**2.3.3 Liver metastases**

US examination is required to detect liver metastases in patients with oncologic history. In addition, the method can incidentally detect metastases in asymptomatic patients. Early identification (small sizes, small number) is important to establish an optimal course of treatment which can be complex (chemotherapy, radiofrequency ablation, surgical resection) but welcomed. In addition, discrimination of synchronous lesions that have a different nature is also important knowing that up to 25-50% of liver lesions less than 2 cm detected in cancer patients may be benign (Kreft et al, 2001). US sensitivity for metastases detection varies depending on the examiner's experience and the equipment used and ranges between 40-80% (Wernecke et al, 1991). Sensitivity is conditioned by the size and acoustic impedance of the nodules. For a lesion diameter below 10 mm US accuracy is greatly reduced, reaching approx. 20%. Other elements contributing to lower US performance are: excessive obesity, fatty liver disease, hypomobility of the diaphragm, and certain patterns of hyperechoic or isoechoic metastases that can be overlooked or can mimic benign conditions. Conventional US appearance of metastases is uncharacteristic, consisting of circumscribed lesions, with clear, imprecise or "halo" delineation, with homogeneous or heterogeneous echo pattern. They can be single (often liver metastases from colonic neoplasm) or multiple. Echogenity is variable. When increased, they can compress the bile ducts (which may be dilated) and the liver vessels. Liver involvement can be segmental, lobar or generalized. In this situation a pronounced hepatomegaly occurs. Generally, metastases have non-characteristic Doppler vascular pattern, with few exceptions (carcinoid metastases). Cyst-adenocarcinoma metastases due to semifluid content may have a transonic appearance. When increasing, they can result in central necrosis. CEUS examination is a real breakthrough for detection and characterization of liver metastases. Increased performance is based on identifying specific vascular patterns during the arterial phase and seeing metastases in contrast to normal liver parenchyma during the sinusoidal phase. CEUS increased accuracy is due to the different behavior of normal liver parenchyma (captures CA in Kuppfer cells) against tumor parenchyma (does not contain Kuppfer cells, therefore CEUS appearance is hypoechoic). To this adds the particularities of intratumoral circulation represented by a reduced arterial bed compared to that of the surrounding normal liver and the absence of the portal vessels (Cosgrove & Blomley 2004). In terms of vascularity, metastases can be hypovascular (in gastric, colonic, pancreatic or ovarian adenocarcinomas) with hypoechoic pattern during arterial phase, and similar during portal venous and late phases, respectively hypervascular (neuroendocrine tumors, malignant melanoma, sarcomas, renal, breast or thyroid tumors) with hyperechoic appearance during arterial phase, with washout during the portal venous phase and hypoechoic pattern 30 seconds after injection (Larsen, 2010).

Fig. 7. Liver metastases (CEUS). Peripheral vascular pattern of the lesion is observed on the left in colon cancer metastasis. Lesion hyperenhancement in ovarian cancer liver metastasis is seen on the right during the arterial phase.

Using CEUS examination to detect metastases a sensitivity of 80-95% is obtained, similar to that of contrast CT and MRI (Quaia et al, 2006 ; Piscaglia et al, 2007). Intraoperative use of the procedure increases its performance even if it does not have a decisive contribution to change the therapeutic behavior (Konopke et al, 2005). Limitations of the method are those related to US penetration (pronounced fatty liver disease, deep lesion, excessive obesity) and to the experience of the examiner. To this the risk of confusion between hypervascular metastases, hepatocellular carcinoma and hemangioma and the confusion between hypovascular metastases and small liver cysts is added. Routine use of CEUS examination to detect liver metastases is recommended when conventional US examination is not conclusive, when precise information on some injuries (number, location) is necessary in conjunction with contrast CT/MRI and to assess the effectiveness of treatment when using a antiangiogenic therapy for hypervascular metastases (Claudon et al, 2008). The method cannot replace CT/MRI examinations which have well established indications in oncology (Larsen, 2010).
