Table 2.
Summary of imaging methods discussed including resolution, depth of imaging, contrast used & safety, quantitation accuracy, clinical translation, and the tumor microenvironment factors the imaging methods have been used to investigate.
| Imaging method | Resolution | Imaging depth | Contrast agents & safety | TME factors | Accuracy in quantitation | Clinical translation | |
|---|---|---|---|---|---|---|---|
| Magnetic resonance | Magnetic resonance imaging[375,376] | 25-100 mm | Whole body | Gd- or iron-oxide-based probes and dendrimer-based macromolecules; conventional MRI is safe for imaging for patients without embedded metals in their body while GD-based contrast may cause some adverse health concerns | ECM proteins, matrix metalloproteinase, mesenchymal stromal cells, cancer associated fibroblasts, immune cells, tumor vasculature, metabolic-choline-phospholipid metabolism, hypoxia, pH, and tumor stroma interactions | Only semi-quantitative, relying on regional differences in signal intensities, and primarily used to reveal gross morphological abnormalities | Currently clinically used technique, making future TME studies feasible; however, some preclinical studies using higher strength magnetic fields may pose challenges |
| Nuclear | SPECT[375,376] | 1-2 mm | Whole body | Radiolabeled antibodies, antibody fragments, and antigens; SPECT requires exposure to radiation | Matrix metalloproteinase, mesenchymal stromal cells, and immune cells | Although clinically only semiquantitative, quantitation is still a key benefit of nuclear medicine and can be improved with quantitative SPECT (only tested preclinically) | Currently a clinically available technique; however, clinical translation suffers from increased attenuation and decreased resolution |
| PET[375,376] | 1-2 mm | Whole body | Radiolabeled antibodies, antibody fragments, and nutrients, as well as activatable probes; PET requires exposure to radiation | Matrix metalloproteinase, mesenchymal stromal cells, immune cells, tumor vasculature, glycolysis, hypoxia, pH, and tumor stroma interactions | Similar to SPECT, quantitative accuracy can be improved with advanced algorithms, but these have only been applied in preclinical imaging | Currently clinically used for evaluation of therapy response, making clinical translation feasible | |
| X-Rays | Computed tomography[375,376] | 50-200 mm | Whole body | Water-soluble iodinated probes; requires exposure to radiation | Immune cells, tumor vasculature, hypoxia, and pH | Accurate quantitation is a benefit of CT | Clinical translation will be limited by approval of appropriate contrast agents |
| Ultrasound | Ultrasound[289,375] | 50-500 mm | mm-cm | Endogenous, targeted microbubbles; no safety concerns with conventional ultrasound and microbubble contrast is FDA approved for cardiac[377] and liver[378] imaging | Immune cells and tumor vasculature | Generally qualitative but use of contrast and mathematical algorithms have improved ability for quantitation | Clinically available technique where microbubble contrast is FDA approved for other types of US imaging, making translation of preclinical techniques feasible |
| Optical | Photoacoustic imaging[379] | 5-300 mm | 0.7-40 mm | Fluorophores, nanoparticles, and quantum dots; no safety concerns, can be used repeatedly on tissue | ECM proteins, immune cells, tumor vasculature, and pH | Generally qualitative but several studies have shown realization of quantitative information | Size and cost of laser sources, building a prototype, and clinical trials limit translation |
| Intravital microscopy[376,380] | 100 nm - 1 mm | 100-300 mm | Endogenous; requires a surgically implanted window | Immune cells, tumor vasculature, and tumor stroma interactions | Quantitation is difficult; however, methods are being developed to improve ability for quantitation | Only feasible for introperative guidance | |
| Bioluminescence imaging[375,376] | 3-5 mm | 1-2 cm | Reporter genes; requires use of lentiviral vectors, although toxicity is low | Matrix metalloproteinase, mesenchymal stromal cells, immune cells, tumor vasculature, and pH | Quantitation is difficult, preclinical methods are being developed for improvements | Only ideally used as a preclinical technique, human tumors do not express luciferases | |
| Fluorescence imaging[375,376] | 2-3 mm | < 1 cm | Fluorophores, fluorescent nanoparticles; fluorescent imaging in the near infrared is biologically safe and fluorescent particles show little to no toxicity | Matrix metalloproteinase, mesenchymal stromal cells, immune cells, tumor vasculature, pH, and tumor stroma interactions | Quantitation is difficult and is only relative to other regions in the tissue | Translation feasibility lies in fluorescence guided surgery, in which clinical trials have been completed | |
| Fluorescence molecular tomagraphy[352] | < 1 mm | 1-2 mm | Near infrared dyes, quantum dots, and reporter genes; imaging with near infrared light is biologically safe; however, quantum dot composite material is toxic in elemental forms | Matrix metalloproteinase | Improves quantitation over fluorescence imaging but is still challenging | Clinical translation limited by the development of scanners for smaller and more superficial structures | |
| Optical coherence tomography[359] | 1-15 mm | 2-3 mm | Endogenous; no biological safety concerns | ECM proteins, immune cells, and tumor vasculature | Typically, quantitation requires specialized proprietary software, but methods are being developed to overcome this | Currently clinically used for ophthalmic imaging, clinical translation for TME imaging depends on development of compact probes | |