Molecular imaging is playing a central role in shaping the future of cell-based therapies. Even as human studies near approval by the FDA, critical questions remain to be answered in preclinical work. Researchers rely on MI to produce those answers.
"One of the big things we want to know and the FDA wants to know about stem cell therapy is, Did the cells get there?" said Dr. Joseph Frank, chief of the Experimental Neuroimaging Section of the Laboratory of Diagnostic Radiology Research at the National Institutes of Health Clinical Center.
Ensuring accurate cell delivery is fundamental to determining whether cell therapy lives up to its early promise.
"If we don't find an effect, we have two questions: Did the cells not work, or did the cells not get there? Imaging can play an important role in assessing that," said Jeff W.M. Bulte, Ph.D, an associate professor of radiology at Johns Hopkins University School of Medicine.
Although MRI, PET, and optical imaging are all used to track stem cells, the underlying imaging strategy, regardless of modality, is based on the nuclear medicine tracer principle discovered by George de Hevesy. He was awarded the 1943 Nobel Prize in chemistry for discovering that molecules labeled with a radioisotope can be followed as they pass through an organism or as they contribute to complex biochemical processes, such as photosynthesis.
Indium-111 oxine SPECT is a logical choice for tracking stem cells for a limited time, said Dr. Sanjiv "Sam" Gambhir, chair of nuclear medicine at Stanford University Medical Center. But, like all radioistopes safe enough use in humans, I-111 has a short half-life that limits its value to measurements taken within a few days after its administration.
Cell therapists need to follow stem cells for weeks, months, and even years after they are introduced as therapy, according to Gambhir. This requirement has led molecular imaging researchers to experiment with contrast-enhanced MR and radically different reporter gene nuclear imaging techniques as potential alternatives.
"This is an area where molecular imaging can play a significant role," Gambhir said at the 2005 Society of Nuclear Medicine meeting.
Reporter gene imaging involves genetically tagging precursor cells to permanently alter their genetic composition (see Molecular Imaging Outlook 2005;2:1-3). The altered gene imprints the cell with a specific protein that distinguishes it and its cellular progeny from other cells after they divide and redivide. Complementary reporter probes, based on either nuclear medicine or optical imaging technologies, can identify reporter gene-labeled cells as long as the original stem cells or their offspring are present in the body.
Dr. Joseph C. Wu, a cardiologist in Stanford's molecular imaging program, is genetically tagging paired myocardial myoblasts transduced with a lentivirus to express a green fluorescent protein and firefly luciferase. The myoblasts are infused in myocardial muscles of rats where they have been successfully tracked with optical imaging for up to two weeks.
Newer reporter gene imaging strategies promise to lengthen the time in which stem cells can be tracked. One MRI approach taken by researchers at Carnegie Mellon University in Pittsburgh uses a metalloprotein from the ferritin family as a reporter gene. Gene expression causes the cell to sequester endogenous iron, which changes MR relaxivity. No external contrast agent is needed.
Another MRI approach, spearheaded by researchers at Johns Hopkins, relies on an artificial lysine-rich protein, which expresses amide protons that can be detected by chemical exchange saturation transfer imaging. Small differences in signal intensity against a background are plotted on statistical maps and highlighted in red. The technique has been successful in vitro, and studies in rodents are under way.
"If this reporter gene approach works, it could revolutionize how we label and track cells with MR," Bulte said. "We're adventurers on a new frontier."
MR IRON OXIDE LABELING
A key technique in optimizing cell therapy is MRI of cells labeled with superparamagnetic iron oxides (SPIOs). With its ability to precisely target cell delivery, track cell migration, and noninvasively evaluate living subjects over time, this technique is helping to bridge the gap between bench and bedside.
Researchers at the University of Nijmegen in the Netherlands, led by Carl Figdor, Ph.D., director of the Nijmegen Centre for Molecular Life Sciences, are conducting one of the first studies of magnetically labeled therapeutic cells in humans. Using dextran-coated SPIO nanoparticles (Endorem in Europe, Feridex in the U.S.) and indium-111 oxine, they are colabeling dendritic cells taken from the blood of melanoma patients. These immune system cells present antigens on their surface, migrate to the lymph nodes, and stimulate T cells to mount an assault against a foreign invader. Phagocytic by nature, they spontaneously take up the iron oxide and indium oxine labels without need for a transfection agent.
After culturing, the researchers inject 15 million labeled dendritic cells under ultrasound guidance into a lymph node at the site of the tumor. To monitor the migration of the dendritic cells, patients undergo both SPECT and 3T MRI before and 48 hours after injection of the dendritic cells. The lymph nodes are then resected and examined histologically and with a 7T MR spectrometer. The researchers have found that MR can easily track dendritic cell migration and assess accurate cell delivery.
TREATING MYOCARDIAL INFARCTION
At Johns Hopkins, cardiologists are beginning to do bone marrow stem cell therapy in patients with acute myocardial infarction, also using unlabeled cells. Dara Kraitchman, Ph.D., an associate professor of radiology at Hopkins, acknowledges that performing these procedures without cell tracking is a pragmatic decision, but she is concerned that it leaves too many questions unanswered.
"The problem with the clinical trials is that people don't know if the cells got to the target. Where else did they end up? Were they still there at some later time?" she said.
Moreover, an assessment of changes in ventricular function offers only a rough gauge of the effectiveness of stem cell therapy. If cardiac function does not improve, no one knows if it is because too few cells arrived at the site of infarct or because the cells simply did not persist long enough. If cardiac function does improve, the question remains whether improvement would have been greater if more cells had been delivered.
Meanwhile, research spadework continues in animals. Early studies of cell tracking with MR showed that a wide range of labeled cells-from organ-specific stem cells to mesenchymal stromal cells and endothelial progenitor cells-could be tracked in the liver, kidney, heart, brain, spinal cord, and other organs.
A team led by Michael Chopp, Ph.D., director of the Neuroscience Institute at Henry Ford Health Sciences Center in Detroit, took neural stem cells from the subventricular zone of the brain, labeled them with ferromagnetic particles, and implanted them in the cisterna magna of rats with ischemic stroke (Figure 2). The cells selectively migrated toward the ischemic zone.
Frank's team at the NIH has shown that MRI can track endothelial precursor cells as they home into sites of tumor vasculogenesis. In a study published in the January 2005 issue of Blood, they labeled endothelial progenitor cells with Feridex, intravenously injecting them in immunocompromised mice implanted with glioma tumor cells. Over time, a continuous dark hypointense ring developed around the outside of the tumor, which was shown on histology to coincide with areas of angiogenesis.
"This shows incorporation of stem cells into the ongoing vasculogenesis that occurs in a growing tumor. We're actually visualizing with MRI the neovasculature of this tumor," Frank said.
Frank is in discussions with the FDA, seeking approval for a follow-up safety study in humans. Feridex-labeled hematopoietic cells-transfected using clinically approved protamine sulfate-would be injected intravenously in human patients with glioblastoma multiforme. Patient enrollment may begin in early 2006.
The clinical implications of such a study are intriguing, said Dr. Neil D. Theise, a professor of pathology at Albert Einstein School of Medicine, who has conducted stem cell research in the liver and other organs. It raises the possibility of using tagged endothelial progenitors to identify early de novo tumors, either as a screening device for high-risk patients or for finding micrometastases in patients with more advanced tumors.
Iron oxides are being used to label more than stem cells. In work that could shed light on the treatment of diabetes in humans, Paula Foster, Ph.D., a scientist with the Imaging Research Laboratories at Robarts Research Institute in London, ON, Canada, is using MR to image SPIO-labeled pancreatic islet cells (Figure 3).
The islets of Langerhans, small clusters of endocrine cells scattered throughout the pancreas, house the beta cells that produce insulin. Islet transplantation is emerging as a possible cure for diabetes, but it is unclear how long islets survive or how well they function after transplantation. Observations made possible because of iron oxide labeling of implanted islets may produce answers.
"We're able to resolve individual islets," Foster said. "The ability to create 3D high-resolution images is another advantage of MR."
Next, Foster plans to monitor islet viability and investigate how the characteristics of the FIESTA signal change if the islet dies and loses its iron.
For all of the advances in MR, experts agree that cell imaging will draw on a range of complementary imaging modalities. MR is ideal for imaging cellular therapy of the heart when the cells are delivered directly by intramyocardial or intracoronary injection, Bulte said. A recent study in dogs, however, showed that intravenous delivery calls for use of nuclear imaging techniques.
In this study, the Johns Hopkins team labeled MSCs with both Feridex and indium oxine, injecting about 100 million cells into dogs in which acute myocardial infarction had been induced (Figure 4). On day one, fused SPECT/CT images showed cells accumulating in the lungs. On day two, researchers observed cells in the liver and spleen, as well as a hot spot in the heart, which persisted for at least five days.
Quantitative analysis showed that only 50,000 to 100,000 cells actually reached the heart, no more than 0.1% of the injected cells. MR was unable to detect any Feridex-labeled cells.
"We could not detect very small quantities of cells with MR, whereas SPECT/CT could," Bulte said.
Researchers continue to explore new MR techniques. Several groups are working on turning darkness into light when imaging SPIO-labeled cells. Iron oxide causes significant signal dephasing due to the magnetic field inhomogeneity induced in the water near the cell. The resulting signal void causes iron-labeled cells to appear black on MR images.
A new technique suppresses the on-resonance signal to which MR is ordinarily tuned and, instead, images the off-resonance water surrounding the labeled cells. A team at Stanford described this technique in the May issue of Magnetic Resonance in Medicine.
At the International Society for Magnetic Resonance in Medicine meeting in May, Matthias Stuber, Ph.D., from Johns Hopkins reported on a similar technique, dubbed inversion recovery with ON-resonant water suppression, or IRON.
"We're taking that typical dark artifact and making it bright to better identify exactly where the stem cells are and also perhaps to have a higher sensitivity to a smaller number of cells," said Kraitchman, who participated in the study.
A potential benefit of IRON and similar techniques is the ability to distinguish iron-labeled cells from other causes of magnetic field inhomogeneity, such as hemorrhage. Frank, whose team is also working on positive contrast techniques, said distinguishing labeled cells from hemorrhage remains an ongoing research challenge.
By creating a virtual window to observe cellular behavior, stem cell tracking promises to become MI's gateway into clinical practice. MI's importance extends far beyond just stem tracking, according to Frank. It may ultimately play a role in patient selection, the characterization of pathology, and determination of whether direct implantation or vascular routing of stem cells will be more effective.
As stem cell implantation techniques grow in sophistication, Frank envisions MI aiding in the optimization of implant surgery.
"This is a big one," he said at the ISMRM meeting. "We won't just be tracking migration, survival, and differentiation in the target anatomy. Through our work, questions of how many, how often, and when to give cells will be answered."
MI techniques will help the therapist to determine whether gene therapy alone will be sufficient or whether it should be combined with other techniques, such as adaptive immunotherapy. Future research will determine which imaging modality or combination of modalities is best suited for specific therapeutic applications.