Identification of Metastatic Lymph Nodes Using Indocyanine Green Fluorescence Imaging

Abstract

Indocyanine green (ICG) has been used to detect several types of tumors; however, its ability to detect metastatic lymph nodes (LNs) remains unclear. Our goal was to determine the feasibility of ICG in detecting metastatic LNs. We established a mouse model and evaluated the potential of ICG. The feasibility of detecting metastatic LNs was also evaluated in patients with lung or esophageal cancer, detected with computed tomography (CT) or positron-emission tomography (PET)/CT, and scheduled to undergo surgical resection. Tumors and metastatic LNs were successfully detected in the mice. In the clinical study, the efficacy of ICG was evaluated in 15 tumors and fifty-four LNs with suspected metastasis or anatomically key regional LNs. All 15 tumors were successfully detected. Among the fifty-four LNs, eleven were pathologically confirmed to have metastasis; all eleven were detected in ICG fluorescence imaging, with five in CT and seven in PET/CT. Furthermore, thirty-four LNs with no signals were pathologically confirmed as nonmetastatic. Intravenous injection of ICG may be a useful tool to detect metastatic LNs and tumors. However, ICG is not a targeting agent, and its relatively low fluorescence makes it difficult to use to detect tumors in vivo. Therefore, further studies are needed to develop contrast agents and devices that produce increased fluorescence signals.

Keywords: fluorescence imaging; indocyanine green; metastatic lymph node; primary cancer.

Figure 1

Identification of tumors and metastatic LNs using ICG fluorescence images in a mouse tumor model. (A) Representative CCD and NIR fluorescence images of normal footpad and popliteal LNs in a normal mouse and footpad tumor model with LN metastasis after intravenous injection of ICG (5 mg/kg). Red arrows indicate cancer, yellow arrows indicate metastatic LNs, and orange arrows indicate normal LNs. (B) Ex vivo fluorescence images of a tumor, metastatic LNs, and normal LNs excised 12 h after ICG injection. (C) Representative H&E staining and CCD, GFP [1], NIR (green), and merged (CCD, GFP, and NIR) microscopic images of normal LNs and metastatic LNs (scale bar: 100 μm). (D) Comparison of the signal-to-background ratios of the tumor, the metastatic LNs, and the normal LNs 12 h after ICG injection (n = 4; statistical analysis was performed using one-way analysis of variance followed by Tukey’s multiple comparison test; tumors vs. normal LNs, **** p < 0.0001; metastatic LNs vs. normal LNs, **** p < 0.0001). LN, lymph node; NIR, near-infrared; GFP, green fluorescent protein; ICG, indocyanine green; H&E, hematoxylin and eosin; CCD, charge-coupled device.

Figure 2

Identification of tumors and metastatic LNs using ICG fluorescence imaging in a patient specimen. (A) Representative ex vivo fluorescence images of tumors, metastatic LNs, and normal LNs after intravenous injection of ICG (2 mg/kg). (B) Signal-to-background ratios of the tumors, metastatic LNs, and normal LNs (n = 5; statistical analysis was performed using one-way analysis of variance followed by Tukey’s multiple comparison test; tumors vs. normal LNs, **** p < 0.0001; metastatic LNs vs. normal LNs, *** p < 0.001). LN, lymph node; ICG, indocyanine green.

Figure 3

Comparison of the detection efficiencies of CT, PET/CT, and ICG imaging for LNs in patients. Representative CT, PET/CT, NIR (green) fluorescence, and pathological images of patients. The patients were classified into two groups (nonmetastatic and metastatic LNs) based on the pathologic results. The yellow arrows indicate metastatic LNs, “O” indicates the presence of metastasis, and “X” indicates the absence of metastasis. CT, computed tomography; PET/CT, positron-emission tomography/computed tomography; NIR, near-infrared; LN, lymph node.