Advances in Imaging Techniques Help Drive Stem Cell Research Forward


Advances in Imaging Techniques Help Drive Stem Cell Research Forward

A 57-year-old man, suffering seizures and headaches, was diagnosed with a brain tumor in 2005. Doctors at the City of Hope cancer center in Duarte, California, removed the mass, but knew his cancer was of a type likely to return.

So they enrolled him in a cell therapy trial, and collected T cells from his blood. When the cancer did indeed recur, nine months later, they combined surgery with infusions of his own T cells, souped up with extra genes to make them seek and destroy the remaining cancer.

But once the new cells were in the patient, would they go to the right place? Would they function as expected? Would they even survive? “If they did not target the tumor, then the therapy would not work,” said Shahriar Yaghoubi, a research radiologist at Stanford University in Palo Alto, California. “And if they went somewhere else, they might cause problems.”

The answers to those questions will be necessary for cell-based therapies, including stem cell treatments, to make it through Food and Drug Administration review and into the clinic. “Right now, a lot of procedures are actually a black box,” said Charles Lin, who is investigating cell tracking at Harvard Medical School in Boston, Massachusetts. “Imaging technology would allow us to peek inside the black box and see what’s happening.” Researchers are developing long-lasting tags to track cells, in vivo, with fluorescence microscopy, magnetic resonance imaging (MRI) and positron emission tomography (PET) (reviewed in Rodriguez-Porcel et al., 2009). “It is a very new technique right now,” said Ali Arbab, a research radiologist at Henry Ford Hospital in Detroit, Michigan.

Doctors can inject cells directly into the target area, or rely on attractors such as cytokines to guide the new cells to the right place. But the majority of cells may fail to reach the target, Arbab said, and may not survive at all. The propensity for cells to find the right site is also likely to be cell type-specific, said Khalid Shah, a neurobiologist interested in cancer at Harvard Medical School and Massachusetts General Hospital in Boston.

Preclinical and Particle Tags

The first place to examine stem cell homing and function is in preclinical animal models, and the most straightforward technique is to mark cells with fluorescent proteins or luciferase. Optical imaging offers single-cell resolution. It is most commonly used with tissue samples once the animal has been sacrificed, but now there are ways to look at cells in a living organism with intravital microscopy. “There has been a real explosion of interest in this technique,” Lin said, although few labs are using it for stem cell tracking yet. Lin had to build his own microscope, but top manufacturers are starting to develop commercial versions, he said.

The advantage of intravital microscopy is not only longitudinal studies of living animals, but also its ability to find rare cell types. Lin and colleagues have used the technique to image hematopoietic stem cells that migrated to the bone marrow inside a mouse skull (Lo Celso et al., 2009: PMID19052546). Only five to ten cells may get there, he said, and picking them out in a set of thin sections would be tedious. With confocal imaging of the entire skull, he can more easily find the cells of interest.

Shah has used fluorescent markers to track stem cells migrating toward mouse gliomas, and determine that neural stem cells move through brain tissue better than mesenchymal stem cells. (Shah et al., 2008: PMID18434519; Sasportas et al., 2009: PMID19264968). But fluorescence only travels through a few millimeters of tissue; Shah must cut a porthole in the skull to image the brain. This kind of microscopy would not be able to reach deep into the brain of someone with a tumor, for example. “The translation of optical imaging to clinics is limited,” Shah said.

The clinical method that has been used most so far it is to force cells to take up super-paramagnetic iron oxide (SPIO) nanoparticles. The iron particles interfere with MRI, appearing as a shadow on the scan. However, the technique has limitations. “It only tells you about the location,” Yaghoubi said. “If the cells multiply, you keep diluting your probe.” And if the cells die, the nanoparticles may hang around for a while.

Four human SPIO trials (reviewed in Bulte, 2009: PMID19620426) have shown that it is feasible to detect cells this way. However, all four were done outside of the United States. Stringent government regulations make it difficult to do human trials in the U.S., said Jeff Bulte, a radiologist at the Johns Hopkins University School of Medicine in Baltimore, Maryland. Another roadblock is that the FDA-approved SPIO particle Feridex has been off the market since 2008, and its producer, AMAG Pharmaceuticals of Lexington, Massachusetts, has no plans to make it again.

It may be worthwhile, then, to explore alternative tracers. Celsense, Inc., of Pittsburgh, Pennsylvania, is developing perfluorocarbon tags (reviewed in Janjic and Ahrens, 2009: PMID19920872). They work much like SPIO, in that cells take up the inert particles. Radiologists detect the marker with an MRI tuned to fluorine. Because fluorine is rare in the body, only tagged cells show up. Doctors can then overlay that image with standard MRI, showing organs and tissues, to see where the labeled cells are. Scientists are “on the verge” of trying the marker in human trials, said Celsense president Charlie O’Hanlon.

It’s in the Genes

Both perfluorocarbon and SPIO signals may fade over time as cells divide or die, and neither will tell doctors what the cells are up to. “The next question is do they survive, and that’s where you need another approach,” said Raphael Guzman, a neurosurgeon at Stanford University. So the next big thing in the field is imaging techniques based on reporter genes that cells transcribe as long as they live.

This is the approach taken in the man (his identity is protected by privacy laws) from the City of Hope trial, which enrolled a handful of patients. When his doctors engineered his T cells, they added a gene for interleukin 13 zetakine, a receptor to home in on the cancer. In case something went wrong with the transplanted cells, they also added a failsafe, a suicide gene from herpes simplex virus 1 called thymidine kinase (HSV1-tk). In addition to phosphorylating nucleosides, its natural role, HSV1-tk also phosphorylates the drug ganciclovir. Harmless on its own, ganciclovir becomes cytotoxic upon phosphorylation. If the new cells went rogue, ganciclovir treatment would destroy only the transplanted cells.

Conveniently, HSV1-tk also forms the basis for a PET imaging system. Yaghoubi and his colleagues, led by Sanjiv Gambhir of Stanford University, piggybacked onto the City of Hope study in 2006 to test-drive the concept. In addition to nucleosides and ganciclovir, HSV1-tk phosphorylates a probe called FHBG. FHBG can move in and out of cells, but once phosphorylated—as would only happen in the transplanted T cells or their descendants—it gains a negative charge and can no longer cross the cell membrane. “The only way the probe gets trapped inside the cells is if these cells are alive,” Yaghoubi said. The researchers used FHBG containing radioactive fluorine so the labeled cells will light up on a PET scan. The approach is “a good way” to follow living cells, said Arbab, who was not involved in the trial.

This study (Yaghoubi et al., 2009: PMID19015650), the first to use reporter gene-based imaging of therapeutic cells in a person, enrolled the man with the brain tumor in 2006. The researchers were able to see the site of the returned tumor, but they also got a surprise: “We noticed there was another signal coming from another site in the brain,” Yaghoubi said. The City of Hope doctors confirmed that there was a second tumor. “Had it not been for this imaging, they might not have detected that until later on,” Yaghoubi said.

The man survived for fourteen months after the recurrence. The researchers have tried the same technique in a second patient, and intend to re-team with City of Hope in a larger cell therapy trial expected to begin enrollment in February, Yaghoubi said.

The More the Merrier

Gambhir founded a company, Cellsight Technologies in Sunnyvale, California, to pursue reporter-based imaging techniques. They will develop methods to image “anything about the status of the cell,” said Yaghoubi, who serves part-time as chief scientific officer. Celsense is also working on a reporter system.

Reporters have their own drawbacks. “Transforming cells genetically adds a significant layer of complication” to safety studies, reducing the clinical appeal, O’Hanlon said. And HSV1-tk, being a viral gene, could cause an immune response. Researchers are working on developing other reporters, such as the human sodium iodide symporter, that might be safer (Terrovitis et al., 2008: PMID18992656). PET also does not give as detailed resolution as MRI.

“I do not think there is a single best method,” Lin said. “It is going to be a combination of methods.” Ultimately, doctors might select an MRI-based technique for the better resolution plus a reporter to confirm cell survival.

Will cell labels become standard procedure for all cell therapies? “The real clinical application of SPIO-based cell tracking will be to make sure the cells go to the right site when performing the injection procedure,” Bulte said. Longer-term tracking may only be necessary at the clinical trial stage, Arbab said, to prove that cells go to the right place and do the right thing. But it is possible, Yaghoubi said, that different people will respond differently to cell infusions. “You may actually want to see what happens to the cell in each individual, so the imaging might actually become part of the therapy.”


Bulte, J.W.M. (2009). In vivo MRI cell tracking: Clinical studies. Am. J. Roentgenol., 314-325.

Janjic, J.M., and Ahrens, E.T. (2009). Fluorine-containing nanoemulsions for MRI cell tracking. Wiley Interdiscip. Rev. Nanomed. Nanobiotechnol., 492-501.

Lo Celso, C., Fleming, H.E., Wu, J.W., Zhao, C.X., Miake-Lye, S., Fujisaki, J., Côté, D., Rowe, D.W., Lin, C.P., and Scadden, D.T. (2009). Live-animal tracking of individual haematopoietic stem/progenitor cells in their niche. Nature, 92-96.

Rodriguez-Porcel, M., Wu, J.C., and Gambhir, S.S. Molecular imaging of stem cells (July 30, 2009), StemBook, ed. The Stem Cell Research Community, StemBook, doi/10.3824/stembook.1.49.1, .

Shah, K., Hingtgen, S., Kasmieh, R., Figueiredo, J.L., Garcia-Garcia, E., Martinez-Serrano, A., Breakefield, X., and Weissleder, R. (2008). Bimodal viral vectors and in vivo imaging reveal the fate of human neural stem cells in experimental glioma model. J. Neurosci., 4406-4413.

Terrovitis, J., Kwok, K.F., Lautamäki, R., Engles, J.M., Barth, A.S., Kizana, E., Miake, J., Leppo, M.K., Fox, J., Seidel, J., et al. (2008). Ectopic expression of the sodium-iodide symporter enables imaging of transplanted cardiac stem cells in vivo by single-photon emission computed tomography or positron emission tomography. J. Am. Coll. Cardiol., 1652-1660.

Yaghoubi, S.S., Jensen, M.C., Satyamurthy, N., Budhiraja, S., Paik, D., Czernin, J., and Gambhir, S.S. (2009). Noninvasive detection of therapeutic cytolytic T cells with 18F-FHBG PET in a patient with glioma. Nat. Clin. Pract. Oncol., 53-58.