Indocyanine green fluorescence in second near-infrared (NIR-II) window

Abstract

Indocyanine green (ICG), a FDA approved near infrared (NIR) fluorescent agent, is used in the clinic for a variety of applications including lymphangiography, intra-operative lymph node identification, tumor imaging, superficial vascular imaging, and marking ischemic tissues. These applications operate in the so-called "NIR-I" window (700-900 nm). Recently, imaging in the "NIR-II" window (1000-1700 nm) has attracted attention since, at longer wavelengths, photon absorption, and scattering effects by tissue components are reduced, making it possible to image deeper into the underlying tissue. Agents for NIR-II imaging are, however, still in pre-clinical development. In this study, we investigated ICG as a NIR-II dye. The absorbance and NIR-II fluorescence emission of ICG were measured in different media (PBS, plasma and ethanol) for a range of ICG concentrations. In vitro and in vivo testing were performed using a custom-built spectral NIR assembly to facilitate simultaneous imaging in NIR-I and NIR-II window. In vitro studies using ICG were performed using capillary tubes (as a simulation of blood vessels) embedded in Intralipid solution and tissue phantoms to evaluate depth of tissue penetration in NIR-I and NIR-II window. In vivo imaging using ICG was performed in nude mice to evaluate vascular visualization in the hind limb in the NIR-I and II windows. Contrast-to-noise ratios (CNR) were calculated for comparison of image quality in NIR-I and NIR-II window. ICG exhibited significant fluorescence emission in the NIR-II window and this emission (similar to the absorption profile) is substantially affected by the environment of the ICG molecules. In vivo imaging further confirmed the utility of ICG as a fluorescent dye in the NIR-II domain, with the CNR values being ~2 times those in the NIR-I window. The availability of an FDA approved imaging agent could accelerate the clinical translation of NIR-II imaging technology.

Fig 1

Schematic of imaging setups, (A) imaging setup for tissue phantom experiments (top view), (B) imaging setup for Intralipid^® phantom and in-vivo experiments (side view).

Fig 2

Fluorescence emission profiles of ICG (in PBS, plasma, and ethanol) and IR-E1050 (in PBS and plasma) at (A) 10, (B) 20, and (C) 50 µM.

Fig 3. Intralipid^® phantom study of ICG in NIR-I and NIR-II window.

(A) Fluorescence images in NIR-II (top panel) and NIR-I window of glass capillary filled with ICG in plasma (50µM) at depths of 1, 2, and 4 mm in 1% Intralipid^®. 785 nm laser used for excitation. Scale bars are 50 mm. (B) Normalized intensity loss of ICG in plasma, in NIR-I and NIR-II as a function of depth. (C) Full-width-half-maximum (FWHM) of capillary glass tube filled with ICG in plasma as a function of depth in Intralipid^®, showing loss of feature consistency in NIR-I compared to the NIR II. For NIR-I camera standard deviation of the signal intensity in the air was 0, which causes an indefinite value of CNR, for the NIR-I and NIR-II comparisons we plotted both Normalized Intensities and FWHM as in [24]. While for tissue phantoms and in-vivo experiments we report CNR values.

Fig 4. NIR-I and NIR-II images of tissue phantoms with inserted tubes containing ICG or IR-E1050 in plasma.

Chicken muscle tissue (top panel) ICG in NIR-I (A) and NIR-II window (B). IR-E1050 in NIR-II (C). Calf liver tissue (bottom panel), ICG in NIR-I (D) and NIR-II window (E). IR-E1050 in NIR-II (F). Top capillary tube is 3 mm and bottom tube is 6 mm from the surface. Scale bar is 5mm.

Fig 5

CNR in capillary tubes embedded at 3 and 6 mm depth from surface in (A) chicken muscle tissue, and (B) calf liver tissue. Tubes were filled with ICG or IR-E1050 in PBS and plasma and imaged in NIR-I (ICG) and NIR-II (ICG and IR-E1050) window.

Fig 6. In vivo sub-surface vascular imaging with ICG and IR-E1050.

Representative images of the hind limb in a mouse in; (A) visible light, (B) NIR-II window obtained after i.v. administration of IR-E1050, (C) NIR-II window obtained after i.v. administration of ICG, and (D) NIR-I window after i.v. administration of ICG. (C) and (D) is the same animal. NIR images at 5 minutes post injection.

Fig 7. CNR in femoral vessel of animals injected with ICG or IR-E1050 in NIR-I (ICG) and NIR-II (ICG and IR-E1050) window (n = 5).