Lectures

Focal liver lesion: nonlinear contrast-enhanced ultrasound imaging

B-DETECTION OF FOCAL LIVER LESIONS

Conventional US is the most widely diffused cross-sectional imaging procedure for the first approach in diagnosis of abdominal organs and liver diseases due to its relative low cost, its availability, safety, and high patient acceptance. In the detection and staging of a liver lesion conventional US is less accurate than contrast-enhanced spiral CT and MR because of the difficulty in detecting small size and/or iso-echoic lesions, deep lesions especially in difficult anatomical areas (sub-capsular or sub-diaphragmatic areas). The recent advances in nonlinear imaging and the introduction of new contrast agents permit a large number of limitations to be overcome, but the role of CEUS is under debate.
Clinical investigations evaluating CEUS have demonstrated that the accuracy in detecting liver metastases is improved and may be comparable to spiral CT Additionally, others have reported that CEUS can detect lesions not visible on CT (Harvey et al., 2000; Albrecht et al., 2003). Data on diagnostic performance have also been published using the much simpler real-time method with low MI and SonoVue (Albrecht et al., 2003). Even if further studies are required, current data from single center and multicenter clinical trials are becoming more robust. It is not expected that US will replace CT or MR; nevertheless, US remains useful as a complementary procedure in some clinical situations.

Incidental Detection of Hepatic Focal Lesion:

The widespread use of a new generation of imaging modalities has permitted a higher frequency in detecting incidental focal liver lesions in patients without symptoms or evidence of liver disease (Ros and Davis, 1998). Benign hepatic lesions are very frequent. In an autoptic series of 95 patients, benign lesions were detected in 52% of cases; the most common were bile duct tumor in 27% and hemangiomas in 20% (Karhunen, 1986). In ~ 50% of the cases multiple lesions were found. It has also been reported that small hepatic lesions are frequently detected on routine imaging evaluations. In a non-spiral contrast enhanced CT review of 2978 cancer patients, 378 (12,7%) small hepatic lesions measuring 10 mm or less were found, among which 59 (15.6%) were considered metastases; while 303 (80.2%) were classified as benign lesions, 16 (4.2%) were not classified (Schwartz et al., 1999). Current US, CT and MR technological advances improve detection of incidental liver focal lesions. Differentiation between the benign or malignant lesions is difficult in small lesions that do not always show enough characteristic features on static cross-sectional imaging as CT or MR; besides, depending on dimension and position, it may also be difficult to perform a biopsy. In these cases CEUS, being a dynamic real-time imaging modality with good detail resolution, may be used  to play a crucial role (Bleuzen and Tranquart 2004).

Staging and Follow-up of Oncologic Patients:

For optimal management of cancer patients with suspected metastases we need accurate tumor staging. Molecular imaging with PET and cross-sectional imaging with CT and MR prove that they have an important role in total body evaluation. However, the liver is the most common target organ for metastases of many primary tumors, especially for those arising from abdominal organs and structures. Conventional US being readily available and cheap remains in many European centers the first-line imaging modality in detecting hepatic metastases. Ultrasound sensitivity has been reported to be as high as 85% (Clarke et al., 1989); especially for rectal-colon tumor, intra-operative US is the most sensitive imaging modality compared to current multi-phasic contrast CT and RM. Thus, it is usually performed before liver resection to identify synchronous metastases (Conlon et al., 2003). Data related to the use of intra-operative CEUS are still scarce (Siösteen and Elvin, 2004).
In rectal-colon cancer, patients who underwent liver resection range between 30%-40% in terms of five years survival at best because other metastases at the time of resection were undetectable. To detect metachronous metastases, a three-monthly conventional US and yearly CT follow-up program may identify 88% of patients who will develop liver metastases in an asymptomatic stage in order to plan liver resection, chemotherapy or percutaneous treatments (Howell et al., 1999). Additionally, it has been reported that patients with diffuse metastatic diseases receiving systemic chemotherapy at an asymptomatic stage have higher response rates, better quality of life, and survival, than those who begin the chemotherapy at a symptomatic stage (Nordic Gastrointestinal Tumor Adjuvant Therapy Group, 1992). In these follow-up programs CEUS results as the more cost-effectiveness imaging procedure in the detection of liver metastases.
Evaluation in the late parenchymal phase with high M.I., using a first-generation contrast agent (Levovist) has been recommended in some reports (Albrecht et al., 2003). A long liver-specific late phase of up to 20 min has been described due to adherence of the microbubbles to the hepatic sinusoid or to phagocytosis by Kupffer’s cells (Albrecht et al., 2001). The late phase is the most useful time to detect metastases that usually appear as hypo-enhanced or lack enhancement areas. In comparison, most benign lesions show uptake at this time and are therefore not likely to be confused with metastases. Second-generation gas filled microbubbles as in SonoVue are considered a truly blood-pool contrast agent without any interstitial equilibrium phase. The appearance of metastases in the arterial phase is variable. Hypo-vascular metastases show up in CEUS as hypo-enhanced lesions with and without an additional marginal enhancement, while hyper-vascular metastases appear as brightly hyper-enhanced more or less homogeneous lesions. Hyper-vascular metastases occur most often from primary tumors of neuro-endocrine origin or from renal or breast cancer.
A common pitfall is that small cysts are sometimes detected on late phase scanning. These can usually be distinguished from metastases as they characteristically show a higher US beam transmission (Bleuzen and Tranquart, 2004). Differential diagnosis may be difficult for metastases with dominant cystic component, as in ovarian metastases. Contrast-enhanced ultrasound could help to observe enhanced septae or nodules that are very frequent in such metastases. Thrombosed hemangiomas may mimic the hypo enhancing behavior of the metastases. Contrast-enhanced ultrasound follow-up could identify the modification of lesion vascularity due to intralesional thrombus changes over time.

Surveillance for Hepatocellular Carcinoma

Surveillance means repeated application of screening tests. Conventional US and serum alpha-fetoprotein are the most widespread screening tests for hepatocellular carcinoma particularly in European countries; a biannual interval between US examinations is generally accepted. A randomized study demonstrated the benefit of such a program considering that in the surveillance arm the HCC-related mortality was reduced by 37% (Zhang et al., 2004). During the Consensus Conference of European Association for the study of the liver (held in Barcelona in the year 2000), a biannual US and AFL level evaluations were recommended as the minimal follow up. A spiral CT investigation is suggested for patients who do not have nodules on US but an increasing AFP level. For patients who have a US finding of a nodule of 10 mm or less in diameter, three monthly US is suggested taking into account that these lesions are too small to be accurately characterized and at least 50% of these lesions will not be HCC. In patients with a nodule of > 20 mm in diameter HCC can be confirmed if the AFP level is over 400 ng/ml and there is contrast-enhanced CT, MR, or angiographic proof of the hyper-vascularity of the lesion. At least two pieces of positive imaging evidence are required for HCC diagnosis (Bruix et al., 2001).
Biopsy is a controversial option especially for small nodules as it has a potential risk of bleeding and tumor seeding in the needle track, whereas it is not always useful due to the heterogeneity of various types of hepatic nodular lesions (Brú et al., 1989). Microbubble contrast as echo-enhancer permits useful information to be obtained for diagnosis if prevalent hyper-vascularity of the lesion is demonstrated (Lencioni et al., 1996; Francanzani et al., 2001). These imaging modalities are associated with motion and blooming artifacts. With first and second-generation contrast media, CEUS reveals a well defined hyper-enhancement in arterial phase and prevalent washout in the late phase. For real- time evaluation of all three phases, the use of second-generation contrast medium and low M.I., is more suitable (Nicolau et al., 2006).
In a series of 104 patients, using a second-generation of contrast media SonoVue and low M.I., one hundred primary tumors (96.2%) demonstrated hyper-enhancement in the arterial phase but four did not (3,8%) (Nicolau et al., 2004). One study did not find arterial enhancement in 3 well differentiated HCCs out of 60 cases consequently highlighted the relationship between absence of enhancement in the arterial phase and histological differentiation (Tanaka et al., 2001). After diagnosis of HCC, the presence and degree of cirrhosis, as well as the tumor extension, need to be evaluated for staging and management. In fact, the presence of satellite nodules, biliary extension, vascular, and extra-hepatic involvement modify therapeutic planning. Entire hepatic evaluation during the arterial phase (20-30 sec) might be difficult, also because the sweeping scanning might fail to detect some lesions. The administration of several doses of contrast agents and scanning of different hepatic segments are suggested as well as the use of a high M.I. at intervals to evaluate the destruction and parenchyma replenishment in order to detect hyper-enhancing foci. To date, there are no data enabling comparison between CEUS and spiral CT MRI in this challenging field. The extra-hepatic staging favors total body cross-sectional imaging. The evaluation of portal tumor invasion is obviously easier with US examination.

C-MONITORING OF PERCUTANEOUS ABLATIVE TREATMENT

In the last ten years numerous changes have occurred in therapeutic approach for primary and metastatic hepatic malignancies. Many minimally invasive therapies as the percutaneous ethanol injection (PEI) and the thermal ablation using different energy sources (radio frequencies, laser or micro-wave) are proposed (Livraghi et al., 1998; Goldberg et al., 2000; Dodd et al., 2000; De Baere et al., 2000; Lencioni et al., 2003). Percutaneous ablation therapies play a key role in the management of patients with liver malignancies (HCC and metastases). Numerous reports have shown the efficacy and safety of these ablating therapies especially in non surgical candidates. The possibility to repeat treatment sessions in the case of recurrent or new lesions, the lower morbidity and mortality as well as the low costs are notable advantages (Solbiati et al., 2001; Livraghi et al., 2004). Comparative data between surgical and percutaneous treated patients show advantages for the latter (Livraghi in http://www.smirg.org/lectures.php).
As previously described, US, CEUS, contrast-enhanced CT, and MRI play a key role in the diagnosis of liver malignancies in patients that are candidates for any treatment plan. For the safety, low cost, and absence of radiation and iodine contrast administration, CEUS play a crucial role in the early and long-term follow-up of percutaneous treated patients, usually in the first week post treatment and after 1, 3, 6 months and then yearly. No reliable information about the outcome of ablation treatments is provided by conventional US, color Doppler, and power Doppler. In fact, the assessment of vascularization and tissue perfusion is essential to distinguish necrosis from residual viable tumor. Triphasic helical CT or dynamic gadolinium-enhanced MR can predict the extent of the coagulation area to within 2-3 mm.
The disappearance, if any was previously visualized, of intralesional enhancement on contrast-enhanced imaging is the most important finding that suggests complete ablation (Fig.3 a-d). While persistence of hyper-enhancing in the arterial phase or clearly enhancing in the portal phase suggest suspicious residual viable tumor. Conatrast-enhanced ultrasound can provide any kind of vascular related information. Comparing pre- and post treatment findings, the size modification of the lesions permits a good evaluation of the necrotic coagulation area and of treatment results. The efficacy of treatment in the hypo-enhanced lesions, as for most liver metastases, can be identified when the lack of enhanced area is increased determining in this manner the sufficiency of the perilesional "safety" margin achieved.
Contrast-enhanced ultrasound and contrast-enhanced CT, within the first 30 days, post-ablative evaluation may reveal a thin and uniform enhancing rim of hyper-vascularity along the periphery of the necrotic area. Misinterpretation of this perilesional hyperemic halo as residual viable tumor can be avoided by comparing post-ablation with pre-ablation findings (Solbiati et al., 2004).