Hepatic metastases of sarcomas have similar appearances to other liver metastases on imaging, except for certain specific types of sarcoma metastases. Examples of such specific types are metastases of osteosarcomas and liposarcomas. Osteosarcoma metastases may show ossification or calcification, which is recognized by high attenuation values on computer tomography (CT) scans and uptake of bone tracers on bone scintigraphy [26]. Liposarcoma metastases may show fat content, recognized as low attenuation values on CT scans and hyperintense signal intensity on T1-weighted magnetic resonance imaging (MRI) scans that is suppressed on fat-saturated images [27]. These features, however, do not occur in all cases of osteosarcoma and liposarcoma metastases [26, 28, 29].

On unenhanced CT scan, liver metastases from other sarcomas appear hypodense. On the early phase of contrast-enhanced CT scans, they are either hypovascular or hypervascular. Hypovascular metastases remain hypodense on the late phase of contrast-enhanced CT scans, whereas hypervascular metastases tend to show peripheral washout [30]. CT scans with contrast of GIST liver metastases show typically well-defined heterogeneous masses with a peripheral enhancing border of variable thickness and central low attenuation, the result of necrosis, hemorrhage, or cystic change [25].

On ultrasonography, hepatic metastases may appear hypoechoic, hyperechoic, or target-like.

On MRI scans, metastases appear hypointense on T1-weighted images and moderately hyperintense on T2-weighted images. They show restricted diffusion on diffusion-weighted images. As with dynamic gadolinium-enhanced CT scans, metastases are either hypervascular or hypovascular on the early phase of dynamic gadolinium-enhanced MRI scans. Hypovascular metastases tend to remain hypointense on the delayed images whereas hypervascular metastases tend to show peripheral washout [30, 31]. Metastases appear hypointense on the hepatocyte-uptake phase after injection of the hepatocyte-specific contrast agent on MRI [32]. Metastases tend to show avid uptake of fluorodeoxyglucose (FDG) on positron emission tomography (PET)/CT scans [33].

Approximately 25–40 % of all patients with abdominal and pelvic STS develop metastases to the liver. Because systemic chemotherapy does not prolong survival remarkably, hepatic metastasis is generally associated with a poor prognosis [34]. Liver resection is the curative treatment of choice for hepatic metastases from GIST or LMS. Patients who undergo complete resection for GIST or LMS can have actuarial survival rates of 80 and 50 % after 1 and 3 years, respectively; patients who do not undergo liver resection have a poor overall survival of 4 % [35].

In 1998, Hirota et al. discovered gain-of-function mutations in exon 11 of the c-Kit proto-oncogene [36]. The cell surface transmembrane receptor c-Kit has tyrosine kinase activity and production of the protein c-Kit proto-oncogene is present in the majority of GISTs. The activated c-Kit receptor tyrosine kinase (CD117) stimulates the proliferation and the survival of neoplastic GIST tumor cells [37]. In 2001, DeMatteo et al. analyzed a tyrosine kinase inhibitor (TKI), STI571 (imatinib mesylate), used as an anticancer drug. In GISTs, there is a somatic mutation in the c-Kit proto-oncogene that activates its product, the c-Kit receptor tyrosine kinase. STI571 is a TKI that blocks c-Kit [35] and has demonstrated efficacy, minimal toxicity, and a partial response rate of 69 % [38]. The therapeutic use of STI571 has been extended to include being used in conjunction with hepatectomy for liver metastases arising from GISTs. STI571 is used before hepatectomy to permit resection in patients with unresectable disease and after liver resection to prevent tumor recurrence [35].

Before imatinib mesylate, the median survival reported was 19 months in patients with metastatic GIST and only 12 months in patients who also had local recurrences [39]. In patients with CD117-positive tumors, a 5-year survival of 80 % was reported compared with 33 % in patients with CD117-negative tumors [3]. Although tumors recurred in most patients; overall 1-, 2-, and 3-year survivals have been reported with rates of 96.7 %, 76.8 %, and 67.4 %, respectively [40]. In long-term follow-up, tumor recurrence or progression can occur despite an initial response to imatinib mesylate therapy [41]. For this reason, surgical resection should remain the mainstay of treatment for GIST liver metastasis (Table 10.2).

Liver metachronous metastases that arise 24 months after a curative resection are associated with a better prognosis in multivariate analysis. The time of development of liver metastasis is an important prognostic factor in the surgical treatment and in the selection of patients for treatment of recurrence after hepatectomy [47]. However, even patients with liver metachronous metastases that arise before 24 months after a curative resection have a median survival of 31 months [12].

In the largest series reported in the literature, only 17 % of patients were able to undergo resection because liver metastases from visceral STS tend to be diffuse and bilobar, rendering the disease of many patients unresectable [35]. Without hepatic resection, the reported survival rate for these patients is 14 months [53].

It is currently well confirmed that surgical resection in association with postoperative TKI treatment improves survival for patients affected by hepatic metastases from GIST. However, two aspects are not well clarified: the timing of TKI therapy and primary resection, and the GIST resistance mechanism. Because these two aspects are not well understood, different therapeutic approaches are currently proposed. Haller et al. [54] treat recurrence of GIST with systemic TKIs and emphasize surgery for patients who demonstrate early signs of TKI resistance such as “stagnation of tumor shrinkage” on radiographic imaging. On the other hand, the group of Memorial Sloan Kettering Cancer Center (MSKCC) [55, 56] reserves surgery for recurrent disease within 6 months of initiating TKI therapy to minimize the risk of acquiring secondary mutations responsible for TKI resistance. The strategy of this second approach depends on the fact that the response to imatinib from GISTs is maximized after 6 months; therefore, tumor debulking will delay the development of secondary Kit mutations that lead to TKI resistance [57]. This second approach is also validated by a second study in which patients who were treated with imatinib alone had a poorer survival than patients who received 6 months of preoperative TKI therapy before surgery [58].

Currently the evidence of long-term survival based only on TKI therapy in case of recurrent or metastatic GIST is critical. Therefore, measures to prevent acquired resistance are vitally important, and because tumor cell exposure to the TKI cannot be decreased, resection of the tumor whenever that is feasible is the best strategy to minimize the risk of developing secondary TKI resistance mutations.

In contrast to pulmonary metastases of nonvisceral STS, which are treated with aggressive surgery in many centers, the attitude toward hepatic metastases seems to discourage aggressive surgery. However, some groups have adopted an aggressive surgical approach that is in contrast with the current typical management. Of 45 patients reported in the literature as affected by hepatic metastases arising from nonvisceral sarcomas, 27 patients underwent hepatic resection. Some of the patients had a “liberal” indication; that is, in eight patients, extrahepatic abdominal recurrence and hepatic metastases were diagnosed and were resected simultaneously. The results were encouraging because the 5-year survival was 49 %, which is much better than found in other series, which report 5-year survivals of between 20 and 39 % [35, 45, 47]. These results encourage the use of hepatic resection in this category of metastases arising from nonvisceral sarcomas, when the procedure is technically and functionally feasible.