Table 3.
Overview of study results analyzing the use of ICG as a fluorescent agent for the detection of lung cancer.
| Author and Year |
Study Population |
#Patients (Lesions) |
Doses | Timing | Highlights |
|---|---|---|---|---|---|
| Jiang 2015 |
LLC mouse models | 25 (25) | 0.71, 2.0, 5.0, 7.5, 10.0 mg/kg IV | 1 min–72 h | The optimal doses and timing in mice were 5 mg/kg and 24 h. This was confirmed in patients. TBRs in the 5 mg/kg group were 3.1–3.7. |
| Patients with pulmonary nodules |
6 (6) | 0.71, 2.0, 5.0 mg/kg IV | 1 day | ||
| Newton 2018 |
Patients with NSCLC | 18 (18) | 1–3 mg/kg IV (n = 9), 4–5 mg/kg IV | 1 day | In the 4–5 mg/kg group, 8 of 9 tumors were fluorescent, with a mean TBR of 2.70. In the 1–3 mg/kg group, 1 of 9 tumors were fluorescent, with a mean TBR of 1.49. |
| Holt 2014 |
Dogs with primary lung cancer |
8 (8) | 5 mg/kg IV | 1 day | All tumors were fluorescent with a mean in situ SBR of 8.8. NIR imaging was able to detect adequate resection margins in 5 tumors. The other 3 tumors had peritumoral inflammation. |
| Patients with pulmonary nodules |
5 (5) | All nodules were fluorescent, with a mean SBR of 8.1. In 4 nodules, NIR imaging was able to detect adequate resection margins, and none of these nodules showed inflammation. The other tumor was surrounded by atelectasis. | |||
| Okusanya 2014 |
Patients with pulmonary nodules |
18 (18) | 5 mg/kg IV | 1 day | Of 18 lesions, 14 showed fluorescence in situ. There were 5 additional nodules detected. All fluorescent nodules were malignant, and the mean SBR was 2.2. Of non-fluorescent lesions, 3 of 4 were malignant. The sensitivity was 86.4%. |
| Mao 2017 |
Patients with pulmonary nodules |
36 (76) | 5 mg/kg IV | 1 day | Of 76 lesions, 68 were detected with IGS, of which 63 lesions were malignant. All 8 non-fluorescent lesions were malignant. The mean SBR was 3.29. In total, 9 additional lesions were identified. Of these, 5 were false positives. The sensitivity was 89.3%. |
| Kim 2016 |
Patients with pulmonary cancer |
11 (11) | 1 mg/kg IV | 1 day | Of 11 lesions, 10 were fluorescent, of which 8 lesions were malignant. The two false-positive lesions had a pathological complete response after neoadjuvant therapy and new obstructive pneumonia. The non-fluorescent lesion was a false negative. The sensitivity was 88.9%. |
| Quan 2020 |
LLC mouse models | 32 (32) | 1.0 mg/kg inhalation | 10 min–24 h | The tumor margin could be visualized by fluorescence imaging, as confirmed with histological examination. Tumor margins were visible 10 min to 24 h after inhalation of ICG, with a signal peak after 1 h. Inhalation of ICG had a significantly higher tumor margin detection efficiency as compared with intravenous injection of ICG. |
| VX2 rabbit models | 20 (20) | 0.1, 0.25, 0.5, 1.0 mg/kg inhalation | Tumor margins were visible from a dose of 0.25 mg/kg, and higher doses resulted in higher fluorescence intensity in normal tissue. One rabbit showed a false-positive lesion, which turned out to be atelectasis. | ||
| Human lung specimen | 6 (6) | NR inhalation |
NR | All tumor margins were visible, with a mean tumor margin detection efficiency (ratio between signal-to-noise ratio in tumor tissue and that in healthy tissue) of 2.9. |
Abbreviations: LLC—Lewis lung carcinoma; ICG—indocyanine green; IV—intravenous; NSCLC—non-small cell lung carcinoma; NR—not reported; TBR—tumor-to-background ratio; SBR—signal-to-background ratio.