Editorial Article
Monday July 26, 2010
by Jeffrey M. Perkel
When it comes to animal research, there's no getting around one simple fact: It's expensive. Between care, housing, and feeding, live animals can take a serious bite out of a researchers' bottom line.
That's especially true if the researcher is performing some sort of time-course—studying tumor development and metastasis over time, for instance. In that case, he or she will have to have enough animals to sacrifice a few at every time point. Add in a few variables, such as different drug treatments, say, and the required animal population can explode.
There is a cheaper alternative, however. Instead of sacrificing animals to monitor their collective progress, why not image individuals over time in a so-called longitudinal study?
Over the past decade, the scientific community has increasingly embraced that idea, and the literature has become, well, littered, with false-color images of intact, immobilized animals whose tumors, visualized without surgery, seem literally to be bursting with light, thanks to a relatively new technique called in vivo imaging.
The first commercial whole-animal in vivo optical imagers appeared in the late 1990s. Since 2003, says James Mansfield, director of multispectral imaging systems at Cambridge Research & Instrumentation (CRi), the market has grown about 20% annually.
Mansfield attributes the sudden popularity of such systems to a confluence of technical and scientific factors. CCD cameras, which suddenly were ubiquitous, inexpensive, and good enough for the work, combined with a ready palette of fluorescent dyes and proteins, supplied the technical know-how needed to get the job done. At the same time, a burgeoning recognition that researchers needed inexpensive, small animal analogs of clinical imaging techniques such as PET and CT, coupled with the longitudinal and cost benefits of whole-animal imaging, provided the scientific impetus.
"It all kind of came together in the early 2000s," Mansfield says.
For many researchers, in vivo imaging devices serve as a bridge between the research and clinical arenas. Animals, after all, are often used as models of human disease. And they are where potential drug compounds first are put through their paces.
"The thing that is driving this animal imaging is the belief that you can cut down the process of developing drugs and treatments by orders of magnitude," says M. Waleed Gaber, co-director of the Small Animal Imaging Facility at Texas Children's Hospital, part of the Baylor College of Medicine.
Researchers can employ a variety of imaging approaches, or modalities, for in vivo imaging, and Gaber's facility, which has been open for about eight months now, supports most of them. These include such small animal analogs of human medical technologies as a 9.4-Tesla magnetic resonance imaging scanner (Bruker Biospec), a combination micro-CT/PET/SPECT imager (Siemens Inveon), which performs both 3D x-ray analysis (CT) and imaging based on injected radioisotopes (PET/SPECT)), and a VisualSonics Vevo 770 ultrasound.
The facility also supports imaging modalities not found in the clinic. One is a homemade system for "intravital imaging" of "microvascular networks and tumor growth via cranial or dorsal windows," according to the facility's brochure.
Comprised of a microscope with long-working-distance objectives and a heavy-duty stage, as well as a pair of extremely sensitive CCDs (a Photometrics Evolve EMCCD Camera and CoolSnap ES), this system "is primarily a microscope that can handle the weight of an animal," explains Gaber, who put the device together to track radiation-induced damage to the blood-brain barrier.
The facility also has a Xenogen IVIS 100 System (Caliper Life Sciences) for detecting fluorescence and bioluminescence—one of a small but growing number of such devices that enables light-based small animal imaging.
According to William Kruka, senior vice president of corporate development and head of the imaging business unit at Caliper Life Sciences, the IVIS (in vivo imaging system) product line, which enables non-invasive optical molecular imaging, is just one part of what he calls the company's IIH (in vitro-to-in vivo-to-human) bridge. Comprising in vivo imagers and the company's automation and lab-on-a-chip in vitro testing and sample preparation systems, IIH is essentially a product pipeline intended to provide researchers with the tools needed to translate medical research from the early preclinical stage into the clinic.
"The idea is to look at the continuum of drug discovery and diagnostic testing, everything from what's done in a test tube all the way through to a human, and do what we can as a company to facilitate the highest quality, lowest cost tests across that spectrum," Kruka says.
Caliper's imaging product line includes such IVIS instruments as the Lumina (a two-dimensional fluorescence and bioluminescence imager), Lumina XR (which adds x-ray capabilities), Kinetic (for real-time video imaging), and Spectrum (for three-dimensional fluorescence and bioluminescence imaging). A more recent addition is the Caliper Quantum FX , capable of generating three-dimensional x-ray images (microCT).
Seth Gammon, in vivo product manager for Carestream Molecular Imaging, explains that multimodal imaging enables researchers to overlay and combine orthogonal datasets to provide a more comprehensive picture of underlying biology than any one modality alone, a concept called multimodal imaging.
Here's a simple example: Imagine injecting fluorescently labeled anti-tumor antibodies into the tail vein of a mouse. Researchers can use optical imaging to identify the location and size of tumors in situ, without having to open the mouse surgically. At the same time, an x-ray image can provide background anatomic detail.
"We can directly coregister the signals from all those different modalities in space," Gammon says. "If I have a fluorescent image, the image will have bright spots on a black background; it has no meaning. The researcher needs to understand where that signal is coming from."
Offering optical, radiographic (i.e., injected radioisotope), and x-ray capabilities, Carestream Molecular Imaging's In-Vivo Multispectral FX PRO can provide that understanding. But it adds something else, as well.
The system uses "excitation-side spectral unmixing" to deconvolve multiple fluorescence signatures, whether of multiple fluorophores, or to distinguish one fluorophore from background autofluorescence.
"The software actually peels apart what is the true fluorescent signal from autofluorescence," Gammon says.
CRI's Mansfield, whose Maestro system also provides multispectral imaging, says most biological samples autofluoresce. "It takes a lot of fluorophore to be seen above the background," he says. "It can be like trying to hear a conversation over a jackhammer."
The Maestro can resolve up to nine different fluorophores, and remove autofluorescence, by taking pictures at a variety of wavelengths and tracking each pixel over the resulting image "stack."
"It's not a simple process of spectral subtraction," he explains. "Rather, we go pixel by pixel and quantify how much signal came from autofluorescence and how much came from fluorophores."
Carestream offers another multimodal imager called the Albira, which focuses on non-optical detection modalities (PET/SPECT/CT). According to Gammon, the advantage of these imaging approaches is the constancy of the signal they produce. By contrast, light emission can be influenced by the environment.
"The advantage of optical systems is you can make activatable probes," Gammon says. "But the fact that you can change the light output makes it basically impossible to truly quantify an optical signal in vivo. With radio-tracers, you have a known specific activity, and the decay rate is not influenced by the environment."
Other in vivo imagers focus exclusively on one imaging mode. LI-COR Bioscience's laser-based Pearl Impulse, for instance, is designed specifically to image near infrared fluorescent dyes (such as the company's own IRDye fluorophores).
According to Jeff Harford, senior product marketing manager for molecular imaging at LI-COR, the system doesn't support bioluminescence because LI-COR has chosen to focus on what they believe to be "translatable" technologies. Bioluminescence, requiring as it does genetic modification of the organism with a luciferase-expression system, does not fit the bill, he says.
Harford says NIR fluorophores have a strong advantage over visible-light fluorophores when it comes to in vivo imaging: "At 800nm, there is very little tissue autofluorescence. The tissue is practically transparent."
Another company, Kubtec, focuses its imaging efforts on high-resolution x-ray imaging, says Marketing Manager Bridget Harrington. The resulting images, she says, are sharp enough to pick out fine detail in bones or tumors (such as vesseling in the marrow or microcalcifications in excised breast tissue) that would be missed by lower-resolution systems. "It's like building a house," Harrington says. "It's the difference between a rendering created to show what the house may look like when complete, versus blueprints that show the precise location of every element."
Not all in vivo imaging is done at the whole-animal level, of course. Multiphoton microscopy through surgical "windows" enables relatively deep tissue imaging of, for instance, the brain as animals learn or interact with their environment. And some companies, such as Olympus, have developed specialized "microprobe" optics to penetrate and image deep into tissue without first sacrificing the animal. Olympus' microprobe looks like a standard microscope objective with a long, pinkie-length projection.
"Imagine sticking your pinkie in pudding, as opposed to touching the surface," Product Marketing Director Edward Lachica says, likening the brain's consistency to that of a dessert. "The microprobe objective will cause damage, but you can get to deeper structures."
Whichever imaging approach you choose, as with any instrumentation decision, the key is good planning. Think about the experiments you need to do, and about how those needs might change moving forward. Just because a system is multimodal doesn't mean it excels at everything. For instance, if you want both bioluminescence and fluorescence, says Mansfield, "no system is perfect and you have to juggle which things are most important to you."