There are several advantages to using silica nano-particle-based dual-modality 99mTc/ICG NIR imaging. Preoperative lymphoscintigraphy can reveal deeply situated sentinel lymph nodes, and intraoperative NIR fluorescence imaging provides simultaneous visualization at anatomical resolution in the surgical field following a single injection. These features may facilitate the sentinel lymph node biopsy procedure. Furthermore, it is well known that ambiguous lymphatic drainage results in atypical sentinel lymph node locations [2, 18]. Nevertheless, in the cases of unpredictable lymph node drainage basins, surgeons are capable of identifying the sentinel lymph nodes by preoperative lymphoscintigraphy and/or intraoperative surveillance using radiation with greater penetration. In this study, the sentinel lymph nodes were imaged using both radioactivity and NIR fluorescence in the surgical field.
The current study used dual-modality imaging with PCSN probes labeled with 99mTc and NIR fluorescence to evaluate sentinel lymph nodes in the recipient animals. The results demonstrate the potential for these approaches to detect sentinel lymph nodes during clinical biopsy procedures. However, in this animal imaging study, the concentration of the imaging agents substantially affected the image quality. Similarly, it has been reported that very little radioactivity accumulates in the sentinel lymph nodes of human patients . For example, in breast cancer specimens, the administered activities range from 10 to 108 MBq . Although the injected radiopharmaceuticals and the timing of the measurements were different, the mean percentages of the injected doses per lymph node were reported to be 0.9%  and 0.60%  in patients with breast cancer and 0.34% to 0.92% in studies of cutaneous melanoma [22, 23].
In one rabbit study, 5% to 9% (198Au-colloid and 99mTc-antimony sulfur colloid) of the injected dose per node was detected . In another rabbit study using 18.5 MBq of 99mTc-phytate, the percentage of the injected dose was 0.21% ± 0.18% (mean ± SD) for lymph nodes in the neck (ranging from 0.01% to 0.62%) . In the current rat study, nodes showing increased uptake contained 0.109% ± 0.067% (mean % per node) of the initial dose of 7.4 MBq. We estimated that, on average, approximately 7.4 KBq of activity accumulated per node. Although this activity is quite small in comparison to that used in clinical practice, the sentinel lymph nodes were accurately visualized using this dose. However, higher doses of 99mTc would have provided much clearer images of the lymph nodes.
We used an NIR imaging video camera system for the in vivo imaging; this system has an excitation wavelength of 760 nm and a fluorescence emission wavelength of 820 nm. Before each imaging study, we confirmed that dissolved ICG-sulfo-OSu had bound to the nanoparticles by TLC to verify the absence of free binding spots or flares of fluorescence. The rats were administered a single 0.1-ml submucosal injection of the PCSN probes containing 0.9 μg of ICG-sulfo-OSu. In previous clinical studies, 2.5 mg of ICG was injected for sentinel lymph node mapping using only NIR fluorescence [6, 25]. Furthermore, the injection volume was approximately 21- to 210-fold larger than the volume used in our animal experiments. This comparison was calculated using weights of 60 kg for human subjects and 500 g for rats. In addition, the amount injected into the rats was also much less than the amount used in other animal sentinel lymph node studies . However, the comparatively small amount of both the 99mTc and the ICG agents used with the PCSN probes in this study depicted the sentinel lymph nodes successfully.
For sentinel lymph node detection, the size of the radiopharmaceutical is critical for optimal imaging. Multiple sizes have been used for sentinel node detection, including approximately 50- to 100-nm radiolabeled colloid, 50- to 200-nm 99mTc-sulfur colloid, or 4- to 100-nm (95% of the particles <80 nm) 99mTc-nanocolloid of human serum albumin . The PCSN size used in the current study was approximately 30 nm, which is within these previously established ranges. Radiopharmaceuticals have been shown to accumulate, primarily via phagocytosis, in macrophages lining the sinusoid spaces in sentinel lymph nodes . It has been reported that silica-coated fluorescent nanoparticles with a 40-nm diameter were observed in macrophages and dendritic cells of sentinel lymph nodes; phagocytosis of these nanoparticles results in fluorescence enhancement in the lymph nodes . We also detected PCSNs in macrophages, which suggest that retention of these particles may be the result of macrophage phagocytosis.
Optical imaging is typically limited by tissue penetration depth and relatively poor quantification. Fluorescence imaging is limited to tissue depths of 1 to 2 mm in the visible range and as much as 1 to 2 cm in the NIR range [28, 29]. Clinically, the detection of the sentinel lymph node using ICG was restricted to 5 mm of tissue depth in oropharyngeal cancer patients . With a wavelength ranging from 750 to 1,000 nm, infrared luminescent light is not directly visible to the human eye. Hemoglobin, water, lipids, and other endogenous chromophores, such as melanin, have their lowest absorption levels within the NIR spectrum, which permits increased photon depth penetration into the tissues . Furthermore, in the NIR spectrum, the autofluorescence emitted from endogenous fluorophores, such as NADH and porphyrins, is extremely low compared with UV and visual light wavelengths, which provides an improved signal-to-background ratio. By contrast, ICG emits infrared fluorescence.
Successful combined imaging using 99mTc and ICG fluorescence for sentinel biopsy of the lymph node has been reported recently [10–12]. Incubation of 99mTc-nanocolloid and ICG provided a favorable agent retention time, with the fluorescence lasting up to 19 h post-injection, for the detection of sentinel lymph nodes in the necks of patients with squamous cell carcinoma of the oral cavity . In addition, another new NIR imaging approach to visualize sentinel lymph nodes has been demonstrated in animal studies using nano-carriers, such as quantum dots [31, 33–35] and dendrimers [36–38].
Silica nanoparticles have also been used for pre-clinical in vivo fluorescence imaging ; these particles are also attractive for approaches aiming to integrate multiple properties for multimodality biomedical imaging . Furthermore, functionalized organically modified silica nanoparticles can serve as an ideal nano-platform for the assembly of multimodal nanoparticles for targeted diagnostics and therapeutics . Benezra et al. used dye-encapsulating silica particles that were surface-functionalized with cyclic arginine-glycine-aspartic acid peptide ligands and radioiodine, which exhibited high-affinity/avidity binding to αvβ3 integrin in mouse melanoma .
The Aerosil 200 used in this study is an amorphous, fumed hydrophilic silica. Of the reported nanoparticles used for imaging, amorphous silica nanoparticles are feasible materials for in vivo imaging due to their biocompatible nature compared to other agents, including quantum dots, which contain cadmium telluride crystals. However, it has been reported that amorphous silica and crystalline silica are phagocytosed and are therefore toxic to mouse alveolar macrophage cells. In addition, the inhalation of crystalline silica has been shown to cause silicosis . Nevertheless, the cytotoxicity induced by nano-scale silica particles is complex and depends on the particle size, the surface functionalization, and the cell line used for the study of toxicity .
The PCSNs used in this study were functionalized using PAMAM, which is a type of dendrimer that is characterized as a branched synthetic molecule composed of spherical nano-structures with a high degree of molecular uniformity. Importantly, PAMAM possesses the potential to contribute to in vivo imaging and drug delivery vehicles as the dendrimer can be used to generate highly functionalized terminal surfaces . We used the terminal NH2 PAMAM for covalent bonding between the PCSN surfaces and ICG with simultaneous 99mTc chelation. Thus, our PCSN probe may help to convey drugs, antibodies, and ligand molecules by acting as a targeted carrier. It is known that the flexibility of dendrimers, along with their size, charge, and nanoparticle hydrophilicity, influences their imaging potential by changing their biodistribution and pharmacokinetics . It is assumed that because our PCSNs comprise the silica core and surface dendrimers, the PCSN probes may exhibit stable biological characteristics due to the lower flexibility of the dendrimers in contrast to probes solely composed of PAMAM dendrimers. However, how this difference would affect in vivo imaging is not well known. Further elucidation will require more study.