Our histopathological examination revealed that the hepatic micrometastases from pancreatic cancer detected by using fluorescence imaging were mainly located at the portal triad (Fig. 38.2). Most of them seemed to be derived from the tumor thrombi of the portal vein in the intrahepatic portal triads, and their extravenous invasion caused desmoplastic reaction around the foci. Such a reaction then resulted in obstruction and stasis of local biliary flow and bile stasis at the peripheral liver (Fig. 38.3). Thereafter, focal dilation of the biliary ducts was frequently observed in the specimens with hepatic micrometastases.

We observed eight cases of false-positive fluorescence in which benign foci were not distinguishable from micrometastases during intraoperative gross examination. Histological examination of the permanent specimens from those cases revealed different biliary congestive conditions including biliary hamartoma, focal steatohepatitis, and cholangiectasis of unknown origin [17]. Thus, some types of bile stasis were found in all cases with abnormal hepatic fluorescence regardless of the histological confirmation of micrometastases. We therefore hypothesize that the ICG-containing stagnant bile is the reason for fluorescence, and hepatic micrometastases themselves do not fluoresce [10, 17]. As a result, the size and location of the fluorescence are not in accord with those of the actual micrometastases (Fig. 38.4). Moreover, diffuse fluorescence was observed in the entire liver of patients with obstructive jaundice, owing to biliary stasis, which supports our hypothesis. In contrast, metastatic foci showed negative fluorescence in those patients (Fig. 38.5). This finding of false-positive results might be associated with inadequate sample collection.

The most important issue regarding hepatic micrometastases from pancreatic cancer is whether they have the same clinical significance as the overt distant metastases. To date, we have treated the two conditions similarly [10]. In cases of histologically confirmed hepatic micrometastases, even if the disease is considered as loco-regionally resectable stage, systemic chemotherapy is administrated after exploratory surgery or palliative bypass operation. On the other hand, patients without hepatic micrometastases (including those with false-positive findings) undergo radical or palliative surgery according to other staging factors. All patients are periodically followed up with MDCT regardless of curative or palliative surgery.

Hepatic overt metastases were more frequently observed in patients with micrometastases (14/18, 77.8 %) within 6 months after the surgery than in those without (7/84, 8.3 %) (Fig. 38.6). The short-term hepatic relapse often developed in multiple lesions and widespread hepatic area. Therefore, the micrometastases detected during surgery seem to have similar clinical impact to that of overt hepatic metastases. Resection of hepatic micrometastases has little clinical validity even when the number of lesions is limited. In addition, systemic chemotherapy is often unable to control the liver disease; therefore, some liver-targeted therapies (i.e., perfusion chemotherapy, radiation) may be needed for patients with hepatic micrometastases [18, 19]. On the contrary, as the probability of hepatic relapse in patients without hepatic micrometastases is significantly lower, loco-regionally curative therapies may play an important role in those patients.