Introduction: Fluorescence-guided surgery has emerged as a powerful tool to detect, localize and resect tumors in the operative setting. Our laboratory has pioneered a novel way to administer an FDA-approved near-infrared (NIR) contrast agent to help surgeons with this task. This technique, coined Second Window ICG, exploits the natural permeability of tumor vasculature and its poor clearance to deliver high doses of indocyanine green (ICG) to tumors. This technique differs substantially from established ICG video angiography techniques that visualize ICG within minutes of injection. We hypothesized that Second Window ICG can provide NIR optical contrast with good signal characteristics in intracranial brain tumors over a longer period of time than previously appreciated with ICG video angiography alone. We tested this hypothesis in an intracranial mouse glioblastoma model, and corroborated this in a human clinical trial.
Methods: Intracranial tumors were established in 20 mice using the U251-Luc-GFP cell line. Successful grafts were confirmed with bioluminescence. Intravenous tail vein injections of 5.0 mg/kg (high dose) or 2.5 mg/kg (low dose) ICG were performed. The Perkin Elmer IVIS Spectrum (closed field) was used to visualize NIR fluorescence signal at seven delayed time points following ICG injection. NIR signals were quantified using LivingImage software. Based on the success of our results, human subjects were recruited to a clinical trial and intravenously injected with high dose 5.0 mg/kg. Imaging was performed with the VisionSense Iridium (open field) during surgery one day after ICG injection.
Results: In the murine model, the NIR signal-to-background ratio (SBR) in gliomas peaks at one hour after infusion, then plateaus and remains strong and stable for at least 48 hours. Higher dose 5.0 mg/kg improves NIR signal as compared to lower dose at 2.5 mg/kg (SBR = 3.5 vs. 2.8; P = 0.0624). Although early (≤ 6 hrs) visualization of the Second Window ICG accumulation in gliomas is stronger than late (≥24 hrs) visualization (SBR = 3.94 vs. 2.32; p<0.05) there appears to be a long plateau period of stable ICG NIR signal accumulation within tumors in the murine model. We call this long plateau period the "Second Window" of ICG. In glioblastoma patients, the delayed visualization of intratumoral NIR signal was strong (SBR 7.50 ± 0.74), without any significant difference within the 19 to 30 hour visualization window (R2 = 0.019).
Conclusion: The Second Window ICG technique allows neurosurgeons to deliver NIR optical contrast agent to human glioblastoma patients, thus providing real-time tumor identification in the operating room. This nonspecific tumor accumulation of ICG within the tumor provides strong signal to background contrast, and is not significantly time dependent between 6 hours to 48 hours, providing a broad plateau for stable visualization. This finding suggests that optimal imaging of the "Second Window of ICG" may be within this plateau period, thus providing signal uniformity across subjects.