Real-time identification of liver cancers by using indocyanine green fluorescent imaging

Background: We have often encountered difficulties in identifying small liver cancers during surgery. Fluorescent imaging using indocyanine green (ICG) has the potential to detect liver cancers through the visualization of the disordered biliary excretion of ICG in cancer tissues and noncancerous liver tissues compressed by the tumor.

Methods: ICG had been intravenously injected for a routine liver function test in 37 patients with hepatocellular carcinoma (HCC) and 12 patients with metastasis of colorectal carcinoma (CRC) before liver resection. Surgical specimens were investigated using a near-infrared light camera system. Among the 49 subjects, the 26 patients examined during the latter period of the study (20 with HCC and 6 with metastasis) underwent ICG-fluorescent imaging of the liver surfaces before resection.

Results: ICG-fluorescent imaging identified all of the microscopically confirmed HCCs (n = 63) and CRC metastases (n = 28) in surgical specimens. Among the 63 HCCs, 8 tumors (13%, including 5 early HCCs) were not evident grossly unless observed by ICG-fluorescent imaging. Five false-positive nodules (4 large regenerative nodules and 1 bile duct proliferation) were identified among the fluorescent lesions. Well-differentiated HCCs appeared as uniformly fluorescing lesions with higher lesion-to-liver contrast than that of moderately or poorly differentiated HCCs (162.6 [71.1-218.2] per pixel vs 67.7 [-6.3-211.2] per pixel, P < .001), while CRC metastases were delineated as rim-fluorescing lesions. Fluorescent microscopy confirmed that fluorescence originated in the cytoplasm and pseudoglands of HCC cells and in the noncancerous liver parenchyma surrounding metastases. ICG-fluorescent imaging before resection identified 21 of the 41 HCCs (51%) and all of the 16 metastases that were examined.

Conclusions: ICG-fluorescent imaging enables the highly sensitive identification of small and grossly unidentifiable liver cancers in real time, enhancing the accuracy of liver resection and operative staging.