NIR-II fluorescence imaging using indocyanine green nanoparticles

Abstract

Fluorescence imaging in the second near-infrared window (NIR-II) holds promise for real-time deep tissue imaging. In this work, we investigated the NIR-II fluorescence properties of a liposomal formulation of indocyanine green (ICG), a FDA-approved dye that was recently shown to exhibit NIR-II fluorescence. Fluorescence spectra of liposomal-ICG were collected in phosphate-buffered saline (PBS) and plasma. Imaging studies in an Intralipid^® phantom were performed to determine penetration depth. In vivo imaging studies were performed to test real-time visualization of vascular structures in the hind limb and intracranial regions. Free ICG, NIR-I imaging, and cross-sectional imaging modalities (MRI and CT) were used as comparators. Fluorescence spectra demonstrated the strong NIR-II fluorescence of liposomal-ICG, similar to free ICG in plasma. In vitro studies demonstrated superior performance of liposomal-ICG over free ICG for NIR-II imaging of deep (≥4 mm) vascular mimicking structures. In vivo, NIR-II fluorescence imaging using liposomal-ICG resulted in significantly (p < 0.05) higher contrast-to-noise ratio compared to free ICG for extended periods of time, allowing visualization of hind limb and intracranial vasculature for up to 4 hours post-injection. In vivo comparisons demonstrated higher vessel conspicuity with liposomal-ICG-enhanced NIR-II imaging compared to NIR-I imaging.

Figure 1

NIR-II emission spectra of liposomal-ICG (Lip-ICG) and free ICG. Lip-ICG and free ICG were diluted in either (A) phosphate buffered saline or (B) bovine plasma. Emission spectra were acquired for three ICG concentrations using an excitation wavelength of 782 nm.

Figure 2

Determination of penetration depth for NIR-II imaging of liposomal-ICG (Lip-ICG) in an Intralipid^® phantom. (A) Comparison of Lip-ICG and free ICG in NIR-II window. Full-width-half-maximum (FWHM) of capillary glass tube filled with Lip-ICG or free ICG as a function of depth in a 1% Intralipid® phantom. FWHM for Lip-ICG were significantly different (p < 0.05) from free ICG at corresponding concentrations, at 4 and 5 mm depths. (B) Comparison of Lip-ICG in NIR-I and NIR-II window. FWHM of capillary glass tube filled with liposomal-ICG as a function of depth in Intralipid®. FWHM NIR-II window were significantly different (p < 0.05) from NIR-I window at their corresponding concentrations, at 4 and 5 mm depths. (C) Representative fluorescence images of glass capillary filled with either Lip-ICG or free ICG at depths of 2 and 4 mm in 1% Intralipid®. Scale bars represent 10 mm. Samples were prepared by diluting Lip-ICG or free ICG in plasma.

Figure 3

In vivo NIR-II imaging of hind limb vasculature. Representative coronal images of hind limb region demonstrating NIR-II imaging of vasculature at various time points after intravenous administration of either liposomal-ICG (top row) or free ICG (bottom row). At 60 min and 240 min, the femoral vessel (yellow arrow) is only visible in liposomal-ICG images.

Figure 4

Contrast-to-noise ratio (CNR) for NIR-II imaging of hind limb vasculature. Normalized CNR values of femoral vessel in the hind limb region in NIR-II images acquired with either liposomal-ICG or free ICG. CNR values were determined for NIR-II images acquired at various time points after administration of contrast agent. CNR values were normalized to ICG dose (mg) per unit body weight (kg). CNR values for liposomal-ICG were significantly different (p < 0.05) from free ICG at all time points.

Figure 5

In vivo NIR-II imaging of brain vasculature through an intact skull. Representative coronal images of mouse brain demonstrating NIR-II imaging of vasculature at various time points after intravenous administration of either liposomal-ICG (top row) or free ICG (bottom row). At 60 min and 240 min, the transverse sinuses (blue arrows) and the sagittal sinus (red arrow) are only visible clearly in NIR-II images acquired with liposomal-ICG.

Figure 6

Contrast-to-noise ratio (CNR) for NIR-II imaging of blood vessels within the brain. Normalized CNR values of transverse sinus in NIR-II images acquired with either liposomal-ICG or free ICG. CNR values were determined for NIR-II images acquired at various time points after administration of contrast agent. CNR values were normalized to ICG dose (mg) per unit body weight (kg). CNR values for liposomal-ICG were significantly different (p < 0.05) from free ICG at all time points.

Figure 7

Comparison of NIR imaging and CT angiography (CTA) for visualization of hind limb vasculature. Representative coronal images of mouse hind limb region demonstrating the visualization of femoral vessel in (A) NIR-I, (B) NIR-II, and (C) CTA. (D) Normalized signal intensity profile across the cross-section of a femoral vessel in CTA, NIR-I and NIR-II window. Full width half maximum (FWHM) values are reported adjacent to each of the profiles. Note that the vessel profile in CT matches closer to the profile in NIR-II than in NIR-I. Scale bars represent 5 mm.

Figure 8

Comparison of NIR imaging and MR angiography (MRA) for visualization of intracranial vasculature. Representative coronal images of mouse brain demonstrating the visualization of intracranial vessels in (A) NIR-I, (B) NIR-II, and (C) MRA. (D) Normalized signal intensity profile across the cross-section of a sagittal sinus in NIR-I and NIR-II imaging window. Full width half maximum (FWHM) values are reported adjacent to each of the profiles. (E) Sagittal view of mouse brain demonstrating the depth of intracranial vessels from the top. All images were acquired in the same animal. Scale bars represent 5 mm.