage during and after ischemic periods. Spin-Density/T2-weighted imaging is a good indicator of edema (hyperintensity), but is not sensitive in the acute phase. It is therefore important to develop new functional imaging methods that can quantitatively assess tissue status when neurologic recovery is still possible. Diffusion imaging can detect ischemic tissue within minutes post-onset, but contrary to early expectations based on animal studies, clinical results generally show that regions of reduced diffusion proceed to infarction at follow-up. In addition, perfusion images generally show an area of reduced flow larger than the region of compromised diffusion, the so-called perfusion-diffusion mismatch. Because it is essential to assess the risk of infarction in this region, which often evolves to reduced diffusion, there is a need for new functional modalities to diagnose this mismatch area at the time of clinical evaluation. Based on recent results obtained by us, we have designed the following hypotheses: (1) quantification of oxygen extraction ratio (OER) can predict the risk for tissue infarction based on the principle of flow thresholds; (2) changes in protein synthesis are reflected in the proton magnetization transfer rate between proteins and water, which can be imaged through the MRI relaxation rate T1rho; (3) It is possible to measure pH using proton MRS, which will provide an additional tissue parameter for stroke evaluation on a standard clinical scanner (proton only). Our corresponding three aims are to develop new methodologies to measure OER, T1rho, and pH, and to subsequently test our three hypothesis on cat brain models of reduced blood flow and of transient global and focal ischemia. Our fourth and final aim is to implement the new technologies on the clinical scanner and to optimize their use for a fast and specific clinical stroke exam.

whole body imaging /scanning; magnetic resonance imaging; nuclear magnetic resonance spectroscopy; biomedical equipment purchase; image processing;

DESCRIPTION (Adapted from applicant's abstract): MRI signals are sensitive to physiological alterations such as changes in cerebral blood flow (CBF), volume (CBV) and oxygenation. As a consequence, MRI methods can be designed that produce blood-oxygenation-level-dependent (BOLD) image contrast. The possibility of detecting such effects has stimulated a boon in the field of functional MRI (fMRI) of the brain, where neuronal activation is reflected in slight focal increases in signal intensity. Despite great progress in understanding some of the mechanisms of these BOLD signal changes, the technique is often criticized because no exact equations relating the measure MRI effects to basic physiological parameters, such as CBF, CBV, oxygen metabolic rate, hematocrit, and arterial oxygen saturation, have been established. The applicants have recently developed a general theory that can quantitatively explain spin-echo (SE) relaxation effects (R2) in terms of hemoglobin deoxygenation and oxygen extraction ratios (OER). This theory has to be tested rigorously using experiments in which hemoglobin deoxygenation (Aim 1) and OER (Aims 2,3) are well understood and can be controlled, after which it can be applied to determine OER effects in fMRI (Aim 4). The aim is to quantitatively measure R2 and the SE signal intensities of water as a function of the inter-echo time spacing in the NMR pulse sequence and as a function of field strength for the following conditions: At different hemoglobin oxygenation levels and hematocrits in isolated blood (AIM); As a function of oxygenation in vivo in the cat brain (AIM 2), and in vivo in the human brain (AIM 3). These experiments will also include simultaneous determination of arterial oxygen saturation, pH, and blood gases (AIM 1-3), as well as measurement of arteriovenous differences and absolute blood flow (microspheres) in the animals (AIM 2). Finally, these relaxation rates and arterial oxygenation and blood gases will be measured during visual stimulation in humans at field strengths of 1.5T and 4.0T (AIM 4). These efforts should lead to a better understanding of the physiological mechanisms underlying the fMRI signal changes in neuronal activation and allow quantitation of cerebro-haemodynamic parameters to be utilized for high resolution mapping of structure/function relationships in intact brain. This understanding should facilitate optimal design of fMRI experiments in terms of the most suitable MRI pulse sequence parameters to obtain maximum effects.

In this project, the overall goal is to design novel molecular MRI methods that are minimally invasive or totally noninvasive and, as such, have a high potential of being translated rapidly into the clinic to be used for tumor assessment and monitoring of treatment. Towards this goal, we exploit so-called chemical exchange saturation transfer (CEST) contrast, which is generated through magnetic labeling of exchangeable protons (such as NH and OH) on either exogenous or endogenous agents, followed by a physical transfer (chemical exchange) of this label to water protons, which allows detection using MRI. To reach our ultimate goal of fast human translation, we will focus our efforts on diamagnetic, biodegradable, non-metallic compounds. Specifically, we will exploit the body's own building blocks, proteins and carbohydrates as CEST biomarkers and develop MRI technology to detect these markers. Tumors are generally characterized by an increased content of small mobile proteins and peptides, rapid glucose metabolism, and increased permeability between blood vessels and extravascular extracellular space. The overall goal therefore is to develop MRI pulse sequence technology and theory for detecting mobile protein content, glucose delivery and metabolism, and tumor perfusion. Our first aim is to assess protein content by employing nuclear interactions within these macromolecules (cross-relaxation) combined with the exchange ofthe protein's amide protons to water protons. In the second aim, glucose metabolism and tumor perfusion will be assessed by monitoring the uptake of non-labeled D-glucose using CEST. These technologies are expected to be applicable for most tumor types, but to demonstrate their applicability, we will apply them first to two human breast cancer lines: less aggressive (MCF-7) and highly aggressive and metastatic (MDA-MB-231). This will be done both ex vivo, in perfused cells and, in vivo, on xenografts in mice. As a third aim, we will perform pilot studies in patients to show feasibility of rapid translation. These aims are expected to result in the availability of molecular MRI technologies in vivo that are suitable for immediate application in humans. Once established, we expect that these methods can be used for tumor detection, imaging tumor perfusion and metabolism, assessing tumor malignancy, and monitoring tumor treatment. This is expected to reduce false-positive detection rates by functioning as an add-on for current high-volume screening approaches and to improve treatment monitoring by MRI.

DESCRIPTION (provided by applicant): The overall goal of this academic-industrial partnership between researchers at Johns Hopkins University and Philips Healthcare/Philips Research is to develop and optimize the novel amide proton transfer (APT) imaging technique for efficient and reliable detection of malignant brain tumors and assessment of treatment response. APT imaging can provide endogenous contrast related to mobile protein content in tissue. Preclinical studies and clinical data suggest that APT imaging may provide unique information about the presence and grade of brain tumors, as revealed by MRI-guided proteomics and in vivo MR spectroscopy. Notably, we recently demonstrated in animal models that the APT-MRI signal is a unique imaging biomarker to distinguish between radiation necrosis and active tumor tissue. Similar to other MRI techniques, the ultimate goal for APT imaging is the standardized use in a clinical setting. However, current APT imaging protocols are far from being optimized. A main reason is that the APT experimental parameters are often limited by scanner hardware constraints, particularly with respect to amplifier duty-cycle and specific absorption rate requirements. In addition to pulse sequence parameter differences, leading to inconsistent effect sizes, data processing strategies vary and may affect the reproducibility of results between hospitals. Therefore, there is an urgent need for industry and academia to work together to develop this emerging technology into a clinically viable, easy-to-use, and reproducible approach. To accomplish this, we (researchers at JHU and Philips) agreed to establish a cooperation to achieve an accelerated APT translation. Together, we formulated the following two specific aims: (1) Optimize and standardize APT imaging technology on 3T human MRI systems; and (2) Test the methodologies developed in Aim 1 on patients with brain tumors. APT imaging has the potential to introduce an entirely new molecular MRI methodology into the clinic that can detect endogenous cellular protein signals in biological tissue non-invasively. Thi research will provide the standard optimized approaches required to translate this new technology into the clinic.