**4.1 Imaging studies**

282 Liver Transplantation – Basic Issues

than or less than 50%) and if these were uninodular, multinodular or extensive tumors. Several subsequent studies with large patient populations have demonstrated its prognostic usefulness, including a recent study from Taiwan that compared this system to four other models in 1713 patients with primary liver cancer (Hsu, 2010; Ueno, 2001). A downside to this system is that its tumor morphology classification is too broad for the current practices of aggressive hepatoma screening in high risk populations. Nowadays, more patients are being diagnosed with very small tumors which could limit the use the CLIP score. This model has also been criticized for not discriminating well the cases of advanced stage (CLIP score 4-6), and for classifying most of the patients as early stage (CLIP 0-2), which hampers

In an attempt to also combine liver function and tumor characteristics in order to provide a more precise prognostic appraisal to patients with hepatocellular carcinoma, researchers in Osaka, Japan integrated the Child-Pugh score and the Japanese TNM staging to create the JIS system. In their evaluation of 722 cases with primary liver cancer, they concluded that compared to the CLIP score, their system was better stratifying patients and superior discriminating those cases that were in the early hepatoma phase. Statistically significant differences were seen for the lower JIS scores (Kudo et al., 2003). These findings suggest that this staging system could be more useful in regions where early detection of hepatocellular carcinoma has become more common. Other studies have found encouraging results of this model's ability to predict survival (Chen, 2007; Kudo, 2004; Nanashima, 2005). The main disadvantage of the JIS classification is the inability to discriminate well the cases of

Biomarkers have been combined to the JIS system to determine if this modification enhances its prognostic value. The biomarker-combined JIS includes the assessment of the following three tumor markers for hepatoma: AFP, *lens culinaris* agglutinin-reactive AFP, and desgamma-carboxyprothrombin. This modified system was studied in 1,924 patients with primary liver cancer, and proved to be more effective predicting prognosis and stratifying patients than the conventional JIS model (Kitai, 2008). Although this was a large provocative study, this combined system needs to be tested in populations outside of the Asian region.

The Barcelona Clinic Liver Cancer staging system has one peculiarity that is not present in any of the other available prognosis models: it incorporates treatment recommendations in its staging algorithm. The independent predictors of mortality in the original studies were constitutional syndrome, performance status, vascular invasion, and extrahepatic spread (Llovet, 1999a, Llovet, 1999b). By combining in a simple format the evaluation of liver function, tumor stage, performance status, and cancer-related symptoms, and providing suggestions to the best available therapeutic modalities for any particular stage, the BCLC system has demonstrated better predictive value of prognosis and survival stratification when compared to several other staging systems (Cillo, 2004; Guglielmi, 2008). This classification has been gaining wide acceptance as the main staging model used in multiple countries, and supported by several well-respected liver societies. This is related in part to the fact that the BCLC has been externally validated in Asia, Europe and the United States

its stratification capacity (Dohmen 2004; Marrero et al., 2010).

**3.4 Barcelona Clinic Liver Cancer (BCLC) staging system** 

advanced stage (JIS score >4).

**3.3 Japanese Integrated Staging (JIS) and biomarker-combined JIS** 

The key radiographic feature of hepatocellular carcinoma is the presence of arterial enhancement and venous washout in contrast enhanced imaging studies (Bruix & Sherman, 2010). The clinical staging of this cancer requires radiologic tests that would provide the finest of details about its gross morphology and extent of involvement. This information will be crucial to direct therapeutic decisions and to provide a better estimate of prognosis to the patient.

Liver transplantation is one of the curative treatment options for early stage hepatocellular carcinomas that meet specific criteria. Careful assessment with imaging studies helps select the most suitable candidates for this treatment option. Particular attention is paid to tumor size, the number of lesions, any presence of vascular invasion, as well as evidence of distant metastasis. The most common sites of extrahepatic metastasis are the lungs, abdominal lymph nodes and bones (Katyal, 2000).

Groundbreaking technological advances in computer tomography (CT), magnetic resonance (MR), ultrasonography, and other imaging modalities have been studied to determine their significance in the workup patients with primary liver cancer. Ultrasound has its merit in screening for hepatomas as it is noninvasive, offers no radiation exposure, is readily available in most centers, and is inexpensive compared to other imaging techniques. Unfortunately its sensitivity is not satisfactory in cirrhotic patients, and lesions can be missed (Kim, 2011). Nonetheless, the corner stones and only accepted techniques for staging hepatocellular carcinoma before liver transplantation are contrast-enhanced CT or MR (J.M. Lee, 2011).

#### **4.1.1 Computer tomography (CT)**

Computed tomography has proven to be a great tool for the evaluation of hepatocellular carcinoma in affected patients. It is also commonly the test of choice when looking for extrahepatic metastatic disease. The improvements in the scanners, their resolution as well as the intravenous contrast material used in these studies, has improved the detection of this hepatic malignancy (B.I. Choi, 2010). The use of multidetector CT scans, for example has markedly highlighted the hypervascular characteristics of hepatomas which were not as evident in earlier CT scan technology (K.H. Lee, 2004). The sensitivity of CT scans has increased due to these advances. However, some studies have shown that the improved accuracy of this imaging technique is more apparent in the detection of larger tumors with classic radiographic features, and when the studies are interpreted by more experienced radiologists (Addley, 2011). It seems that in the setting of cirrhosis and when the lesions are smaller than 1 centimeter in diameter, CT has difficulties over and underestimating the diagnosis of hepatocellular carcinoma (Luca, 2010; Ronzoni, 2007). Similar observations have been shown in large and experienced liver transplant centers were the false positive diagnosis rate of primary liver cancer has been reported as high as 8%, and predominantly seen in lesions that were between 0.75 and 1.5 centimeters in diameter (Brancatelli, 2003). Perhaps the continued progress in the field of computed tomography might some day better characterize these small lesions.

#### **4.1.2 Magnetic resonance (MR)**

Magnetic resonance imaging provides another suitable option in the staging of hepatocellular carcinoma during the liver transplantation workup. There is ample evidence that attests the important role of MR in the diagnosis and description of hepatomas. The development of faster MR techniques has contributed to obtaining multiphase intravenous contrast-enhanced images that capture with more detail the hypervascular characteristics of this cancer. One of its most important attributes is its superior ability for the detection of small liver tumors, particularly those less than 2 centimeters in diameter (Burrel, 2003; Colli, 2006; Golfieri,2009, D.H. Lee, 2009,). Given the reported higher sensitivities and accuracy detecting these malignant lesions with the newer MR techniques and contrast agents, many transplant centers have adopted MR as their study of choice when evaluating potential transplant recipients with cirrhosis. One cannot forget that as in the case of CT scans and other imaging modalities, the detection and characterization of hepatocellular carcinoma could also be in part influenced by the experience of the radiologists interpreting these studies. MR use is limited in the setting of patients that have certain types of metallic medical implants or other devices, and on those who are claustrophobic or cannot hold their breath. Recently, gadolinium contrast has also been linked to nephrogenic systemic sclerosis which also limits the use contrast-enhanced MR in patients with significant renal failure, a not uncommon situation in the cirrhotic patient population (Idee, 2009). Nonetheless, the current available evidence has MR as the clear frontrunner in the search for the best imaging modality for diagnosing and characterizing hepatocellular carcinoma.

#### **4.1.3 Contrast enhanced ultrasound (CEUS)**

Contrast-enhanced ultrasonography's (CEUS) capability of detecting vascular liver lesions, particularly hepatomas, has been studied and shown some promise (Forner, 2008; Jang, 2007). This modality has even demonstrated remarkable sensitivities and accuracy of 87 and 93% respectively, for diagnosing liver cancer in lesions less than 2 centimeters in diameter (Jang, 2009). Other studies have looked at its utility characterizing portal vein thrombosis as malignant or benign. This differentiation is critical as it has major implications for those patients being evaluated for liver transplantation. When compared to spiral CT for the detection and characterization of portal vein thrombosis in 50 patients with hepatocellular carcinoma and biopsy-proven portal vein thrombosis, CEUS outperformed CT by detecting 100% of the thrombi and correctly characterizing 49/50 (98%) of them. CT modality detected and correctly characterized 68% of these portal vein thrombi, respectively (Rossi, 2008). CEUS has also demonstrated to be as valuable when compared to biopsy for the assessment of the benign or malignant nature of these thrombi (Sorrentino, 2009). Major drawbacks of this modality are that CEUS is still not available in many countries, and that it is operator dependent.
