Molecular imaging and stroke.

Molecular imaging (MI) is a rapidly developing field that encompasses many new (and old) imaging modalities that seeks to present patient-specific and disease-specific molecular and genetic information in conventional 2-dimensional and 3-dimensional anatomic imaging readouts. The foundations of MI are based on the fusion of a “promoter” agent that will be altered in a particular environment or disease state with an observable “reporter” agent that would register any change in the signal or contrast from the promoter. Much of MI is being done in experimental models of cancer, but only some of the developments can be scaled up to a clinical reality for eventual diagnostic and prognostic usefulness in stroke. The most commonly used modalities in MI research have been optical-based near-infrared or visible light sensors, bioluminescence,1,2 cameras sensitive to the firefly luciferase–luciferin generation of green light, nuclear medicine (NM)-based single-photon and positron tomography, and now, most recently, magnetic resonance (MR). Optical imaging techniques developed early for molecular and cellular biology using a wide variety of wavelengths. The noninvasive imaging in vivo with light photons has largely come from the advances in targeted bioluminescence probes, near-infrared fluorochromes, activated near-infrared fluorescence agents, and primarily from light emitted from the luciferase entity (reporter) in the presence of a substrate (luciferin).1,2 Optical techniques using multiple wavelength probes such as quantum dots holds the potential for multichannel imaging. However, most fundamental to the widespread use of in vivo optical imaging of stroke or ischemia in living subjects is the difficulty of detecting light from the brain, primarily because of the presence of the skull. Because of this, MI using optical techniques has been largely limited to nonneuro studies of cancer in rodent models. Nonetheless, optical imaging has a bright future in neuroimaging research. Advances in the use of near-infrared for diagnostic,3–5 prognostic,6,7 and, more recently, in potential stroke therapy applications8–10 have been exciting. Look for near-infrared to open research areas in mapping of edema, oxygenation dynamics, blood flow (with injected contrast agents), and NAD/NADH turnover in stroke applications. Nuclear medicine molecular imaging methods such as single photon emission CT and PET are much more relevant and indeed have been used for decades in stroke and neuro applications such as brain mapping, flow mapping, metabolism, and so on. The advent of small single photon emission CT (micro single photon emission CT), PET (microPET), and CT scanners has made the NM methods the main focus of stroke research. Here, NM relies solely on the design and use of injected MI probes to provide the imaging signal contrast. The term “probe” is commonly used to refer to the tracer, beacon, smart probe, reporter agent, contrast agent, and/or nanoparticle used. The biggest advantages of NM PET aside from the near-picomolar probe sensitivity is that existing probes can be modified with a radiolabel while minimally perturbing the parent molecule, again indirectly related to the exquisite sensitivity of NM to the radiolabel.11 This has not been widely possible for optical or MR, because the need for the contrast agent has involved bulky, less sensitive probes such as gadolinium in MR. Conversely, because NM relies on injected tracers, soft tissue contrast and resolution was poor. With the advent of PET/CT hybrids, however, 2-dimensional and 3-dimensional fusions provide both agent sensitivity and tissue contrast. The PET/CT hybrid is a potent methodology in stroke. Aside from the excellent usefulness of CT in stroke with hemorrhage detection, CT angiography, CT perfusion as well as xenon-enhanced CT, the PET/CT hybrid will allow for the addition of PET detection of receptor ligands, cerebral metabolism and blood flow, neuronal integrity as well as hypoxia and apoptosis markers.12–14 MR, with superb specificity to the water hydrogen proton, lacks a relatively low sensitivity. MR using other nuclei such as carbon-13 or fluorine-19 is much less efficient. This occurs because of the Boltzmann Distribution; that is, that at 1.5 Tesla, only approximately 5 to 10 protons per million ever contribute to the MR signal. As we will see, new methods of dynamically increasing this inherently low signal magnetization by as much as 5 orders of magnitude are now available, termed “hyperpolarization.”3,15–17 Using metabolically active carbon-13 labeled sugars, hyperpolarized MR will one day be routinely used in patients with stroke for mapping dynamic glucose usefulness, pH, oxygen extraction, and flow all in one examination. MR contrast for molecular imaging in stroke research is derived from 4 major sources: endogenous contrast from the water proton dynamics in the microenvironment, endogenous chemical shifts of proton-bearing metabolites such as lactate,