DISCUSSION

This study demonstrates the high sensitivity of ICG imaging, with which tumors < 1 mm in diameter could be detected, although a considerable number of false-positive results were obtained. Histological examination of several ICG-positive tumor-negative specimens revealed thrombi, hemorrhage, and hemosiderin pigmentation, implying disturbances of the microenvironmental pulmonary bloodstream. These histological findings suggest that ICG is detected not only when it is absorbed by tumor cells, but also when it remains in the pulmonary tissue because the bloodstream is disturbed. This hypothesis is reasonable insofar as ICG is used for fundus angiography and lymphangiography during various operations13,14,16–18. Because ICG may be distributed unevenly when the lesions are very small, it is difficult to distinguish whether the tumors themselves are detected with ICG imaging or if it is the surrounding tissues that are affected by the disturbance of the bloodstream associated with the tumor.
The histological subtype of all true-positive specimens was the embryonal type, and no mesenchymal component was detected. Therefore, the applicability of ICG imaging to lesions composed of a predominantly mesenchymal component, which are occasionally observed after chemotherapy, has not yet been clarified.
In this study, 8 of the 250 ICG-positive lesions were not palpable during surgery, and viable tumor nodules were identified in these specimens on histological examination. These lesions may not be detectable without ICG imaging. All these lesions were small, and ICG imaging might not detect the tumors themselves, but it might usefully detect circulatory disturbances in the surrounding tissues. Consequently, ICG imaging is also useful in detecting these small lesions, which are often <1 mm in diameter.
Twenty specimens were ICG negative but contained palpable nodules, and viable tumors were confirmed on histological examination. Of these 20 specimens, 19 were taken from the same patient. The specimens from this patient were initially ICG positive, but reverted to ICG negative, so the character of the tumors was considered to have changed. These tumors might have become refractory to the uptake of ICG, or might have discharged ICG very rapidly. The remaining specimen might not have taken up ICG, probably in response to a circulatory disturbance. Small nonpalpable lesions that failed to take up ICG for some reason could not be detected during surgery.
The microscopic examination of a considerable number of ICG-positive tumor-negative (false-positive) specimens also revealed nonspecific histological changes, such as fibrosis. To identify a false-positive specimen during surgery is considered difficult. To achieve thorough metastasectomy with shorter surgeries and fewer operations, further optimization of the dose of ICG and the timing of its administration may be essential. To minimize the resected lung volume, a wedge-shape resection with a minimal margin performed with open thoracotomy may be more efficient than linear-stapler thoracoscopic surgery.10
Although the sensitivity of ICG imaging is high, several false-negative (ICG-negative tumor-positive) specimens are confirmed, suggesting that some metastatic tumors cannot be detected with ICG imaging. Recognizing the limitations of metastasectomy with ICG imaging is also necessary. Palpation during operation and preoperative imaging studies may still be mandatory for the detection of metastatic lesions, even in the era of ICG imaging.