Introduction: Complete surgical resection resection of malignant tumors ultimately results in improved patient prognosis. Since surgery is the primary treatment for most tumors, technologies that indicate the presence of tumor, intraoperatively, could greatly reduce the risk of recurrent disease. We previously reported development of an image-guided surgery (IGS) system that detects near infrared fluorophores[1], e.g. ICG and nanoparticles that entrap of ICG for image-guided surgery[2], which are based on hyaluronic acid (HA) that is modified with a hydrophobic moiety, self-assemble in aqueous conditions, and entrap ICG). Here, we examine the effect of hydrophobic moiety chemical structure, time-dependency of tumor contrast in vivo, and tumor detection depth in a surgical simulation model.
Materials and Methods: Amino-propyl-pyrenebutanamide (PBA) or amino-propyl-5-β-cholanamide (5βCA) was conjugated to HA (10 kDa) at 30 wt%. ICG was loaded by dialyzing in DMSO:H2O and then H2O, yielding NanoICG, Hydrodynamic diameter (HD) was measured using dynamic light scattering. Optical properties were determined by absorbance and fluorescence spectroscopy. Under an IACUC approved protocol, NanoICG NPs or ICG were injected at 310 ug/kg ICG via a tail vein. At either 4 or 24 h, mice were euthanized. Signal to noise ratio (SNR) and tumor contrast (relative to surrounding tissue) to noise ratio (CNR) were determined using a LI-COR Pearl Impulse Animal Imaging system and an in-house built IGS system.[1] Resected tumor xenografts were then imbedded at various depths in a breast phantom with optical properties that mimic human breast tissue[3], providing the estimated depth that NanoICG could be detected in human tissue.
Results: NanoICG HD ranged from 40-70 nm and was not dependent on the hydrophobic moiety conjugated to HA. SNR depended on whether ICG was in the HA-derived NP and time (Fig. 1), but not on the hydrophobic moiety used to drive self-assembly. ICG had higher tumor SNR at 4 h, but NanoICG had higher tumor SNR at 24 h, along with higher liver and spleen SNR. Average CNR was similar between all agents at 4 h and had high SD. At 24 h, SD decreased and NanoICG CNR was more than double ICG CNR. In the breast phantom, tumors could be detected at greater than 5 mm using the LI-COR and IGS system, where NanoICG resulted in tumors that could be detected deeper than those enhanced with ICG (Fig. 2).
Discussion: Similar SNR between ICG and NanoICG, initially, is likely due to initial bloodflow to the tumor. The increased NanoICG at 24 h is likely due to slower clearance and because ICG may initially be quenched fluorescence in the NP state.[2] Thus, the increasing tumor contrast over time could be due to ICG release in tumor cells or microenvironment. Higher signal in liver and spleen is indicative of RES clearance associated with NP formulations. Xenograft tumor models, while useful for initial evaluation of relative tumor accumulation of contrast agent, are not ideal to determine depth at which a contrast agent can be detected. Thus, harvesting a contrast-enhanced tumor and placing that in a phantom provides critical information on the size tumor that can be detected at a given depth.
Conclusion: NanoICG increases tumor contrast and depth that it can be detected compared to ICG alone. Future studies are underway to fully understand why NIR signal remains strong over time. Use of immunocompetent mice for tumor enhancement and biodistribution studies are planned.
This work was supported in part by the National Institutes of Health; P30 CA012197 (Wake Forest University Comprehensive Cancer Center), P30 CA036727 (Fred and Pamela Buffett Cancer Center, UNMC), R00 CA153916 (AMM), and R01 EB019449 (AMM) and the University of Nebraska Medical Center.
References:
[1] Mohs AM, et al. IEEE Trans Biomed Eng 2015, 62, 1416-24.
[2] Hill TK, et al. Bioconjug Chem 2015, 18, 294-303.
[3] Pleijhuis, et al. Eur J Surg Oncol 2011, 37, 32-39.