Contrast-Enhanced Ultrasound of Focal Liver Lesions

Transcription

1 Gastrointestinal Imaging Review D Onofrio et al. EUS of Focal Liver Lesions Gastrointestinal Imaging Review FOUS ON: Mirko D Onofrio 1 Stefano rosara Riccardo De Robertis Stefano anestrini Roberto Pozzi Mucelli D Onofrio M, rosara S, De Robertis R, anestrini S, Pozzi Mucelli R Keywords: contrast-enhanced ultrasound (EUS), focal liver lesions, H, metastasis, microbubbles DOI: /JR Received November 29, 2014; accepted after revision January 10, ll authors: Department of Radiology, G.. Rossi Hospital, University of Verona, Piazzale L.. Scuro 10, Verona, Italy. ddress correspondence to M. D Onofrio (mirko.donofrio@univr.it). WE This is a web exclusive article. JR 2015; 205:W56 W X/15/2051 W56 merican Roentgen Ray Society ontrast-enhanced Ultrasound of Focal Liver Lesions OJETIVE. The purpose of this article is to discuss the use of contrast-enhanced ultrasound (EUS) in focal liver lesions. ONLUSION. Focal liver lesions are usually detected incidentally during abdominal ultrasound. The injection of microbubble ultrasound contrast agents improves the characterization of focal liver lesions that are indeterminate on conventional ultrasound. The use of EUS is recommended in official guidelines and suggested as a second diagnostic step after ultrasound detection of indeterminate focal liver lesions to immediately establish the diagnosis, especially for benign liver lesions, such as hemangiomas, avoiding further and more expensive examinations. ontrast-enhanced ultrasound (EUS) is an imaging method that has been used in European and sian countries for more than 10 years. The use of ultrasound contrast agents has been approved in several countries, but the Food and Drug dministration (FD) in the United States has not yet approved their application for a noncardiac use. The injection of microbubble contrast agents improves the accuracy of ultrasound study. The main liver application of EUS is for focal liver lesions. Focal liver lesions are usually detected incidentally during an abdominal ultrasound examination, during first evaluation or follow-up for a primary neoplasm, or during surveillance in chronic liver diseases and cirrhosis. In cases of incidental findings, focal liver lesions can be characterized by conventional -mode and color Doppler ultrasound when a typical pattern is identified, as in the case of homogeneously hyperechoic hemangiomas [1] or focal nodular hyperplasia with a spoke-wheel shaped central vascular pattern on color Doppler ultrasound [2], but the accuracy of the final definitive diagnosis can be limited. In fact, even though color Doppler imaging during an ultrasound study of the liver can improve diagnostic confidence in the characterization of focal liver lesions, it has important limitations because of limited sensitivity and specificity because benign and malignant lesions may show similar appearance on -mode and Doppler ultrasound [3 9]. Moreover, it is impossible to characterize a focal liver lesion in the presence of underlining oncologic history of the patient or chronic liver disease and cirrhosis because the aspect of a malignant lesion may be similar to a benign lesion. EUS has improved the detection and characterization of focal liver lesions [10 16]. Technique and Protocol efore EUS, a thorough conventional ultrasound examination of the entire liver must be performed. The baseline study includes the assessment of the lesions on -mode imaging and by means of color Doppler ultrasound together with the use of tissue harmonic imaging [17]. The currently used contrast agents (second-generation) are gas-filled (sulfur hexafluoride) microbubbles stabilized by a shell made by albumin, surfactants, or phospholipids. They are designed to be smaller than 7 µm (mean diameter, 2.5 µm) to circulate freely in the capillary beds, showing both macrovasculature and microvasculature, with no interstitial phase because they are exclusively intravascular. In EUS, 2.4 ml (a much lower dose than with T and MRI) of microbubble contrast agent is rapidly injected via an antecubital vein followed by a 5-mL saline flush. sufficiently large needle (20-gauge minimum diameter) should be used to avoid causing bubble rupture. W56 JR:205, July 2015

2 EUS of Focal Liver Lesions ecause microbubbles resist compression better than expansion, when they are insonated at a low mechanical index (lower than 0.3) the expansion and contraction phases are no longer equal and the returning signal shows nonlinear characteristics resulting in the generation of multiple harmonics from bubble-filled vessels. t higher energy levels of insonation, disruption of the bubbles occurs, producing strong but transient harmonic signal. ontrast-specific software integrated in ultrasound scanners cancels the linear ultrasound signal from tissues and uses the nonlinear responses from microbubbles [18]. The use of ultrasound contrast agents has been approved in several countries, but the FD in the United States has not yet approved their application for a noncardiac use. The EUS contrast agent widely used in Europe is a sulfur hexafluoride contrast medium (SonoVue, racco) with a main bubble diameter of 2.5 µm. SonoVue has recently been approved by the U.S. FD for cardiac use. Ultrasound contrast agents are safe, with an extremely low incidence of side effects [19]. There are no cardio-, hepato-, or nephrotoxic effects. Thus, it is not necessary to perform liver or kidney function blood tests before contrast medium administration. The incidence of severe hypersensitivity events is lower than with T contrast agents. Life-treating reactions have been seen, with a rate of 0.001% and no deaths reported in the literature [19]. Microbubbles are eliminated partly by means of metabolization in the liver (stabilizing shells) and partly by the patient breathing out of the lungs (gas filling) in approximately minutes [20], with no risk of nephrotoxicity. Different dynamic phases of contrast enhancement may be identified in the liver study after the injection of a microbubble-based contrast agent. The arterial phase starts within seconds and lasts for seconds after the injection. The portal phase is characterized by the arrival of contrast agent through the portal system; it lasts for 2 minutes after the injection, and the overall liver echogenicity becomes more intense. ecause microbubbles are purely intravascular, unlike the T and MR contrast media that present an interstitial phase, there is no interstitial or equilibrium phase, and for the remaining observation time (late phase) they are progressively cleared [18]. In noncirrhotic patients, the first aim of liver EUS is to distinguish between malignant and benign lesions. The late phase of liver parenchyma contrast enhancement has been shown to be the most useful for this task. enign lesions typically present persistent microbubble uptake with a hyper- or isovascular appearance relative to the adjacent liver, whereas malignant lesions typically display microbubble washout with a hypovascular appearance because their vascularization is almost exclusively arterial and microbubble contrast agents do not present an interstitial phase [18, 21 24]. Video frames of the entire liver have to be recorded in all three contrast phases because it is not possible to simultaneously examine multiple lesions in the arterial and portal phases [25]. If more than one lesion has to be studied, a second bolus of contrast medium should be administered after a washout time of almost half an hour or a flash replenishment technique can be applied, which is an ultrasound technique that uses high-power flash pulses to destroy contrast microbubbles when scanning a liver lesion followed by low-power pulses to show lesion replenishment. However, this technique is not recommended because of the possibility of introducing changes and therefore inhomogenicity of microbubble distribution in the liver and impairing detection. The late phase of EUS is useful for detection of malignant (metastases) focal liver lesions. The late phase of EUS enables better ultrasound hepatic staging in oncologic patients. However in the late phase, simple cysts that are very small can be misinterpreted as metastatic lesions because both appear hypoechoic. This explains the importance of a complete -mode and color Doppler study prior to the administration of the ultrasound contrast agent. In general, if an examination of the liver by ultrasound is insufficient, examination by EUS will be insufficient. The limitations that apply to EUS are the same as those that apply to ultrasound. Thus, the quality of the examination still depends on the skill of the operator. In addition, EUS provides limited ability to observe certain parts of the liver, especially in obese patients. Further limitations are related to the width of the acoustic window and movement artifacts. enign Focal Liver Lesions enign focal liver lesions are a common finding [26]. Noninvasive accurate characterization using the most economic and convenient workup is preferred [27, 28]. Hemangiomas Hemangiomas are the most common benign tumor of the liver [29]. On -mode ultrasound they typically present as homogeneously hyperechoic rounded lesions with distinct margins, sometimes with slight posterior acoustic enhancement [1]. When hemangiomas are found in patients with no history of malignancy, no further imaging is needed. When -mode ultrasound is not sufficient to make a certain diagnosis, EUS is indicated. The typical EUS pattern of hemangiomas is globular enhancement in the peripheral area of the lesion in the arterial phase with progressive centripetal fill-in (Fig. 1). During the late phase the centripetal fill-in appears complete in 40 50% of cases, with a persistent hyper- or isoechogenicity. This pattern is considered diagnostic of hemangioma whatever the -mode appearance of the lesion [30 32]. entripetal filling is most often seen in large lesions but can also be present in lesions smaller than 2 cm. In smaller lesions (16% of cases), the fill-in is more rapid, probably due to the abundant arterioportovenous shunts, and the whole lesion may be enhanced in the arterial phase [1, 24, 33], remaining hyperechoic or isoechoic, but never hypoechoic, in the portal and late phases. Occasionally, lesions that are isoechoic in the arterial and portal phases become progressively more echoic than the surrounding parenchyma in the late phase [11]. Persistent isoechogenicity, instead, strongly suggests the diagnosis of a benign lesion but is not specific for hemangioma. It also may, in fact, be present in focal fatty infiltration. Unenhanced areas in the portal and late phases of hemangiomas represent intralesional thrombosis or fibrosis that may occur in large hemangiomas. The finding of sustained progressive enhancement of the nonthrombosed and nonfibrotic portions in the portal and late phases provides a reliable diagnosis of a benign lesion. typical features also include hemangiomas with calcifications and abundant thrombosis or fibrosis and arteriovenous shunts. Focal Nodular Hyperplasia Focal nodular hyperplasia (FNH) is the second most common benign lesion of the liver [26]; it cannot be defined as a truly neoplastic lesion but rather a regenerative mass of variable size resulting from a vascular abnormality [34, 35]. It may be detected more frequently in women between 30 and 50 JR:205, July 2015 W57

3 D Onofrio et al. years old, and patients are usually asymptomatic. In 15 30% of cases, FNH is multiple [36]. The typical pathologic feature is the presence of a large central (sometimes peripheral) fibrous scar in which an artery larger than usual is located; the artery originates from outside the nodule. This vascular disorder is responsible for the formation of the nodule (composed of normal hepatic structures) and is characterized by the absence of the central terminal hepatic vein and by capillarization of the sinusoids derived from the outside feeding artery [34]. This peculiar vascularization is responsible for the typical dynamic appearance of FNH. There is no specific -mode imaging pattern of FNH, it may be slightly hyperechoic, slightly hypoechoic, isoechoic, or slightly inhomogeneous. The central scar and radiating fibrous septa can usually (70 80% of cases) be identified as linear hyperechoic structures. The central vessel radiating from Fig. 1 avernous hemangioma in 35-year-old man., -mode conventional ultrasound image shows small hypoechoic nodule (calipers) because of fatty background liver visible in left lobe. D, In ultrasound images obtained after contrast medium administration, nodule shows progressive globular centripetal contrast enhancement ( and ) and appears homogeneously hyperechoic (calipers) in late phase (D). the center to the periphery (a highly specific finding) can be identified at baseline color Doppler ultrasound with a high flow and low resistance index pattern [36, 37]. t EUS, FNHs are typically hypervascular, appearing homogeneously hyperechoic in the arterial phase (Fig. 2). The fill-in is characteristically centrifugal and very rapid; the feeding vessel may be clearly seen a few moments before nodule enhancement. The lesion can show hyperechogenicity in the early portal and occasionally in the late phase. Usually, however, FNHs are isoechoic during the portal and late phases [38, 39]. The central scar appears as a hypoechoic area in the late phase [36]. The spoke-wheel pattern is of key importance in distinguishing FNH from high-flow hemangiomas, adenomas, and hypervascular malignant focal liver lesions. The management is conservative because FNHs do not undergo malignant changes. D denoma denoma is a rare benign hepatic neoplasm arising in normal liver, mainly in young women and in people using steroidcontaining medications (e.g., oral contraceptives or anabolizing hormones). The clinical presentation is usually related to the mass effect (40%) or pain due to intralesional bleeding (40%) because of arterial pressure blood perfusion and the lack of a portal venous supply; the ease to bleeding is also due to the architecture, which has large hepatocytes and poor connective tissue support. denoma is asymptomatic in 20% of cases [37]. Usually, adenomas are well encapsulated, 8 10 cm in size, and may be multiple (more than 10). t -mode ultrasound, adenomas can be hyperechoic, hypoechoic, isoechoic, or inhomogeneous. Sometimes, intralesional calcification can be present as the evolution of bleeding episodes. olor Doppler ultrasound shows arterial signals with high peak flow and low impedance [40]. On EUS, adenoma usually shows homogeneous contrast enhancement in the arterial phase, typically with rapid complete centripetal filling. In the early portal venous phase, it usually becomes isoechoic or, more rarely, remains slightly hyperechoic. Intratumoral nonenhancing areas are due to previous bleeding episodes or to necrotic portions. The correct differential diagnosis from FNH is made because of the absence of the central spoke-wheel pattern and because the adenoma shows a centripetal enhancement pattern [36, 41]. The management is very often surgical because of the risk of hemorrhagic complications and possible degeneration. Regenerative and Dysplastic Nodules Regenerative nodules (macroregenerative nodules if larger than 5 mm) represent a region of liver parenchyma enlarged in response to necrosis, abnormal circulation, or other stimuli and can be seen in both noncirrhotic and cirrhotic liver diseases. Dysplastic nodules contain cellular atypia without frank malignant changes [37]. They usually show a hypoechoic, often heterogeneous, appearance on -mode ultrasound with color Doppler ultrasound displaying peripheral arterial and venous vessels [42]. fter microbubble injection, most macroregenerative nodules show absent or persistent dotted contrast enhancement, with a hypo- or isovascular appearance during the arterial phase followed by an isovascular ap- W58 JR:205, July 2015

4 EUS of Focal Liver Lesions pearance in the late phase [24, 43]. Dysplastic macroregenerative nodules may display diffuse contrast enhancement [24] in the arterial phase and are isovascular in the late phase. When diffuse contrast enhancement in the arterial phase is present, the differential diagnosis from well-differenced hepatocellular carcinoma (H) can be difficult [44]. Focal Fatty Deposit or Sparing Focal fatty deposits, or sparing, usually are polygonal areas located along the portal bifurcation or close to the gallbladder and hepatic hilum. They do not produce mass effect and are typically hyper- and hypoechoic, respectively, in comparison with the surrounding liver parenchyma. On EUS, the dynamic behavior is the same as that of the surrounding liver; the arterial and portal vascularization are preserved so the areas (pseudolesions) appear isovascularized in relation to the liver parenchyma in all dynamic phases [36]. Malignant Lesions Malignant lesions may be primary or secondary, the latter much more common than the former in patients not presenting with cirrhosis. The -mode imaging appearance may suggest a malignant nature of the lesions (large number, known primary tumor, hypoechoic halo, and infiltration of intrahepatic vessels). Malignancies are often hypoechoic, but any pattern may be represented. However, in a large number of cases, contrast-enhanced imaging is necessary [36]. Fig. 2 Focal nodular hyperplasia (FNH) in 41-yearold woman., -mode conventional ultrasound image shows small hypoechoic nodule (calipers) on right lobe., olor Doppler ultrasound image with waveform shows artery going into lesion with arterial-type pattern with regular systolic peak and diastolic flow. and D, Ultrasound images obtained after contrast medium administration show nodule to be hypervascular in arterial phase () and slightly hyperechoic in late phase (D). Hepatocellular arcinoma H is the sixth most common neoplasm and the most common primary liver malignancy [45, 46]. In most cases, H develops within an established background of chronic liver disease (70 90% of cases), and most of the patients have a background of liver cirrhosis [47]. Ultrasound is the most common imaging modality for H surveillance in high-risk patients because of its efficiency, availability, noninvasiveness, and low cost [48]. However, Doppler technique applied to -mode ultrasound has low sensitivity in studying blood flow features within a newly discovered lesion. EUS can overcome this problem because it can clearly show blood flows and display tissue perfusion information [49]. H is characterized by arterial phase hypervascularity and later and low washout on EUS (Fig. 3). The arterial phase enhancement is often diffuse or heterogeneous. Peripheral rimlike enhancement is unusual in H and is commonly seen in metastases or intrahepatic cholangiocellular carcinoma (I) [50]. Large Hs often show nonenhancving areas due to necrosis or internal hemorrhage. Washout in H tends to be late and often begins later than 90 seconds after injection of contrast medium, whereas metastases or Is consistently show rapid washout (< 60 seconds) [51]. Most small cholangiocarcinomas, which are infrequently detected during H surveillance, can be differentiated from H because of the presence of rimlike arterial enhancement and rapid washout. Washout is slower or may even not be seen in occasional cases of welldifferentiated H. Washout timing is related to the pathologic differentiation of H: Well-differentiated H tends to show later washout or no washout, whereas poorly differentiated H tends to show rapid washout [52 54]. It is important to understand that most newly discovered hypervascular nodules on EUS detected during H surveillance are H regardless of washout if the nodules do not show the typical appearance of hemangioma [55]. However, a biopsy D JR:205, July 2015 W59

5 D Onofrio et al. is needed to confirm H for hypervascular nodules without washout. EUS is superior to T or MRI for detecting hypervascularity of H because of real-time evaluation of arterial phase enhancement that can be missed by T and MRI because of the predetermined scanning delay [56 58]. There is a small subset of hypovascular Hs, particularly those that are well-differentiated. These lesions usually show transient hypovascularity followed by gradual enhancement, and the lesions become isoechoic relative to normal in the portal venous and late phases [59]. EUS is also a useful tool in the guidance of radiofrequency ablation procedures as well as in the monitoring of procedure efficacy. It can help to localize the lesion before the procedure in cases of nodules not easily detectable on -mode ultrasound. Moreover, marginal disease recurrence adjacent to the ablation zone may not be easily located on gray-scale ultrasound because of the preexisting abnormality. EUS can therefore easily identify the exact location of recurrent H [60]. The role of EUS in the differential diagnosis between H and I is controversial. I always enhances later and more slightly and washes out more quickly than H on EUS [61]. The other major concern in cirrhotic livers is to make a distinction between H and other nodules, such as large regenerative nodules and low- and high-grade dysplastic nodules. Pathologically, large regenerative nodules and low-grade dysplastic nodules generally show arterial and capillary supply similar to that detected in the adjacent cirrhotic nodules, whereas high-grade dysplastic nodules and H may show abnormally increased arterial supply with a decreased portal supply [62]. bout % of high-grade dysplastic nodule cases show arterial hyperenhancement, whereas % show hypoenhancement [63, 64]. Washout is seldom found in the late phase for high-grade dysplastic nodules in contrast to typical H, which is supplied by abnormal arteries alone, best displayed with the use of a flash replenishment technique with maximum intensity processing. Occasionally, cancerous foci of well-differentiated H are encountered within dysplastic nodules, which are called nodule-in-nodule lesions, or dysplastic nod- Fig. 3 Hepatocellular carcinoma in 77-year-old man., -mode conventional ultrasound image shows hypoechoic mass (calipers) in right lobe. and, Ultrasound images obtained after contrast medium administration show nodule to be hypervascular in arterial phase () and slightly hypoechoic in comparison with surrounding liver parenchyma in portal and late phases (). ules with a focus of H. Differentiation between H and these nodules is always a major concern in cirrhotic livers because the appearance on conventional ultrasound may be similar but their prognosis is substantially different from each other. EUS facilitates the detection of the H portion in dysplastic nodules because the H portion generally shows arterial hyperenhancement [61]. Little information is available concerning the role of EUS in the diagnosis of H in noncirrhotic livers. On EUS, the enhancement pattern has no difference in comparison with that in cirrhotic livers, but the differential diagnosis is different, comprising FNH, adenomas, and small hemangiomas [61]. Intrahepatic holangiocarcinoma I is a highly malignant epithelial tumor that originates in the second branch of the intrahepatic bile duct. It is the second most common primary liver tumor. lthough it is a relatively rare neoplasm, its incidence and mortality are increasing because of its late clinical presentation with nonspecific symptoms and lack of an effective nonsurgical therapy [65 68]. combined H-cholangiocarcinoma also exists. Is usually arise on a healthy liver background. lthough rare, Is can grow in a cirrhotic liver (1 2% of newly discovered nodules in a cirrhotic liver are Is). In these cases, the main issue is the differential diagnosis with H nodules [69 71]. In previous studies, EUS had similar diagnostic accuracy to T for I and was suggested as an alternative diagnostic option when T was not available (patients with iodine allergy or impaired renal function) [72]. I is subcategorized into three different types: mass-forming, periductal infiltrating, and intraductal growing. Mass-Forming I Mass-forming I is the most common form [73, 74]; it spreads between hepatocyte plates and expands via the hepatic sinusoidal spaces. It often invades the adjacent peripheral branches of the portal vein. On conventional -mode ultrasound, it usually appears as an ill-defined irregularly hypoechoic mass [75, 76]. On the arterial phase of EUS, four enhancement patterns can be recognized, with descending frequency: peripheral irregular rimlike enhancement, heterogeneous hyperenhancement, homogeneous hyperenhancement, and heterogeneous hypoenhancement [72, 77 79]. In the portal phase, mass-forming I W60 JR:205, July 2015

6 EUS of Focal Liver Lesions usually shows hypoechogenicity (Fig. 4). ll mass-forming Is are invariably hypoechoic on the EUS late phase [75, 76]. Peripheral enhancement during the arterial phase corresponds with the pathologic features of the lesion: marked fibrous stroma in the center and abundant carcinoma cells at the edge [75, 76]. On the other hand, in homogeneously or heterogeneously hyperenhancing mass-forming Is, carcinoma cells were prominent in the center as well as in the periphery. This dynamic behavior can be related to the tumor size: Usually small Is show homogeneous enhancement in the arterial phase because they are abundant in tumor cells with little fibrous tissue; on the other hand, in large Is, more fibrous tissue and central necrosis appear, giving rise to the characteristic enhancing peripheral rim [80, 81]. Periductal Infiltrating I Periductal infiltrating Is tend to spread along the bile duct wall via the nerves and perineural tissue of Glisson capsule toward the porta hepatis. On conventional - mode ultrasound, they usually appear as hypoechoic ill-circumscribed irregular lesions with ambiguous boundaries and adjacent bile duct dilation [75]. On EUS, they appear heterogeneously enhancing in the arterial phase, with hypoenhancement in both the portal and late phases [75]. Intraductal Growing I Intraductal growing Is are usually small or polypoid and do not invade deeply into the submucosal layer, often spreading superficially along the mucosa surface. ecause of this nonaggressive behavior, the prognosis is much better than that of the previously mentioned subtypes [73, 82]. On conventional -mode ultrasound, intraductal Is usually appear as hyperechoic lesions with well-circumscribed boundaries and local bile duct dilation [75]. On EUS, in most cases, homogeneous hyperenhancement in the arterial phase is visible, with hypoenhancement in both the portal and late phases [75]. The areas of typical delayed enhancement in I at T have been reported to correspond to fibrotic stroma at histopathologic examination [73, 82 84]. However, Is do not show delayed enhancement on EUS. The delayed enhancement of I is due to the T contrast material characteristics; it is rapidly cleared from the blood pool and retained within the interstitial space in the fibrous stroma of the tumor during the late phase [83, 84]. Ultrasound contrast agents are exclusively blood pool tracers and are confined to the intravascular space. Metastases Metastases are the most common malignant liver lesions. They mainly arise from cancers in the gastrointestinal tract, pancreas, breast, and lung. EUS has markedly improved the detection rate of liver metastases, with reported sensitivity and specificity ranging from 80% to 95% [85 87]. Hypovascular Metastases Most liver metastases (e.g., from gastrointestinal adenocarcinomas or squamous cell carcinomas) usually are hypovascular or weakly enhanced during the arterial phase (15 30 seconds after contrast injection), with enhancement, when present, more pronounced at the periphery of the lesion [18, 24, 50]. This phase of hypervascularity is often undetected on T and MRI because of its brevity (its washout usually takes place after 20 seconds from the injection of the contrast medium, whereas the arterial phase of MRI and T starts after about 40 seconds from the injection). On the other hand, EUS enables a real-time dynamic lesion study, with continuous target lesion monitoring. fter the washout, all metastases are invariably Fig. 4 Intrahepatic mass-forming cholangiocarcinoma (I) in 70-year-old woman., -mode conventional ultrasound image shows small hypoechoic nodule (calipers) in left liver lobe. and, Ultrasound images obtained after contrast medium administration show nodule to be hypervascular in arterial phase () and markedly hypoechoic in comparison with surrounding liver parenchyma in portal and late phases (). I contrast washout in portal and late phases is more intense than that of hepatocellular carcinoma. JR:205, July 2015 W61

7 D Onofrio et al. hypoechoic on EUS (Fig. 5) in the portal and late phases. Hypervascular Metastases Hypervascular metastases frequently arise from neuroendocrine tumors (NETs); melanomas; sarcomas; and renal, breast, and thyroid neoplasms. Most NET liver metastases show increased arterial enhancement on EUS. EUS behavior of gastroenteropancreatic NET liver metastases shows hypervascularization (Fig. 6) during the arterial phase with portal venous and late phase washout; however, contrast washout is slightly delayed in H in comparison with hypervascular metastases [88, 89]. In the portal ( seconds after contrast injection) and delayed (up to 5 minutes after injection) phases, both hyper- and hypovascular metastases usually appear as dark defects on the liver background because they lack portal supply [85 90]. Several authors have stated that EUS has significantly higher sensitivity in detecting liver metastases than -mode ultrasound and a similar detection rate to T and MRI, with some studies showing higher, although not statistically significant, EUS sensitivity in comparison with T and MRI [13, 85, 90, 91]. ccording to the European Federation of Societies for Ultrasound in Medicine and iology (EFSUM) guidelines [92], EUS is indicated for lesion characterization when a lesion or suspected lesion is detected with ultrasound in patients with a known history of malignancy, as an alternative to T or MRI, or when T and MRI are contraindicated or inconclusive. However, it is important to emphasize that EUS is complementary to T or MRI in preoperative staging before liver resection. It cannot replace the other imaging modalities in the preoperative workup or follow-up of patients with liver metastases during chemotherapy because T and MRI provide more comprehensive information. ccuracy of EUS and Review of the Literature EUS has improved the characterization of focal liver lesions, offering comparable results to those with T and MRI when ultrasound exploration is technically satisfactory [10, 11]. EUS, when performed by experi- Fig. 5 Hypovascular liver metastasis in 42-year-old woman., -mode conventional ultrasound image shows hypoechoic mass in left lobe. and, Ultrasound images obtained after contrast medium administration show nodule to be hypovascular with small peripheral rim of enhancement in arterial phase () and intensely hypoechoic in comparison with surrounding liver parenchyma in portal and late phases (). Fig. 6 Hypervascular liver metastasis in 65-year-old man., -mode conventional ultrasound image shows small hypoechoic nodule (calipers) in right liver lobe. and, Ultrasound images obtained after contrast medium administration show nodule to be hypervascular in arterial phase () and clearly hypoechoic in comparison with surrounding liver parenchyma in portal and late phases (). Larger metastatic lesion is visible on late phase (). W62 JR:205, July 2015

8 EUS of Focal Liver Lesions enced operators, significantly improves overall diagnostic accuracy by more than 30% compared with unenhanced ultrasound [24]. The two most important multicenter studies regarding EUS application for the characterization of focal liver lesions were the German Society of Sonography (DEGUM) multicenter study [93] and a French study [94], both of which showed good value for EUS for focal liver lesion characterization. The accuracy was slightly different in the German versus the French study and also different when accuracy in focal liver lesions on noncirrhotic liver was compared with lesions occurring in cirrhotic liver. The accuracy of focal liver lesion characterization has been proven to be higher for hemangiomas or metastasis and lower for Hs or adenomas. The DEGUM study [93] included 1349 patients with focal liver lesions on ultrasound. total of 1328 focal liver lesions (755 malignant and 573 benign) were assessed. The reference standard diagnosis was made by means of liver biopsy in 75% of cases and by contrast-enhanced T or contrast-enhanced MRI in the other cases. The accuracy of EUS for the diagnosis of focal liver lesions was 90.3%. EUS showed 95.8% sensitivity and 83.1% specificity, with 95.4% positive predictive value and 95.9% negative predictive value for differentiating benign versus malignant lesions. Moreover, EUS correctly diagnosed 82.2% of the hemangiomas, 87.1% of the FNHs, 57.9% of the adenomas, 84.9% of the Hs, and 91.4% of the metastases. Thus, EUS has been shown to be an accurate method for the diagnosis of liver metastases and Hs but less accurate for the diagnosis of adenoma. The aim of another study based on the DEGUM multicenter study was to assess the value of tumor-specific vascularization patterns [28], such as spoke-wheel pattern and arterial hyperenhancement followed by isoenhancement in the late phase in FNH, globular peripheral enhancement and partial or complete fill-in pattern in hemangiomas, or late phase hypoenhancement in metastases. The tumor-specific vascularization pattern could be assessed in the majority of cases, and the diagnostic accuracy of EUS was 83.1% for all benign lesions, 95.8% for all malignant lesions, 91.4% for liver metastases, and 84.9% for hepatocellular carcinomas. The French study [94] assessed the clinical value of EUS using SonoVue for the characterization of focal liver lesions discovered in patients with a cancer history or in those with chronic liver disease. The study included 1034 focal liver lesions undiagnosed on ultrasound alone. The reference standard methods were contrast-enhanced T, contrast-enhanced MRI, or liver biopsy. In this study, EUS had 79.4% sensitivity and 88.1% specificity in differentiating benign versus malignant focal liver lesions. In another German study [95], EUS was the more cost-effective method in all scenarios in which EUS examinations were performed at specialized centers compared to T. With about 40,000 relevant liver lesions per year, systematic implementation of EUS would result in a cost savings of 4m per year [95]. Guidelines for Use of EUS in Liver Studies In the 2005 edition [96] of the merican ssociation for the Study of Liver Diseases (SLD) guidelines for H, EUS, T, and MRI were included as indicated imaging modalities for H diagnosis. In 2010, Vilana et al. [77] from the arcelona linic Liver ancer Group found that the I contrast-enhancement pattern corresponds with findings that were diagnostic of H on approximately half of the patients of their cohort. On the basis of this study, EUS was excluded from the 2011 edition of the SLD guidelines [48] and subsequently from the 2012 European ssociation for the Study of the Liver (ESL) guidelines [97] because of the risk of misdiagnosing H and I. oth of these guidelines accept only contrastenhanced multiphase T and dynamic MRI. nother reason for the removal of EUS from the SLD guidelines is that SonoVue is not yet approved by the FD and thus is not available in the United States. The EFSUM released the first guidelines on the use of EUS for liver applications in 2004 [98], followed by updates in 2008 [18] and 2012 [92]. The EFSUM 2012 guidelines recommended the use of EUS for focal liver lesion characterization in the following most common clinical situations: incidental findings on routine ultrasound, lesion or suspected lesion detected with ultrasound in patients with a known history of malignancy, need for a contrast-enhanced study when T and MRI contrast media are contraindicated, and inconclusive MRI or T. In addition, the EFSUM 2012 guidelines recommended the use of EUS for the characterization of all nodules found on surveillance and routine ultrasound. The sian Pacific ssociation for the Study of the Liver (PSL) and Japanese Society of Hepatology (JSH) guidelines [99, 100], on the other hand, accept EUS as an imaging modality for H diagnosis. However, the PSL guidelines recommend the use of T and MRI for first-line diagnosis and EUS as a second or third line in case of inconclusive T and MRI findings, whereas JSH guidelines accept EUS even as a firstline modality. onclusion EUS is an imaging method used in European and sian countries for more than 10 years. The injection of microbubble contrast agents improves the accuracy of the ultrasound study. Whether a patient is symptomatic or not, in Europe, usually at the beginning of a study for abdominal evaluation, ultrasound is performed because it is inexpensive and available in all hospitals. Focal liver lesions are usually detected incidentally during an abdominal ultrasound study, during the first evaluation or follow-up for a primary neoplasm, or during surveillance in chronic liver diseases and cirrhosis. EUS improved the detection and characterization of focal liver lesions. Thus, the diagnosis of an indeterminate focal liver lesion, such as in the case of the common hypoechoic aspect, can be immediately made. EUS is the most economically appropriate second-line imaging method after inconclusive baseline ultrasound for the diagnosis of benign focal liver lesions. References 1. Vilgrain V, oulos L, Vullierme MP, Denys, Terris, Menu Y. Imaging of atypical hemangiomas of the liver with pathologic correlation. RadioGraphics 2000; 20: Wang LY, Wang JH, Lin ZY, et al. Hepatic focal nodular hyperplasia: findings on color Doppler ultrasound. bdom Imaging 1997; 22: Taylor KJ, Ramos I, Morse SS, Fortune KL, Hammers L, Taylor R. Focal liver masses: differential diagnosis with pulsed Doppler US. Radiology 1987; 164: Hosten N, Puls R, echstein WO, Felix R. Focal liver lesions: Doppler ultrasound. Eur Radiol 1999; 9: Tanaka S, Kitamura T, Fujita M, Nakanishi K, Okuda S. olor Doppler flow imaging of liver tumors. JR 1990; 154: Nino-Murcia M, Ralls PW, Jeffrey R Jr, Johnson M. olor flow Doppler characterization of focal hepatic lesions. JR 1992; 159: JR:205, July 2015 W63

9 D Onofrio et al. 7. Numata K, Tanaka K, Mitsui K, Morimoto M, Inoue S, Yonezawa H. Flow characteristics of hepatic tumors at color Doppler sonography: correlation with arteriographic findings. JR 1993; 160: Lee MG, uh YH, ho KS, hung YH, Lee I, Kang EM. olor Doppler flow imaging of hepatocellular carcinomas: comparison with metastatic tumors and hemangiomas by three-step grading for color hues. lin Imaging 1996; 20: González-ñón M, ervera-deval J, García-Vila JH, et al. haracterization of solid liver lesions with color and pulsed Doppler imaging. bdom Imaging 1999; 24: ertolotto M, Dalla Palma L, Quaia E, Locatelli M. haracterization of unifocal liver lesions with pulse inversion harmonic imaging after Levovist injection: preliminary results. Eur Radiol 2000; 10: rannigan M, urns PN, Wilson SR. lood flow patterns in focal liver lesions at microbubble-enhanced US. RadioGraphics 2004; 24: Quaia E, Stacul F, Gaiani S, et al. omparison of diagnostic performance of unenhanced vs SonoVue-enhanced ultrasonography in focal liver lesions characterization: the experience of three Italian centers. Radiol Med (Torino) 2004; 108: Quaia E, D Onofrio M, Palumbo, Rossi S, runi S, ova M. omparison of contrast-enhanced ultrasonography versus baseline ultrasound and contrast-enhanced computed tomography in metastatic disease of the liver: diagnostic performance and confidence. Eur Radiol 2006; 16: Leen E, eccotti P, Kalogeropoulou, ngerson WJ, Moug SJ, Horgan PG. Prospective multicenter trial evaluating a novel method of characterizing focal liver lesions using contrast-enhanced sonography. JR 2006; 186: Strobel D, Seitz K. lank W, et al. Tumor-specific vascularization pattern of liver metastasis, hepatocellular carcinoma, hemangioma and focal nodular hyperplasia in the differential diagnosis of 1,349 liver lesions in contrast-enhanced ultrasound (EUS). Ultraschall Med 2009; 30: D Onofrio M, Martone E, Faccioli N, Zamboni G, Malago R, Mucelli RP. Focal liver lesions: sinusoidal phase of EUS. bdom Imaging 2006; 31: laudon M, Tranquart F, Evans DH, Lefevre F, orreas M. dvances in ultrasound. Eur Radiol 2002; 12: laudon M, osgrove D. lbrecht T, et al. Guidelines and good clinical practice recommendations for contrast enhanced ultrasound (EUS): update Ultraschall Med 2008; 29: Piscaglia F, olondi L; Italian Society for Ultrasound in Medicine and iology (SIUM) Study Group on Ultrasound ontrast gents. The safety of Sonovue in abdominal applications: retrospective analysis of investigations. Ultrasound Med iol 2006; 32: Morel DR, Schwieger I. Hohn L, et al. Human pharmacokinetics and safety evaluation of SonoVue, a new contrast agent for ultrasound imaging. Invest Radiol 2000; 35: lomley MJ, Sidhu PS, osgrove DO, et al. Do different types of liver lesions differ in their uptake of the microbubble contrast agent SH U 508 in the late liver phase? Early experience. Radiology 2001; 220: lbrecht T, Oldenburg, Hohmann J, et al. Imaging of liver metastases with contrast-specific low- MI real-time ultrasound and SonoVue. Eur Radiol 2003; 13(suppl 3):N79 N ryant TH, lomley MJ, lbrecht T, et al. Improved characterization of liver lesions with liver-phase uptake of liver-specific microbubbles: prospective multicenter study. Radiology 2004; 232: Quaia E, alliada F, ertolotto M, et al. haracterization of focal liver lesions with contrast-specific US modes and a sulfur hexafluoride-filled microbubble contrast agent: diagnostic performance and confidence. Radiology 2004; 232: Seitz K. ontrast-enhanced ultrasound in the diagnosis of hepatocellular carcinoma and liver metastases. Ultraschall Med 2005; 26: Karhunen PJ. enign hepatic tumours and tumour like conditions in men. J lin Pathol 1986; 39: Faccioli N, D Onofrio M, omai, ugini. ontrast-enhanced ultrasonography in the characterization of benign focal liver lesions: activitybased cost analysis. Radiol Med (Torino) 2007; 112: Romanini L, Passamonti M, iani L, et al. Economic assessment of contrast-enhanced ultrasonography for evaluation of focal liver lesions: a multicentre Italian experience. Eur Radiol 2007; 17(suppl 6):F99 F Ros PR. omputed tomography: pathologic correlations in hepatic tumors. In: Ferrucci JT, ed. dvances in hepatobiliary radiology. St Louis, MO: Mosby, 1990: Lee JY, hoi I, Han JK, Kim Y, Shin SH, Moon SG. Improved sonographic imaging of hepatic hemangioma with contrast-enhanced coded harmonic angiography: comparison with MR imaging. Ultrasound Med iol 2002; 28: Nicolau, atala V, ru. haracterization of focal liver lesions with contrast-enhanced ultrasound. Eur Radiol 2003; 13(suppl 3):N70 N Kim JH, Kim TK, Kim S, et al. Enhancement of hepatic hemangiomas with Levovist on coded harmonic angiographic ultrasonography. J Ultrasound Med 2002; 21: Quaia E, ertolotto M, Dalla Palma L. haracterization of liver hemangiomas with pulse inversion harmonic imaging. Eur Radiol 2002; 12: Fukukura Y, Nakashima O, Kusaba, Kage M, Kojiro M. ngioarchitecture and blood circulation in focal nodular hyperplasia of the liver. J Hepatol 1998; 29: Gaiani S, elli N, Piscaglia F, et al. Usefulness of contrast-enhanced perfusional sonography in the assessment of hepatocellular carcinoma hypervascular at spiral computed tomography. J Hepatol 2004; 41: Piscaglia F, Lencioni R, Sagrini E, et al. haracterization of focal liver lesions with contrast-enhanced ultrasound. Ultrasound Med iol 2010; 36: Quaia E, D Onofrio M, artolotta TV, et al. haracterization of focal liver lesions. In: Quaia E, ed. ontrast media in ultrasonography: basic principles and clinical applications. Heidelberg, Germany: Springer Science & usiness Media, 2006: Dietrich F, Schuessler G, Trojan J, Fellbaum, Ignee. Differentiation of focal nodular hyperplasia and hepatocellular adenoma by contrast-enhanced ultrasound. r J Radiol 2005; 78: Venturi, Piscaglia F, Silvagni E, et al. Role of real-time contrast-enhanced ultrasound in the assessment of metastatic portal vein thrombosis. Ultraschall Med 2007; 28: Gaiani S, Volpe L, Piscaglia F, olondi L. Vascularity of liver tumours and recent advances in Doppler ultrasound. J Hepatol 2001; 34: Kim TK, Jang HJ, urns PN, Murphy-Lavallee J, Wilson SR. Focal nodular hyperplasia and hepatic adenoma: differentiation with low-mechanical-index contrast-enhanced sonography. JR 2008; 190: Quaia E, Forgacs, alderan L, ertolotto M, Pozzi Mucelli R. haracterization of focal hepatic lesions in cirrhotic patients by pulse inversion harmonic imaging US contrast specific technique with Levovist. Radiol Med 2002; 104: Quaia E, lomley MJ, Patel S, et al. Initial observations on the effect of irradiation on the liverspecific uptake of Levovist. Eur J Radiol 2002; 41: Longchampt E, Patriarche, Fabre M. ccuracy of cytology vs. microbiopsy for the diagnosis of well-differentiated hepatocellular carcinoma and macroregenerative nodule: definition of standardized criteria from a study of 100 cases. cta ytol 2000; 44: Ferlay J, Shin HR, ray F, Forman D, Mathers, Parkin DM. Estimates of worldwide burden of cancer in 2008: GLOON Int J ancer 2010; 127: Forner, Llovet JM, ruix J. Hepatocellular carcinoma. Lancet 2012; 379: Sherman M. Hepatocellular carcinoma: epidemiology, surveillance, and diagnosis. Semin Liver W64 JR:205, July 2015

10 EUS of Focal Liver Lesions Dis 2010; 30: ruix J, Sherman M; merican ssociation for the Study of Liver Disease. Management of hepatocellular carcinoma: an update. Hepatology 2011; 53: Wong GL, Xu HX, Xie XY. Detection of focal liver lesions in cirrhotic liver using contrast-enhanced ultrasound. World J Radiol 2009; 1: Jang HJ, Kim TK, Wilson SR. Imaging of malignant liver masses: characterization and detection. Ultrasound Q 2006; 22: hayana D, Kim TK, Jang HJ, urns PN, Wilson SR. Hypervascular liver masses on contrast-enhanced ultrasound: the importance of washout. JR 2010; 194: Jang HJ, Kim TK, urns PN, Wilson SR. Enhancement patterns of hepatocellular carcinoma at contrast-enhanced US: comparison with histologic differentiation. Radiology 2007; 244: Fan ZH, hen MH, Dai Y, et al. Evaluation of primary malignancies of the liver using contrast-enhanced sonography: correlation with pathology. JR 2006; 186: Iavarone M, Sangiovanni, Forzenigo LV, et al. Diagnosis of hepatocellular carcinoma in cirrhosis by dynamic contrast imaging: the importance of tumor cell differentiation. Hepatology 2010; 52: Jang HJ, Kim TK, Wilson SR. Small nodules (1 2 cm) in liver cirrhosis: characterization with contrast-enhanced ultrasound. Eur J Radiol 2009; 72: Takahashi M, Maruyama H, Shimada T, et al. haracterization of hepatic lesions (</= 30 mm) with liver-specific contrast agents: a comparison between ultrasound and magnetic resonance imaging. Eur J Radiol 2013; 82: Maruyama H, Takahashi M, Ishibashi H, Yoshikawa M, Yokosuka O. ontrast-enhanced ultrasound for characterisation of hepatic lesions appearing non-hypervascular on T in chronic liver diseases. r J Radiol 2012; 85: Sugimoto K, Moriyasu F, Shiraishi J, et al. ssessment of arterial hypervascularity of hepatocellular carcinoma: comparison of contrast-enhanced US and gadoxetate disodium-enhanced MR imaging. Eur Radiol 2012; 22: Kim TK, Jang HJ. ontrast-enhanced ultrasound in the diagnosis of nodules in liver cirrhosis. World J Gastroenterol 2014; 20: Kim YJ, Lee MW, Park HS. Small hepatocellular carcinomas: ultrasonography guided percutaneous radiofrequency ablation. bdom Imaging 2013; 38: Zheng SG, Xu HX, Liu LN. Management of hepatocellular carcinoma: the role of contrast-enhanced ultrasound. World J Radiol 2014; 6: Kim TK, Lee KH, Khalili K, Jang HJ. Hepatocellular nodules in liver cirrhosis: contrast-enhanced ultrasound. bdom Imaging 2011; 36: olondi L, Gaiani S, elli N, et al. haracterization of small nodules in cirrhosis by assessment of vascularity: the problem of hypovascular hepatocellular carcinoma. Hepatology 2005; 42: Xu HX, Lu MD, Liu LN, et al. Discrimination between neoplastic and non-neoplastic lesions in cirrhotic liver using contrast-enhanced ultrasound. r J Radiol 2012; 85: de Groen P, Gores GJ, LaRusso NF, Gunderson LL, Nagorney DM. iliary tract cancers. N Engl J Med 1999; 341: Khan S, Thomas H, Davidson R, Taylor- Robinson SD. holangiocarcinoma. Lancet 2005; 366: Khan S, Taylor-Robinson SD, Toledano M, eck, Elliott P, Thomas H. hanging international trends in mortality rates for liver, biliary and pancreatic tumours. J Hepatol 2002; 37: Patel T. Increasing incidence and mortality of primary intrahepatic cholangiocarcinoma in the United States. Hepatology 2001; 33: Galassi M, Iavarone M, Rossi S, et al. Patterns of appearance and risk of misdiagnosis of intrahepatic cholangiocarcinoma in cirrhosis at contrast enhanced ultrasound. Liver Int 2013; 33: Li R, Zhang X, Ma KS, et al. Dynamic enhancing vascular pattern of intrahepatic peripheral cholangiocarcinoma on contrast-enhanced ultrasound: the influence of chronic hepatitis and cirrhosis. bdom Imaging 2013; 38: ohle W, lemens PU, Heubach T, Zoller WG. ontrast-enhanced ultrasound (EUS) for differentiating between hepatocellular and cholangiocellular carcinoma. Ultraschall Med 2012; 33:E191 E hen LD, Xu HX, Xie XY, et al. Enhancement patterns of intrahepatic cholangiocarcinoma: comparison between contrast-enhanced ultrasound and contrast-enhanced T. r J Radiol 2008; 81: Lim JH, Park K. Pathology of cholangiocarcinoma. bdom Imaging 2004; 29: hoi I, Lee JM, Han JK. Imaging of intrahepatic and hilar cholangiocarcinoma. bdom Imaging 2004; 29: Xu HX, hen LD, Liu LN, Zhang YF, Guo LH, Liu. ontrast-enhanced ultrasound of intrahepatic cholangiocarcinoma: correlation with pathological examination. r J Radiol 2012; 85: Li, Wang W, Ding H, et al. Value of contrastenhanced sonography in the diagnosis of peripheral intrahepatic cholangiocarcinoma. J lin Ultrasound 2011; 39: Vilana R, Forner, ianchi L, et al. Intrahepatic peripheral cholangiocarcinoma in cirrhosis patients may display a vascular pattern similar to hepatocellular carcinoma on contrast-enhanced ultrasound. Hepatology 2010; 51: Xu HX, Liu GJ, Lu MD, et al. haracterization of focal liver lesions using contrast-enhanced sonography with a low mechanical index mode and a sulfur hexafluoride-filled microbubble contrast agent. J lin Ultrasound 2006; 34: hen LD, Xu HX, Xie XY, et al. Intrahepatic cholangiocarcinoma and hepatocellular carcinoma: differential diagnosis with contrast-enhanced ultrasound. Eur Radiol 2010; 20: Yoshida Y, Imai Y, Murakami T, et al. Intrahepatic cholangiocarcinoma with marked hypervascularity. bdom Imaging 1999; 24: Valls, Guma, Puig I, et al. Intrahepatic peripheral cholangiocarcinoma: T evaluation. bdom Imaging 2000; 25: Lim JH. holangiocarcinoma: morphologic classification according to growth pattern and imaging findings. JR 2003; 181: sayama Y, Yoshimitsu K, Irie H, et al. Delayedphase dynamic T enhancement as a prognostic factor for mass-forming intrahepatic cholangiocarcinoma. Radiology 2006; 238: Lacomis JM, aron RL, Oliver JH 3rd, Nalesnik M, Federle MP. holangiocarcinoma: delayed T contrast enhancement patterns. Radiology 1997; 203: Dietrich F, Kratzer W, Strobe D, et al. ssessment of metastatic liver disease in patients with primary extrahepatic tumors by contrast-enhanced sonography versus T and MRI. World J Gastroenterol 2006; 12: orreas JM, Low G, Needleman L, et al. ontrast enhanced ultrasound in the detection of liver metastases: a prospective multi-centre dose testing study using a perfluorobutane microbubble contrast agent (N100100). Eur Radiol 2011; 21: Hohmann J, Muller, Skrok J, et al. Detection of hepatocellular carcinoma and liver metastases with R14: a multicenter phase II study. Ultrasound Med iol 2012; 38: D Onofrio M, Mansueto G, Falconi M, Procacci. Neuroendocrine pancreatic tumor: value of contrast enhanced ultrasonography. bdom Imaging 2004; 29: Malagò R, D Onofrio M, Zamboni G, et al. ontrast-enhanced sonography of nonfunctioning pancreatic neuroendocrine tumors. JR 2009; 192: Oldenburg, Hohmann J, Foert E, et al. Detection of hepatic metastases with low MI real time contrast enhanced sonography and SonoVue. Ultraschall Med 2005; 26: Zheng YL, Yin XY, Xie XY, et al. Value of contrast-enhanced ultrasonography in assessing the vascularity of liver metastases: comparison with contrast-enhanced computed tomography. J Ultrasound Med 2010; 29: laudon M, Dietrich F, hoi I, et al. Guidelines and good clinical practice recommendations JR:205, July 2015 W65