High-resolution, 3D multimodal imaging technology enables molecular, anatomical, and functional co-registration of features.
MARK LITTLE PHOTOSOUND TECHNOLOGIES INC.
The development of anticancer metastatic therapies for human clinical trials requires meticulous evaluation of efficacy and optimization of small animal test models in preclinical experimentation1. Critical information on morphology and the molecular microenvironment of tumors is currently obtained and monitored using noninvasive, in vivo imaging methods. Detection of individual, small tumors (separated by 2 mm or less) pushes the limits of small animal imaging modalities currently on the market, which can prohibit noninvasive quantification of the volume of metastatic lesions.
Photoacoustic and optical images of the ventrodorsal view of a mouse specimen: 532 nm (yellow, skin level) + 890 nm (red, deep anatomy) (a); 532 nm (yellow, skin level) + 780 nm (green, ICG detecting) (b) and 532 nm (yellow, skin level) + 780 nm (green, ICG detecting) + FMT (blue, 780-nm excitation) (c). Injection site, right subiliac lymph node drainage from injection site, and vertebrae (for reference) labeled 1, 2, and 3, respectively. Courtesy of the Ultrasound Imaging and Therapeutics Research Laboratory at Georgia Tech.
Commercial small animal imaging platforms that enable rendering and anatomical registration of metastatic lesions with true 3D isotropic submillimeter spatial resolution would help solve this problem. Such platforms would alleviate subjective interpretation and provide molecular and functional information on blood content. A multimodal approach could provide a more cost-effective implementation of all these features in a single configuration.
Metastatic therapy analyses
If preclinical researchers were asked to provide a wish list of anticancer metastatic therapy analyses, three main types would be included: molecular, anatomical, and functional analysis. Preferably, all analyses should be performed on a live specimen close to its natural physiological state. The first type, molecular analysis, involves imaging of features based on the identification and location of natural constituents such as blood oxygenation levels or injected therapeutic compounds. Tumor growth mechanisms and a drug’s ability to inhibit tumor growth are important results found during molecular analysis. Most molecular imaging of live animals is poorly resolved and shows the location of constituents in a general area of the specimen. Typically, confirmation of whether a constituent is present can be gained from molecular analysis.
Identification of constituents in a general area does not provide a complete picture unless the researcher knows precisely where in the specimen’s anatomy the constituent is concentrating. Anatomical analyses, the second type, provide finely resolved images of vasculature, internal organs, and bone. The high resolution required to image these features limits the total field of view (FOV) and increases the analysis time. However, the benefits of being able to take a highly resolved 3D picture of a specimen’s internal physiology in vivo cannot be overstated.
The third important type of analysis is functional. Functional analyses address the “how” of cancer metastasis and drug treatment. How does a tumor form and grow? How long does a drug take to reach the tumor site after administration? How are nutrients being delivered to the tumor? And what, if anything, hinders their delivery? Answering these questions is a much more difficult task for in vivo imaging, as one is not just taking an image in time but taking a series of images over time (for example, oxygen levels in blood flow) and relating changes in each image that contribute to tumor metastasis.
Experimental scheme for 3D in vivo lymphatic mapping and detection of regional tumor metastasis. Courtesy of PhotoSound Technologies Inc./Ultrasound
Imaging and Therapeutics Research Laboratory at Georgia Tech.
Within each of these analysis types, the resolution of fine biological features in real time is required to observe the effects of therapies at the earliest stages of cancer metastasis. To resolve features on the submicron cellular level, technologies are currently limited to histological samples that require sacrificing the specimen, excising the region of interest, and placing it under a microscope. However, in vivo imaging techniques are now pushing the envelope, and sub-100-μm spatial resolution of anesthetized specimens is possible.
Current imaging techniques
Preclinical imaging laboratories in academic core facilities and contract research organizations (CROs) incorporate several different imaging modalities by purchasing separate instruments that specialize in one of the core analysis types described above. Five main imaging modalities stand out: optical, nuclear, acoustic, x-ray, and MRI. Relevant features of each modality are summarized in the table below. Each modality delivers its designated analysis type; however, imaging with various types requires moving the specimen from instrument to instrument. Numerous drawbacks to specimen transport include but are not limited to: increased analysis time, loss of imaging and injection reproducibility, difficulty in co-registering image features, changes to specimen morphology, chance of the specimen regaining consciousness, and ethical consequences of administering multiple injections and anesthesia sessions to the animal. In addition, the total laboratory space, purchasing budget, and cost of ownership is quite high if the research facility desires use of multiple modalities.
3D photoacoustic tomography imaging of mouse vasculature at various angles of rotation. Courtesy of the Ultrasound Imaging and Therapeutics Research
Laboratory at Georgia Tech.
While an exhaustive comparison of each modality is not within the scope of this article, the information in the table highlights obvious reasons that some modalities are better than others for certain needs. An approach that can combine multiple modalities in a single instrument would be very beneficial. For instance, achieving the high resolution of nuclear or CT imaging without ionizing radiation, while adding molecular analysis capabilities from optical imaging in a smaller and lower-cost form, would benefit early detection of cancer metastasis and drug therapy development.
Comparison of Various In Vivo Small Animal Imaging Modality Features
FMT: Fluorescence Molecular Tomography
A multimodal revolution
Optical (fluorescence and bioluminescence)2,3 and ultrasound4,5 in vivo imaging methods have found great popularity among researchers because the methods have enabled affordable and sensitive molecular and anatomical imaging tools for preclinical studies and development of animal models. However, their stand-alone application is impeded by poor spatial resolution and/or limitations imposed by two-dimensionality of the images.
A high-resolution in vivo 3D imaging method, which could be easily integrated with optical imaging in a single instrument, would have a great impact. Photoacoustic tomography is an emerging whole-body 3D imaging modality capable of 150- to 500-μm resolution without the use of ionizing radiation6,7. In addition, optical imaging uses many of the same components (e.g., a tunable laser system) for fluorescence excitation and generation of the photoacoustic effect. A multimodal approach provides a way to defeat shortcomings of each individual modality and enables high-resolution whole-body 3D imaging of fluorescent biomarkers by integrating fluorescence and photoacoustics in a single co-registered modality8-10.
PhotoSound TriTom imaging platform based on photoacoustic and fluorescence technology (PAFT).
Courtesy of PhotoSound Technologies Inc.
Combining multiple modalities into a single platform to enable multiple analysis types is not a simple task. The combination must be completely transparent to the end user, with no obvious separation between modalities, so that instrument operation is seamless. As is usually the case with push-button, easy-to-use instrumentation, the end user cares only about the final answer and not the idiosyncrasies of hardware/software engineering. While a number of factors should be considered, three will be briefly discussed here: user interface, mode switching, and image co-registration.
The user interface should consist of one main window containing all elements that control sample manipulation, data acquisition, data storage, experimental parameters, and image reconstruction. Other programs can be called upon to perform other functions, but the main window should have controls for launching these programs without the end user having to navigate to another location on the computer. The best data is obtained from efficient workflows that move the end user from starting a scan, to previewing the result for parameter optimization, if needed, and then finally to saving the data with all relevant instrument parameters. If the end user must launch separate applications for different imaging modes, then the workflow is interrupted. The time it takes to switch applications can lead to the specimen regaining consciousness or to problems co-registering images.
Switching between imaging modes should occur automatically. There should be no hardware changes, add-ons, or transport of the specimen to another location. The time between switching should be zero (simultaneous data acquisition between modes), or as short as possible.
Finally, co-registration of spatial imaging coordinates between images captured from each mode must be precisely aligned and should not require end user action. Toggling between image layers of each mode allows rich visualization of analysis types. For example, cancer metastasis target sites can be found with optical imaging anywhere in the whole specimen body, overlapped on highly resolved anatomical features obtained by photoacoustic imaging.
Advancements in imaging technology and the combining of multiple modalities into a single instrument are enabling in vivo images with simultaneous anatomical and molecular information from target features such as early cancer metastasis. The extra information and 3D imaging are pushing the envelope of image viewing software, with some end users turning to virtual reality to see all features in an informative way. However, without considering the above hardware/software challenges, the benefits of multimodal technology cannot be realized.
Changes to labs coming
A combination of platforms exists that can cover many of the benefits described above, such as ultrasound, nuclear imaging, optical imaging, MRI, and x-ray/CT. A preclinical imaging facility equipped with these platforms is extremely expensive (up to $1 million per instrument). A single, relatively inexpensive, multimodal platform that is easy to use can serve many purposes in one location. The platform could include a single workstation lab bench design. The platform would have tools and stations for small animal preparation from sample restraining, contrast agent injection, multi-inlet anesthesia exposure (induction box, sample restrainer, and imaging chamber connections), and full life-support monitoring. Next to the preparation station would be the imaging chamber equipped with photoacoustic and optical imaging detectors. The simplified operation and easy-to-use software would combine multiple imaging platforms into one compact version that could be operated by any biologist studying early cancer metastasis and developing new antimetastatic therapies to inhibit or stop tumor growth.
Meet the author
Mark Little, Ph.D., is an application scientist and marketing director. He received a doctorate in analytical chemistry, investigating fundamental processes and novel applications of tunable and fixed laser systems using mass spectrometry for the analysis of biomolecular samples. Little is currently the director of business development at PhotoSound Technologies Inc.
References
1. R.S. Balaban and V.A. Hampshire (2001). Challenges in small animal noninvasive
imaging. ILAR J, Vol. 42, Issue 3, pp. 248-262.
2. R. De Souza et al. (2015). Preclinical imaging and translational animal models of cancer for accelerated clinical implementation of nanotechnologies and macromolecular agents. J Control Release, Vol. 219,
pp. 313-330.
3. V. Ntziachristos et al. (2003). Fluorescence imaging with near-infrared light: new technological advances that enable in vivo molecular imaging. Eur Radiol, Vol. 13, Issue 1, pp. 195-208.
4. F. Foster et al. (2009). A new 15-50 MHz array-based micro-ultrasound scanner for preclinical imaging. Ultrasound Med Biol, Vol. 35, Issue 10, pp. 1700-1708.
5. X. Wang et al. (2012). Novel single-chain antibody-targeted microbubbles for molecular ultrasound imaging of thrombosis: validation of a unique noninvasive method for rapid and sensitive detection of thrombi and monitoring of success or failure of thrombolysis in mice. Circulation, Vol. 125, Issue 25, pp. 3117-3126.
6. R. Su et al. (2012). Three-dimensional optoacoustic imaging as a new noninvasive technique to study long-term biodistribution of optical contrast agents in small animal models. J Biomed Opt, Vol. 17, Issue 10,
p. 101506.
7. H. Wang et al. (2014). In vivo photoacoustic molecular imaging of breast carcinoma with folate receptor-targeted indocyanine green nanoprobes. Nanoscale, Vol. 6, Issue 23,
pp. 14270-14279.
8. C. Kim et al. (2010). Sentinel lymph nodes and lymphatic vessels: noninvasive dual-modality in vivo mapping by using indocyanine green in rats — volumetric spectroscopic photoacoustic imaging and planar fluorescence imaging. Radiology, Vol. 255, Issue 2, pp. 442-450.
9. L. Xi et al. (2013). HER-2/neu targeted delivery of a nanoprobe enables dual photoacoustic and fluorescence tomography of ovarian cancer. Nanomedicine, Vol. 10,
Issue 3, pp. 669-677.
10. B. Wang et al. (2012). Photoacoustic tomography and fluorescence molecular tomography: a comparative study based
on indocyanine green. Med Phys, Vol. 39, Issue 5, pp. 2512-2517.