contrast-enhanced ultrasonography in small liver tumors ( 3cm) · 2017-01-28 · contrast-enhanced...

26 J Med Ultrasound 2008 • Vol 16 • No 1 ©Elsevier & CTSUM. All rights reserved.

R E V I E W

A R T I C L E

Contrast-enhanced Ultrasonography in Small Liver Tumors (< 3 cm)

Jing-Houng Wang, Chi-Sin Changchien*

Ultrasonography is a safe, convenient, low cost and noninvasive diagnostic modality forliver tumors. Power Doppler sonography may demonstrate fine tumor vessels in smalllesions and hypovascular lesions. However, it has limitations including motion artifacts,less sensitivity to slow vascular flow, poor demonstration of deep-seated lesions (> 7 cm indepth), and high sensitivity to tissue motion (heart beat or aortic pulsation). Owing toimprovements in contrast agents and new technologies such as harmonic and pulseinversion imaging, contrast-enhanced ultrasound (CEUS) has improved the detectionrate compared with Doppler ultrasound in studies of liver lesions. The enhanced vascularpatterns have been proved to correlate well with the findings from dynamic computedtomography or magnetic resonance imaging. CEUS provides the ability to detect smallfocal liver lesions and even metastatic liver tumors of less than 1 cm in diameter. Thisreview attempts to determine ways to allow the diagnosis of small hepatocellular carcino-mas (HCCs), especially in cirrhotic patients, using CEUS. Because HCCs are small, thefeeding arteries are fine and the arterial blood flow to the tumor is slow, CEUS used in the diagnosis of nodules of 1–2 cm in cirrhotic patients is not satisfactory. The portal andlate phases in pulse inversion imaging may provide more information to detect smalllesions in the cirrhotic liver and improve the diagnostic sensitivity and specificity.Contrast-enhanced flash echo with subtraction mode is another way of detecting thistype of small tumor. In the arterial phase, some tumors are hard to identify, owing to the isoechoic status of the tumors with respect to the surrounding liver parenchyma.However, these small lesions may be shown by flash echo subtraction imaging.Concurrent delayed phase imaging is useful in the diagnosis of small hypovascular HCCs.In conclusion, CEUS improves the diagnostic accuracy of focal liver lesions, even intumors as small as 1–2 cm. This safe, convenient, low cost and noninvasive diagnosticmodality should be promoted in routine clinical practice.

KEY WORDS — contrast-enhanced ultrasonography, focal liver lesion, hepatocellularcarcinoma, liver, ultrasound contrast agent

■ J Med Ultrasound 2008;16(1):26–40 ■

Received: September 17, 2007 Accepted: January 22, 2008Division of Hepato-Gastroenterology, Department of Internal Medicine, Chang Gung Memorial Hospital–Kaohsiung Medical Center, Chang Gung University College of Medicine, Kaohsiung, Taiwan.

*Address correspondence to: Dr. Chi-Sin Changchien, Division of Hepato-Gastroenterology, Department ofInternal Medicine, Chang Gung Memorial Hospital–Kaohsiung Medical Center, 123, Ta-Pei Road, NiaoSung Hsiang, Kaohsiung, Taiwan. E-mail: [email protected]

Contrast-enhanced US in Small Liver Tumors

27J Med Ultrasound 2008 • Vol 16 • No 1

Introduction

Because real-time ultrasonography (US) demon-strates small liver tumors (< 3 cm in diameter) [1],conventional US is used for the detection of focalliver lesions because of its efficiency, availability, non-invasiveness, and relatively low cost. On conven-tional US, it is sometimes difficult to differentiatebenign lesions from malignant lesions, especiallyin smaller lesions. Anatomically, the blood supplyof a malignant tumor is mainly from the hepaticartery, and Doppler ultrasound can demonstratethe vessels in and around the tumor [2]. ColorDoppler US and power Doppler US have proved tobe useful in demonstrating tumor vascularity[3–6]. However, color Doppler sonography failedto show the blood flow in a hypovascular lesion[2]. Although power Doppler sonography is supe-rior to color Doppler sonography in demonstrat-ing fine tumor vessels in small lesions (≤ 2 cm indiameter) and hypovascular lesions, power Dopplertechniques have limitations such as motion arti-facts, less sensitivity to slow vascular flow, poordemonstration of deep-seated lesions (> 7 cm indepth), and high sensitivity to tissue motion (heartbeat or aortic pulsation) [7]. The detection ofincreased arterial flow in a tumor by Doppler US is dependent on the size, depth, and blood flow of the lesion. However, Doppler US studies inpatients at risk of developing hepatocellular carci-nomas (HCCs) have not yet been proven to affectthe sensitivity of detection of small HCCs. Sen-sitivity can be increased by improving contrast resolution and by new technologies involving har-monic frequencies (tissue harmonic imaging) [8].Contrast-enhanced ultrasound (CEUS) improvedthe detection rate up to 90–97% compared withgrayscale US in studies of liver metastatic lesions[9,10]. In a study by Solbiati et al, CEUS detectedmiliary metastases of 5–10 mm in diameter; thedetection rate was as high as 82%, even betterthan that of helical computed tomography (CT) [8].In an overview, Konopke et al stated that the use ofCEUS in 100 patients increased the correct diagnosisof focal liver lesions from 64% to 87% compared

with B-mode US, especially in the case of smallmetastases and following chemotherapy [11].

The enhancement pattern of CEUS in the vas-cular phase is based on the dynamic contrastbehavior of different focal lesions in the liver. Theenhanced vascular patterns were proved to corre-late well with the findings from dynamic CT ormagnetic resonance imaging (MRI) [12]. Theenhancement patterns of liver HCC were also char-acteristic [11,12]. Furthermore, for the detectionof focal liver lesions, CEUS was shown to improveaccuracy in liver metastasis detection and wascomparable to spiral CT [13]. Some studies havesuggested that CEUS can detect lesions not visibleon CT [5,13–18].

An initial report on the use of ultrasound con-trast agents was published in 1969 [19]. Thedevelopment of effective US contrast agents andof new sonographic techniques such as harmonicand pulse inversion imaging has considerablyimproved the possibilities of CEUS in the assess-ment of liver tumors [20,21]. Small focal liverlesions were correctly detected by CEUS [11,22–25],and even metastatic liver tumors of less than 1 cmin diameter as well as tumors located near the liversurface or situated around the ligamentum tereswere also detected [25]. However, in small HCCs,especially in cirrhotic patients, current imagingtechniques are still not accurate enough to estab-lish a reliable diagnosis in nodular lesions of < 2 cm[14,26] or in well-differentiated small HCCs [27].The accurate and early diagnosis of HCC is essen-tial for treatment planning in affected patients.This review attempts to find ways of using CEUS inthe diagnosis of small HCCs.

Ultrasound Contrast Agents (UCAs)

Microbubbles with a diameter of < 8 µm have beenproved to pass through capillary vessels, and anultrasound pulse with a frequency of 2 MHz and a negative pressure of about 700 kPa has the abil-ity to disrupt the microbubbles and generate echo signals [28]. Thus, contrast agents with

J.H. Wang, C.S. Changchien

28 J Med Ultrasound 2008 • Vol 16 • No 1

transpulmonary stability, which are administeredintravenously into peripheral veins, have becomecommercially available for use in sonographicenhancement studies.

Among the UCAs, which include SonoVue,Levovist, Sonazoid, Optison, Definity and Imagent,three types of UCAs are commonly used in Europe[29,30]. These are Levovist, SonoVue and Optison.1. Levovist (SH U 508A): Levovist bubbles contain

air with galactose/palmitic acid surfactant(introduced by Schering in 1996). The bubblesin Levovist are coated with a thin layer ofpalmitic acid. Flash echoes are not observed at a depth of > 7 cm, because the acoustic pressurefall below the threshold of bubble collapse is dueto tissue attenuation. Assuming the tissue atten-uation is 0.6 dB/MHz/cm, transmission powerwould attenuate at −10.5 dB or around 420 kPaat a depth of 7 cm. These results suggest thatmore power output is needed for Levovist toobtain flash echoes deeper than 7 cm [31]. Mainindications for the use of Levovist include heart,abdomen (including vesicoureteral reflux) andtranscranial studies.

2. SonoVue (second-generation agent): Bubblesof SonoVue contain sulfur hexafluoride with a phospholipid shell (introduced by Bracco,Milan, Italy in 2001). SonoVue is a blood poolperfluoro gas agent, which consists of micro-bubbles of sulfur hexafluoride stabilized by aphospholipid shell [32]. The microbubbles areisotonic to human plasma and stable and resis-tant to pressure. SonoVue improves the displayof focal tumor vascularity and normal parenchy-mal liver vascularity [33]. Main indications forthe use of SonoVue are cardiac, macrovascular,liver and breast lesions.

3. Optison: Optison contains octafluoropropane(perflutren) with an albumin shell (introduced byAmersham in 1998). Sole indication is cardiac.UCAs consist of gas bubbles stabilized by a

shell; Levovist contains air, whereas SonoVue (sul-fur hexafluoride) and Optison (perflutren) containlow solubility gases which improve microbubble stability [29].

Techniques

Harmonic imagingBecause of the difference in acoustic impedancebetween air and liquid, air microbubbles in the liq-uid can reflect ultrasound signals which containsignificant energy, not only the fundamental fre-quency from the original transducer, but also higherorder harmonics. Using these nonlinear propertiesof microbubbles, a harmonic imaging study oncapillary blood flow was reported in 1992 [34]. Inthis method, signals are transmitted at the funda-mental frequency, but are received as higher orderharmonics. That is, an ultrasound scanner is trans-mitting at one frequency and receiving double ortriple times the frequency transmitted. This tech-nique improves the detection rate of the microbub-ble contrast agents [35,36]. Harmonic imagingmakes it possible to visualize capillary blood flowin tissues that cannot be detected by conventionalB-mode or color Doppler imaging [35,36].

Pulse inversion imagingBecause of the limitations of resolution due tocompromises forced by harmonic imaging thatrestrict the bandwidth [37], pulse inversion imag-ing mode was designed to allow low incidencepower and nondestructive, continuous imaging ofmicrobubbles in an organ (such as the liver) toproduce high-quality contrast-enhanced images.The pulse inversion imaging mode is superior tosecond harmonic imaging or conventional Dopplerimaging [37,38] and may present as harmonicangio images [39]. In pulse inversion imaging, twoseparate pulses (normal pulse and inverted pulseat 180° out of phase) are transmitted in rapid suc-cession into the tissue. The inverted pulse is a mir-ror image of the normal pulse. The echoes fromthese two successive pulses are received by thescanner which forms their sum. The resultingechoes from tissue behave in a linear manner can-celing each other, and the sum of these two pulsesis zero. For an echo with nonlinear components,such as that from a bubble, the echoes producedfrom these two pulses will not be simple mirror

images of each other. Because of the asymmetricbehavior of the bubble radius with time, the sumof these two pulses is not zero. That means theecho is present and contains nonlinear harmoniccomponents of the signal (including second har-monics). So, the signal can be detected from a bubble but not from tissues [37,40]. Pulse inver-sion imaging mode has no restriction on band-width; the full frequency range of sound emittedfrom the transducer can be detected and thismode provides an effective result [37,41].

Low mechanical index (MI) techniqueUltrasound pulse with an acoustic power has theability to disrupt the microbubbles of UCAs. LowMI with very low acoustic power (such as < 0.2)avoids disruption of the microbubble. This low MIcontrast technique allows dynamic imaging withsubsequent evaluation of the three different vascu-lar phases using a low solubility gas UCA.

UCA is administered as a bolus injection, fol-lowed by a 5–10 mL saline flush. The needle diam-eter should not be smaller than 20 gauge to avoidloss of bubbles due to mechanical impact duringinjection. Continuous scanning for 60–90 secondsis recommended to continuously assess the arterialand portal venous phases. Under this assessmentuntil the disappearance of the UCA from the tis-sues, the microvasculature has been observed[29]. The setting of low MI is an insonating fre-quency of 3 MHz, acoustic power of −75 to −90 dB, and frame rate of 17–20. The scanningtime of a vascular study is up to 3.5 minutes,including the arterial phase of 0–49 seconds, theportal phase of 50–179 seconds and the late phaseof > 180 seconds.

High MI techniqueHigh MI technique in which microbubbles aredeliberately destroyed is probably more useful forfocal liver lesion detection and can be used forcharacterization of these lesions. High MI tech-nique requires intermittent scanning of the lesionduring all three phases [29]. The destruction ofmicrobubbles by high MI ultrasound has been

studied to allow a greater separation between thetissue and the contrast agent. The high MIdestruction imaging technique has been referredto as agent detection imaging (ADI). With Dopplertechniques, ADI displays microbubble signals as a color overlay on the grayscale tissue image. Instudies with ADI, a bubble destruction imageshowed the normal liver as bright and the metas-tases as black without any signal. Thus, ADI makesCEUS very sensitive [30].

Intermittent harmonic imaging withsubtraction modeUsing the second harmonic imaging technique[31], intermittent harmonic (flash echo) imagingwith subtraction mode can be performed to evalu-ate the dynamic perfusion of small lesions in whichpower Doppler US failed to demonstrate the ves-sels. Flash echo imaging in subtraction mode isobtained in the following way. The scanner trans-mits the ultrasound beam at, for example, 2.1 MHzand receives echoes at 4.2 MHz and is set to generate two bursts of high acoustic power (highMI, 1.0–1.2) in rapid succession. The subtractionimage is automatically obtained by setting theultrasound machine to subtract the second frameimage from the first frame image. The flash echoimage with subtraction mode can be designed tooperate depending on the operator’s demand. Thereal-time, low acoustic power imaging (low MI,0.2) is used for monitoring during the intervalsbetween flash echo imaging.

Vascular Phase Study in Liver

Hepatic artery supply usually starts 10–20 secondspost-injection into a peripheral vein and lasts forapproximately 10–15 seconds. This is followed bythe portal vein phase, which usually lasts 2 min-utes after contrast agent injection. The late phaselasts until clearance of the contrast agent from thehepatic parenchyma, up to approximately 15–20minutes post-injection for Levovist and 4–6 minutesfor SonoVue [29].



1. Arterial phase: The contrast agent reaches theliver first via the hepatic artery and providesinformation on the degree and pattern of vascu-larity. Tumors with abundant blood supply showhypervascularity during this phase.

2. Portal vein phase: The contrast agent has passedthrough the circulation and spreads through theliver via the portal branches. This phase usuallylasts 2 minutes after contrast agent injection.

3. Late (parenchyma) phase: The late or parenchy-mal phase follows the portal phase, in which the agent is slowly distributed throughout theentire liver parenchyma. The origin of the latephase is the subject of ongoing scientific discus-sion, and suggested mechanisms include sinusoidpooling and reticuloendothelial system/Kupffercell uptake [42].The portal and late phases provide information

regarding the wash-out of contrast agent from thelesion compared with normal liver tissue. In thecase of hemangiomas, a progressive filling can beobserved during these phases. The portal and latephase enhancement can provide important infor-mation regarding the character of the lesions.Most malignant lesions are hypo-enhancing, whilethe majority of solid benign lesions are iso- orhyper-enhancing [43–46].

The limitations of CEUS for characterization ofliver lesions are subject to the same limitations asother types of ultrasound, and sensitivity is markedlyreduced in attenuating livers and deep lesions. Asa general rule, if baseline ultrasound is suboptimal,results from CEUS may be disappointing.

Small Liver Lesions of < 3 cm

HCCPathologically, HCCs receive an arterial blood sup-ply. In small HCCs of distinctly nodular type, mostof these tumors show hypervascularity despite thesmall tumor size. While in small HCCs of indistinctlynodular type, many of them show hypovascularity.Meanwhile, the majority of well differentiated HCCsof indistinctly nodular type (early HCC) receive

portal blood supply in addition to the arterial bloodsupply, because they contain portal tracts within thetumor. The number of arterial vessels per squaremillimeter in early-stage HCCs of < 1.5 cm in diam-eter is about two-thirds of that in advanced tumors,and it is only less than one-third in tumors smallerthan 1.0 cm. Early HCCs are not encapsulated, andHCC cells proliferate as if they are replacing theliver cell cords at the boundaries. The sinusoidalblood spaces are incompletely vascularized, andthe sinusoidal spaces of the tumor are continuouswith the sinusoids of the surrounding liver tissue.So, a certain proportion of blood flows into thesinusoids of the surrounding liver tissue [47]. Thecontrast agent will wash out from the tumor to the liver parenchyma rapidly, and the tumor mayappear hypoechoic with respect to the surroundingliver in the late phase.

With CEUS, HCCs are characterized by hyper-vascularity in the arterial phase. Using real-timeevaluation with low MI, early intense enhance-ment is usually identified and the feeding artery isclearly visible in most cases. Tumor vessels usuallyshow a basket-like or an irregular branching pat-tern extending from the periphery to the center ofthe tumor [48]. Arterial enhancement of the tumormay be inhomogeneous, because tumors may havesepta, different cell differentiation and arteriovenousshunting among the neoformed vessels [15,26]. In small HCCs, the arterial phase shows hyper-echoic enhancement, and in the portal and latephase, enhancement also provides important infor-mation regarding the character of the lesions. In thelate phase, the most malignant lesions are hypo-enhancing, while the majority of solid benign lesionsare iso- or hyper-enhancing [43–46]. In a study byNicolau et al on the differentiation of benign frommalignant focal liver lesions using contrast-enhancedimaging, the results showed that evaluation of allthree vascular phases was superior to the evaluationof enhancement in the late phase alone. The sensi-tivity increased from 78.4% to 98%, and the accu-racy from 80.9% to 92.7% [49]. However, thesecharacteristic patterns are sometimes hard to de-monstrate in small HCCs, owing to the small tumor





size, fine feeding arteries, and slow arterial bloodflow to the tumor [50]. Contrast-enhanced flashecho with subtraction mode is another way ofdetecting liver tumors as small as 8 mm in diame-ter. In the arterial phase, some tumors show fainthyperperfusional status, which is isoechoic withrespect to the surrounding liver parenchyma dur-ing arterial enhancement and is difficult to iden-tify. Smaller lesion may be shown using flash echosubtraction imaging (Fig. 1) [51]. In a study of 14small liver tumors (diameter, 0.8–3.0 cm; mean,1.8 ± 0.5 cm) by Wang et al, contrast-enhancedflash echo with subtraction mode was sensitive andeffective in detecting these small liver tumors [51].

In cirrhotic liver, multistage processes can exist,including regenerative nodules, dysplastic nodulesand HCC. Differentiation between these processes

is somewhat difficult. On a histopathologic basis,evolving malignant change in a cirrhotic nodularlesion shows that the arterial flow supply progres-sively increases to the lesion [52,53]. This progres-sive neoangiogenesis provides the clue for clinicaldiagnosis with imaging techniques [54]. Therefore,the European Association for the Study of the Liverpanel of experts determined that the diagnosis ofnodules of > 2 cm must be confirmed by two dif-ferent imaging techniques. However, no referencewas made to the vascularity of nodules of 1–2 cmin diameter. Many studies have used CEUS in thecharacterization of liver lesions. In a study of 41cirrhotic patients with small monofocal lesions(< 3 cm in diameter) by Fracanzani et al [16], it wasreported that contrast-enhanced Doppler US usingLevovist was a noninvasive, sensitive technique in

A B C

D E F

Fig. 1. Intermittent harmonic (flash echo) imaging with subtraction mode in a patient with a small hepatocellular carcinoma of1.0 cm in diameter. (A) Conventional ultrasound demonstrates a small hypoechoic lesion (arrow). (B) In the arterial phase of contrast enhanced ultrasound, the hypervascular lesion cannot be differentiated from the enhanced surrounding normal liverparenchyma. (C) Using subtraction mode, the lesion is clearly shown (arrow). (D, E) In the portal and late phases, the lesion showsas hypoechoic (arrow). (F) By subtraction of the late phase image, the lesion is still hypoechoic (arrow).

differentiating malignant and premalignant focallesions. In this study, the intratumoral arterial bloodflow was detected in 19 of 20 HCCs (95%), and in6/21 nonmalignant nodules (28%) by contrast-enhanced Doppler US. Four of the six false-positivefindings were high-grade dysplastic nodules, andthe remaining two evolved to HCC during follow-up [16]. The results showed that the likelihood ofdetecting arterial flow in a nodular lesion in a cir-rhotic patient with HCC was high. Bolondi et al [23]reported 72 small nodules (1–2cm, n=41; 2.1–3cm,n = 31) in 59 cirrhotic patients, which depended onthe coincident arterial hypervascularity at contrastperfusional sonography using SonoVue and helicalCT to detect small HCCs in cirrhotic nodules. Thedetection rate was only 61% (44/72) in nodules of< 3 cm in cirrhotic patients (44% in nodules 1–2 cm,84% in nodules 2–3 cm). Relying on imaging tech-niques in nodules of 1–2 cm, the missed diagnosisof HCC was up to 38%, and any nodules of > 2 cmshould be regarded as highly suspicious for HCC[23]. Thus, even with arterial hypervascularity shownby CEUS, the diagnosis of nodules of 1–2 cm in cirrhotic patients is not satisfactory.

The detection of small HCCs, especially in thecirrhotic liver, is a problem. Small HCCs in cirrhoticliver may be detected as areas of increasedenhancement in the arterial phase, and the shortduration of the arterial phase does not allow sur-veillance of the whole liver. The portal and latephases may provide more information in the de-tection of small lesions in cirrhotic liver [43–46].Forsberg et al reported that Levovist showed a tis-sue-specific late phase with selective enhancementof liver and spleen parenchyma [55]. The accumu-lation of Levovist in parenchyma is known in retic-ular endothelial (RE) cells [56], and Kono et alsuggested that this accumulation is the uptake ofLevovist by mechanisms including sinusoid poolingand RE/Kupffer cells [42]. Therefore, pulse inversionimaging allows visualization of liver parenchymaon sonography [57]. Pathologically, HCC and me-tastatic malignancies in liver lack RE cells, and noHCC or metastasis have shown uptake of Levovistin late-phase pulse-inversion sonography [17]. For

the discrimination of malignant versus benign liver lesions, late-phase pulse-inversion contrast-enhanced sonography improved diagnostic sensi-tivity from 85% to 100% and specificity from 30%to 63% compared with baseline sonography, andwith lower interobserver variability [17]. In addi-tion, Wang et al reported the highly specific visual-ization of small HCCs (≤ 2 cm) in cirrhotic patientsusing a combination of arterial enhancement (AE)and absence of delayed phase enhancement (ADE)using Levovist [58]. Scanning for the delayedphase was carried out about 5–6 minutes laterafter arterial phase evaluation. In the evaluation of30 small hepatic nodules (diameter, 1–2 cm; mean,1.5 ± 0.3 cm) using both AE and ADE for HCCdiagnosis, the sensitivity, specificity, accuracy, pos-itive predictive value and negative predictive valuewere 55.6%, 91.7%, 70%, 90.9% and 57.9%,respectively. When using either AE or ADE for thediagnosis of HCC, the same parameters were94.4%, 66.7%, 83.3%, 81% and 88.9%, respec-tively. The authors concluded that concurrentdelayed phase imaging is useful in the diagnosis of small hypovascular HCCs [58].

For grading of small HCCs on the basis of thepresence of Kupffer cells in small tumors, Kitamuraet al reported that contrast-enhanced color Dopplersonography appeared to reflect the histopatho-logic features of HCCs in 20 tumors (mean diame-ter, 2.8 ± 1.2 cm) and was useful for differentiatingliver tumors [59]. von Herbay et al [17] suggestedthat CEUS could not grade HCCs. However, bothhighly differentiated and less differentiated HCCswere detected without contrast enhancement inlate-phase images [17], and Nicolau et al reportedthat the echogenicity in the portal and late phasescorrelated with cellular differentiation [27]. In 104HCCs (36 cases; diameter, < 2 cm), 96.2% of HCCs(including 27 well-differentiated HCCs) showedenhancement in the arterial phase. Four (3.8%) ofthe well-differentiated cases showed an isoechoicpattern (p<0.05). Therefore, in the arterial phase, noenhancement may represent a well-differentiatedHCC. In the early portal phase, isoechoic echogeni-city was found in well- and moderately differentiated



HCCs, and hypoechoic echogenicity in poorly dif-ferentiated HCCs. In the late phase, isoechoicechogenicity was found in well-differentiated HCCsand hypoechoic in moderately and poorly differ-entiated HCCs [27]. In a report by Wang et al on18 small HCCs (diameter, ≤ 2 cm), no significantcorrelations between cellular differentiation andCEUS enhancement patterns were observed [58].These results were limited by the small number ofcases. However, the correlation between cellulardifferentiation and CEUS enhancement patternsdepends upon the vascularity of HCC in the arte-rial phase, the speed of blood feeding in the portalphase, and the number of Kupffer cells or sinusoidspaces in the late phase. More research is neededto confirm these findings.

MetastasisHepatic metastatic lesions are not uncommon andare not always multiple. Histologically, most hepaticmetastatic lesions are hypovascular. Characteristicsof metastatic tumors include intratumoral hypovas-cularity and hypervascularity in the tumor periph-ery. Ten to fifteen percent of hepatic metastatictumors are hypervascular. Conventional grayscaleUS can detect metastatic lesions as hypo-, hyper-or mixed echogenicity, but conventional US canmiss isoechoic lesions and lesion size of < 1 cm indiameter [10]. Many studies have confirmed theimprovement in accuracy of CEUS in diagnosingliver metastatic lesions [13,17,18,60–62]. The detec-tion rate of hepatic metastatic lesions was reportedby Bernatik et al in their study of 28 patients to be97% [9]. Oldenburg et al reported the sensitivityto be 90% in 128 metastases [10]. Solbiati et alalso reported that CEUS improved the detection ofmiliary metastases (0.5–1 cm) [8]. Characteristicfeatures of hepatic metastatic lesions can bedemonstrated in three phases of continuous low-MI imaging CEUS. In the arterial phase, hypovas-cular metastases appear as hypo-effective lesions,usually with a typical rim enhancement of varyingsize, whereas hypervascular metastases appear asbright enhancing hyper-effective and homogeneouslesions. Rapid wash-out of arterial enhancement is

found in the late arterial and portal phases [63].Therefore, at the beginning of the portal phase,the arterial enhancement fades and the entire hypo-vascular lesion becomes hypoechoic. In the latephase, both hypovascular and hypervascular metas-tases invariably appear dark compared with theenhanced background of normal liver parenchyma.During this late phase, both portal venous andlate-phase imaging markedly increase the contrastbetween the enhanced normal liver. Thus, non-enhancing metastases improve detection, especiallyof small lesions of < 1 cm in diameter and of lesionswhich are isoechoic at baseline [11]. Benign focalliver lesions (FLL) contain Kupffer cells and/or sinu-soids (except hemangiomas), including dysplasticnodules, which also consist of sinusoidal capillariza-tion [47]. So, in the late parenchymal phase, FLL canbe enhanced in an isoechoic pattern with respectto the surrounding liver. An isoechoic pattern inrelation to the adjacent liver in the last phase isdemonstrated in benign lesions, whereas metastasesappear more hypoechoic by the enhanced normalliver background and are easily distinguished fromnormal liver. A high diagnostic accuracy in the dif-ferentiation between metastases and benign FLL inthe late phase has been reported [17,64,65].

HemangiomasHemangiomas are the most frequent benign tumorfound in the liver. Pathologically, a hemangioma iscomposed of cavernous vessels. In the tumor, theblood flow in the vascular space is extremely slowand blood pooling exists [66]. The velocity of flowin the tumor is so slow that Doppler US can notdetect any signals [67]. Although most heman-giomas show homogeneous hyperechoic patternson conventional US, the sonographic features ofhemangiomas are not specific. Furthermore, theimaging diagnostic accuracy is low for small hem-angiomas even by MRI [68]. CEUS has been shownto increase the detection sensitivity of slow-velocityblood flow in hepatic tumors [50] and provides a progressive filling in hemangiomas in the portaland late phases. Thus, in the arterial phase ofCEUS, characteristic images show the presence of





peripheral globular or rim-like enhancement intypical and atypical hemangiomas [48,69,70]. How-ever, these characteristic patterns are also found inmalignant FLL, such as metastases [48,62]. In theportal and late phases, the enhancement of heman-giomas has a specific pattern [12,41,48,71]. In theportal phase, progressive filling may gradually showcentripetal enhancement first and then completefilling within several seconds in small hemangiomaand within 30–60 minutes in giant hemangioma.Incomplete centripetal enhancement and/or incom-plete tumor fill-in may occur in some tumors. Inthe late phase, owing to the contrast agent, thetime elapsed, the filling velocity and intratumoralthrombosis or fibrosis, the enhanced patterns showcomplete enhancement or non-enhancing centralareas (partial thrombosis or fibrosis). As a result ofthe persistence of contrast in the vascular bed ofhemangioma, the enhancement remains hyper-echoic compared with the surrounding tissue [29].In small hemangiomas (< 2 cm in diameter), thearterial phase shows diffuse enhancement, whichmay occur in hypervascular malignant tumors suchas HCC or metastases. However, in the portal andlate phases, small hemangiomas usually have hy-perechoic or stronger enhancement with respectto the surrounding liver tissue, while malignantlesions become hypoechoic [11,12,29]. CEUS canbe used in the diagnosis of hemangioma, whencentripetal fill-in enhancement is a positive findingin hemangioma. Ding et al reported that the sensi-tivity of CEUS in the diagnosis of hemangioma was96.23% and the specificity was 97.5% [72]. Whentypical contrast-enhanced patterns in the vascularand late phases were regarded as positive findings,Nicolau et al showed that the correct identificationof 22 hemangiomas was 81.8% in the vascularphase and 86.4% in the late phase [12].

Focal nodular hyperplasia (FNH)FNH is the second most common benign liver neo-plasm. FNH is a hyperplastic lesion composed his-tologically of all the components of normal livertissue, and about 45% of cases have a central stel-late fibrous scar [73]. In the arterial phase of CEUS,

FNH rapidly enhances with atypical centrifugalradiating enhancement (spoke-wheel pattern) [46,12,74], then shows whole-lesion homogeneoushyperechoic enhancement. The spoke-wheel pat-tern represents a central feeding artery and cen-trifugal blood supply from the center of the lesionto the periphery.

In the portal phase, enhancement in FNH grad-ually changes to isoechoic compared with the sur-rounding liver parenchyma. In the late phase,because the lesion contains all the components ofnormal liver tissue, the lesion remains isoechoic;some lesions even show as slightly hyperechoicwith respect to the surrounding normal liver[17,41]. Due to the central stellate fibrosis scar,CEUS does not show enhancement in the portaland late phases and represents a key feature fordiagnosing FNH. The characteristic features ofFNH on CEUS include: (1) hypervascularity in thearterial phase with a centrifugal blood supply; (2)isoechogenicity in the late phase; and (3) a centralscar. Those features aid in the differential diagnosisof FNH from other hypervascular tumors, includ-ing hypervascular metastases, HCCs, small heman-giomas and adenomas [12]. Di Stasi et al reportedon the characteristic features, and the sensitivityand specificity of diagnosing FNH were 87.6% and94.5%, respectively [74]. Yen et al [75] reportedthat the sensitivity of the spoke-wheel sign or centralscar for FNH in 35 FNH lesions (1.3–7.0 cm; mean,2.9 ± 1.4 cm) was 97.1%, 40%, 28.6%, and 50%for CEUS, color/power Doppler US, CT scan, MRIand hepatic angiography, respectively. Regardlessof the size of FNH, CEUS demonstrated the spoke-wheel sign or central scar well (Fig. 2) [75].

Hepatocellular adenoma (HA)Among the FLL, HA is relatively rare and mainlyfound in young women with a history of oral con-traceptive use, androgen steroid therapy andglycogen storage disease [76]. Histologically, HA iscomposed of cords of tumor cells, which closelyresemble hepatocytes and contain fat and glycogen.On conventional US, HA showed variable US pat-terns, such as hypo-, iso-, hyper- or mixed-echoic



images, depending on the presence of intratumoralhemorrhage, necrosis or fatty change [77]. In thearterial phase of CEUS, early and homogeneoushyperechoic enhancement is found in most cases.However, no enhancement will be seen if thetumor contains hemorrhage or necrosis. Due tothe subcapsular feeding artery in HA, CEUS using acontinuous low-MI image may demonstrate anearly and hyper-enhanced tumor periphery. Thisphenomenon is useful in distinguishing HA fromother hypervascular tumors [12]. In the portal andlate phases, the enhancement of HA is almost thesame as that of liver parenchyma and can remainslightly hypoechoic in relation to the adjacent livertissue in later stages because of varying numbersand activity of Kupffer cells [61,78]. However,larger studies regarding the value of CEUS in HAhave not been carried out [11].

Focal fatty change and spared area in diffuse steatosis of liverOn conventional US, focal fatty change and focalspared areas are usually demonstrated adjacent to the right main portal vein at area 4, the gall-bladder bed or the falciform ligament. However, a single well-demarcated nodule can be foundanywhere in the liver. Because this type of lesionhas normal liver components, CEUS shows thesame enhancement pattern with respect to the

normal liver in all phases and remains isoechoic inthe post-vascular phase [62,79].

CEUS evaluation of small HCCs afterpercutaneous ablation therapyHCCs of < 2 cm are well differentiated and notassociated with abundant neoangiogenesis [80,81].Although contrast-enhanced color Doppler sonog-raphy was reported to be useful in the detection ofresidual HCC after percutaneous local therapy[82,83], the sensitivity of detection was low becauseof slow blood flow [84] and reduced vascularity inthe residual mass. Flash echo contrast sonographywith subtraction (FECS) mode allows microbubblecontrast agents to flow into the capillaries before themicrobubbles are imaged and destroyed [31]. Usingthis method, Wang et al reported the results of as-sessing perfusion and the therapeutic effects afterpercutaneous ablation in small HCCs (Fig. 3) [51,85,86]. The agreement between FECS and CT, FECSand hepatic angiography, and all three imaging mo-dalities in a study of 35 small tumors (mean, 2.0 ±0.5 cm) were 80%, 85.7% and 77.1%, respectively.The sensitivity, specificity, accuracy, and positive andnegative predictive values of FECS in detecting viabletumors were 53.8% (7/13), 90.9% (20/22), 77.1%(27/35), 77.8% (7/9) and 76.9% (20/26), respec-tively [86]. These results were lower than thosereported by Ding et al [87]. The explanation by

A B

Fig. 2. Contrast-enhanced color Doppler ultrasound in 2-cm focal nodular hyperplasia. (A) Conventional US shows an isoechoicmass of about 2 cm in diameter (arrows). (B) In the arterial phase, contrast-enhanced color Doppler ultrasound shows the spoke-wheel pattern of the central artery (arrow) in the tumor.



Wang et al for the lower sensitivity in detecting viabletumor tissue was the smaller tumor size when com-pared with those of other studies [87–90] and thelonger follow-up interval [87,90]. Although CEUShas its limitations in the depiction of tumor vascular-ity for HCC located more than 7 cm from the ab-dominal wall [31,51,85], CEUS has potential in theevaluation of the therapeutic effects of percutaneousablation for HCC, even in small tumors [31,86].

Conclusion

Owing to the easy application of contrast agentsinto peripheral vessels and improvements in tech-nologic devices and techniques, CEUS can clearly

demonstrate the vascular pattern and parenchy-mal contrast in liver lesions. The effectiveness ofCEUS in diagnosing liver tumors (even smalltumors) is not less than that of CT scan, MRI orangiography. CEUS also improves the diagnosticaccuracy of focal liver lesions, even for those assmall as 1–2 cm. This safe, convenient, low costand non-invasive diagnostic modality should bepromoted in routine clinical practice.

References

1. Sheu JC, Sung JL, Chen DS, et al. Ultrasonography ofsmall hepatic tumors using high-resolution linear-arrayreal-time instruments. Radiology 1984;150:797–802.

A B

C D

Fig. 3. A patient with a small hepatocellular carcinoma. After ablation, the viable tissue is demonstrated by intermittent harmonic(flash echo) imaging with subtraction mode. (A) A hypoechoic lesion of about 1.8 × 1.1 cm in diameter is found in the liver (arrow).(B) After ablation, the hypoechoic area is larger than the original lesion (arrow). (C) Flash echo imaging with subtraction in thearterial phase. A clear enhanced vessel is demonstrated at the edge of the hypoechoic lesion (arrow). (D) In the late phase, faintenhancement at the edge of the lesion is still seen in the viable tissue (arrow).

2. Tanaka S, Kitamura T, Fujita M, et al. Color Dopplerflow imaging of liver tumors. AJR Am J Roentgenol1990;154:509–14.

3. Liang JD, Yang PM, Liang PC, et al. Three-dimensionalpower Doppler ultrasonography for demonstratingassociated arteries of hepatocellular carcinoma. J Formos Med Assoc 2003;102:367–74.

4. Reinhold C, Hammers L, Taylor CR, et al. Characteri-zation of focal hepatic lesions with duplex sonogra-phy: findings in 198 patients. AJR Am J Roentgenol1995;164:1131–5.

5. Choi BI, Kim TK, Han JK, et al. Power versus conven-tional color Doppler sonography: comparison in thedepiction of vasculature in liver tumors. Radiology1996;200:55–8.

6. Gaiani S, Volpe L, Piscaglia F, et al. Vascularity of livertumours and recent advances in Doppler ultrasound.J Hepatol 2001;34:474–82.

7. Hosoki T, Mitomo M, Choi S, et al. Visualization oftumor vessels in hepatocellular carcinoma. PowerDoppler compared with color Doppler and angiogra-phy. Acta Radiologica 1997;38:422–7.

8. Solbiati L, Tonolini M, Cova L, et al. The role of contrast-enhanced ultrasound in the detection of focalliver lesions. Eur Radiol 2001;11(Suppl 3):E15–26.

9. Bernatik T, Strobel D, Hahn EG, et al. Detection ofliver metastases: comparison of contrast-enhancedwide-band harmonic imaging with conventionalultrasonography. J Ultrasound Med 2001;20:509–15.

10. Oldenburg A, Hohmann J, Foert E, et al. Detectionof hepatic metastases with low MI real time contrastenhanced sonography and SonoVue. Ultraschall Med2005;26:277–84.

11. Konopke R, Bunk A, Kersting S. The role of contrast-enhanced ultrasound for focal liver lesion detection:an overview. Ultrasound Med Biol 2007;33:1515–26.

12. Nicolau C, Bru C. Focal liver lesions: evaluation withcontrast-enhanced ultrasonography. Abdom Imaging2004;29:348–59.

13. Albrecht T, Blomley MJ, Burns PN, et al. Improveddetection of hepatic metastases with pulse-inversionUS during the liver-specific phase of SH U 508A:multicenter study. Radiology 2003;227:361–70.

14. Strobel D, Kleinecke C, Hansler J, et al. Contrast-enhanced sonography for the characterisation ofhepatocellular carcinomas: correlation with histolog-ical differentiation. Ultraschall Med 2005;26:270–6.

15. Kim TK, Kim AY, Choi BI. Hepatocellular carcinoma:harmonic ultrasound and contrast agent. AbdomImaging 2002;27:129–38.

16. Fracanzani AL, Burdick L, Borzio M, et al. Contrast-enhanced Doppler ultrasonography in the diagnosisof hepatocellular carcinoma and premalignant le-sions in patients with cirrhosis. Hepatology 2001;34:1109–12.

17. von Herbay A, Vogt C, Haussinger D. Late-phasepulse-inversion sonography using the contrast agentLevovist: differentiation between benign and malig-nant focal lesions of the liver. AJR Am J Roentgenol2002;179:1273–9.

18. Harvey CJ, Blomley MJ, Eckersley RJ, et al. Hepaticmalignancies: improved detection with pulse-inversionUS in late phase of enhancement with SH U 508A:early experience. Radiology 2000;216:903–8.

19. Gramiak R, Shah PM, Kramer DH. Ultrasound car-diography: contrast studies in anatomy and func-tion. Radiology 1969;92:937–47.

20. Basilico R, Blomley MJ, Harvey CJ, et al. Which con-tinuous US scanning mode is optimal for the detec-tion of vascularity in liver lesions when enhancedwith a second generation contrast agent? Eur J Radiol2002;41:184–91.

21. Sodhi KS, Sidhu R, Gulati M, et al. Role of tissue har-monic imaging in focal hepatic lesions: comparisonwith conventional sonography. J Gastroenterol Hepatol2005;20:1488–93.

22. D’Onofrio M, Rozzanigo U, Masinielli BM, et al.Hypoechoic focal liver lesions: characterization withcontrast enhanced ultrasonography. J Clin Ultrasound2005;33:164–72.

23. Bolondi L, Gaiani S, Celli N, et al. Characterizationof small nodules in cirrhosis by assessment of vascu-larity: the problem of hypovascular hepatocellularcarcinoma. Hepatology 2005;42:27–34.

24. Xu HX, Liu GJ, Lu MD, et al. Characterization ofsmall focal liver lesions using real-time contrast-enhanced sonography: diagnostic performance analy-sis in 200 patients. J Ultrasound Med 2006;25:349–61.

25. Konopke R, Kersting S, Bergert H, et al. Contrast-enhanced ultrasonography to detect liver metas-tases: a prospective trial to compare transcutaneousunenhanced and contrast-enhanced ultrasonographyin patients undergoing laparotomy. Int J Colorectal Dis2007;22:201–7.

26. Nicolau C, Vilana R, Bru C. The use of contrast-enhanced ultrasound in the management of the cir-rhotic patient and for detection of HCC. Eur Radiol2004;14(Suppl 8):63–71.

27. Nicolau C, Catala V, Vilana R, et al. Evaluation ofhepatocellular carcinoma using SonoVue, a second



generation ultrasound contrast agent: correlation withcellular differentiation. Eur Radiol 2004;14:1092–9.

28. Holland AK, Apfel RE. An improved theory for theprediction of microcavitation thresholds. IEEE TransUltrason Ferroelectr Freq Control 1989;36:204–8.

29. Albrecht T, Blomley M, Bolondi L, et al; for theEFSUMB Study Group. Guidelines for the use of con-trast agents in ultrasound. Ultraschall Med 2004;25:249–56.

30. Averkiou M, Powers J, Skyba D, et al. Ultrasoundcontrast imaging research. Ultrasound Q 2003;19:27–37.

31. Kamiyama N, Moriyasu F, Mine Y, et al. Analysis offlash echo from contrast agent for designing optimalultrasound diagnostic systems. Ultrasound Med Biol1999;25:411–20.

32. Schneider M, Arditi M, Barrau MB, et al. BR1: a newultrasonographic contrast agent based on sulfurhexafluoride-filled microbubbles. Invest Radiol 1995;30:451–7.

33. Leen E, Angerson WJ, Yarmenitis S, et al. Multi-centreclinical study evaluating the efficacy of SonoVue, a new ultrasound contrast agent in Doppler investi-gation of focal hepatic lesions. Eur J Radiol 2002;42:200–6.

34. Schrope B, Newhouse VL, Uhlendorf V. Simulatedcapillary blood flow measurement using a nonlinearultrasonic contrast agent. Ultrason Imaging 1992;14:134–58.

35. Burns PN. Harmonic imaging with ultrasound con-trast agents. Clin Radiol 1996;51(Suppl 1):50–5.

36. Kono Y, Moriyasu F, Mine Y, et al. Gray-scale secondharmonic imaging of the liver with galactose-basedmicrobubbles. Invest Radiol 1997;32:120–5.

37. Burns PN, Wilson SR, Simpson DH. Pulse inversionimaging of liver blood flow: improved method forcharacterizing focal masses with microbubble con-trast. Invest Radiol 2000;35:58–71.

38. Jang HJ, Lim HK, Lee WJ, et al. Ultrasonographicevaluation of focal hepatic lesions: comparison ofpulse inversion harmonic, tissue harmonic, and con-ventional imaging techniques. J Ultrasound Med 2000;19:293–9.

39. Wen YL, Kudo M, Zheng RQ, et al. Characterizationof hepatic tumors: value of contrast-enhanced codedphase-inversion harmonic angio. AJR Am J Roentgenol2004;182:1019–26.

40. Albrecht T, Hoffman CW, Schettler S, et al. B-modeenhancement at phase-inversion US with air-basedmicrobubble contrast agent: initial experience inhumans. Radiology 2000;216:273–8.

41. Dill-Macky MJ, Burns PN, Khalili K, et al. Focalhepatic masses: enhancement patterns with SH U508A and pulse-inversion US. Radiology 2002;222:95–102.

42. Kono Y, Steinbach GE, Peterson T, et al. Mechanismof parenchymal enhancement of the liver with amicrobubble-based US contrast medium: an intra-vital microscopy study in rats. Radiology 2002;224:253–7.

43. Dietrich C, Ignee A, Trojan J, et al. Improved charac-terisation of histologically proven liver tumours bycontrast enhanced ultrasonography during the por-tal venous and specific late phase of SH U 508A. Gut 2004;53:401–5.

44. Bryant T, Blomley M, Albrecht T. Improved charac-terization of liver lesions with liver-phase uptake ofliver-specific microbubbles: prospective multicenterstudy. Radiology 2004;232:799–809.

45. Quaia E, Degobbis F, Tona G, et al. Differential pat-terns of contrast enhancement in different focal liverlesions after injection of the microbubble US con-trast agent SonoVue. Radiol Med (Torino) 2002;107:155–65.

46. Leen E. The role of contrast-enhanced ultrasound inthe characterisation of focal liver lesions. Eur Radiol2001;11(Suppl 3):E27–34.

47. Kojiro M, Roskams T. Early hepatocellular carci-noma and dysplastic nodules. Semin Liver Dis2005;25:133–42.

48. Furuse J, Nagase M, Ishii H, et al. Contrast enhance-ment patterns of hepatic tumours during the vas-cular phase using coded harmonic imaging andLevovist to differentiate hepatocellular carcinomafrom other focal lesions. Br J Radiol 2003;76:385–92.

49. Nicolau C, Vilana R, Catala V, et al. Importance ofevaluating all vascular phases on contrast-enhancedsonography in the differentiation of benign frommalignant focal liver lesions. AJR Am J Roentgenol2006;186:158–67.

50. Hosten N, Puls R, Lemke AJ, et al. Contrast-enhancedpower Doppler sonography: improved detection ofcharacteristic flow patterns in focal liver lesions. J Clin Ultrasound 1999;27:107–15.

51. Wang JH, Lu SN, Changchien CS, et al. Flash-echogray-scale imaging in the subtraction mode forassessing perfusion of small hepatocellular carci-noma. J Clin Ultrasound 2003;31:451–6.

52. Park Y, Yang C, Fernandez G, et al. Neoangiogenesisand sinusoidal “capillarization” in dysplastic nodulesof the liver. Am J Surg Pathol 1998;22:656–62.





53. Roncalli M, Roz E, Coggi G, et al. The vascular pro-file of regenerative and dysplastic nodules of the cirrhotic liver: implications for diagnosis and classifi-cation. Hepatology 1999;30:1174–8.

54. Gaiani S, Celli N, Piscaglia F, et al. Usefulness of contrast-enhanced perfusional sonography in theassessment of hepatocellular carcinoma hypervascu-lar at spiral computed tomography. J Hepatol 2004;41:421–6.

55. Forsberg F, Goldberg BB, Liu JB, et al. Tissue-specific US contrast agent for evaluation of he-patic and splenic parenchyma. Radiology 1999;210:125–32.

56. Schlief R. Galactose-based echo-enhancing agents.In: Goldberg BB, ed. Ultrasound Contrast Agents.London: Dunitz, 1997:75–82.

57. Harvey CJ, Blomley MJ, Eckersley RJ, et al. Pulse-inversion mode imaging of liver specific microbub-bles: improved detection of subcentimetre metastases.Lancet 2000;355:807–8.

58. Wang JH, Lu SN, Hung CH, et al. Small hepatic nodules (≤ 2 cm) in cirrhosis patients: characteriza-tion with contrast-enhanced ultrasonography. LiverInt 2006;26:928–34.

59. Kitamura H, Kawasaki S, Nakajima K, et al.Correlation between microbubble contrast-enhancedcolor Doppler sonography and immunostaining forKupffer cells in assessing the histopathologic gradeof hepatocellular carcinoma: preliminary results. J Clin Ultrasound 2002;30:465–71.

60. Albrecht T, Hoffmann CW, Schmitz SA, et al. Phase-inversion sonography during the liver-specific latephase of contrast enhancement: improved detectionof liver metastases. AJR Am J Roentgenol 2001;176:1191–8.

61. Blomley MJK, Sidhu PS, Cosgrove DO, et al. Do dif-ferent types of liver lesions differ in their uptake ofthe microbubble contrast agent SH U 508A in thelate liver phase? Early experience. Radiology 2001;220:661–7.

62. Tanaka S, Ioka T, Oshikawa O, et al. Dynamic sonog-raphy of hepatic tumors. AJR Am J Roentgenol 2001;177:799–805.

63. Hohmann J, Albrecht T, Hoffmann CW, et al. Ultra-sonographic detection of focal liver lesions: increasedsensitivity and specificity with microbubble contrastagents. Eur J Radiol 2003;46:147–59.

64. Quaia E, Bertolotto M, Forgacs B, et al. Detection of liver metastases by pulse inversion harmonic imaging during Levovist late phase: comparison with

conventional ultrasound and helical CT in 160patients. Eur Radiol 2003;13:475–83.

65. Yucel C, Ozdemir H, Gurel S, et al. Detection and dif-ferential diagnosis of hepatic masses using pulseinversion harmonic imaging during the liver-specificlate phase of contrast enhancement with Levovist. J Clin Ultrasound 2002;30:203–12.

66. Wright TL, Venook AP, Millward-Sadler GH. Hepatictumors. In: Millward-Sadler GH, Wright R, ArthurMJP, eds. Wright’s Liver and Biliary Disease, Volume 2,3rd edition. Philadelphia: WB Saunders, 1992:1079–121.

67. Bleuzen A, Tranquart F. Incidental liver lesions: diag-nostic value of cadence contrast pulse sequencing(CPS) and SonoVue. Eur Radiol 2004;14(Suppl 8):53–62.

68. Sasaki K, Ito K, Koike S, et al. Differentiationbetween hepatic cyst and hemangioma: additivevalue of breath-hold, multisection fluid-attenuatedinversion-recovery magnetic resonance imaging usinghalf-Fourier acquisition single-shot turbo-spin-echosequence. J Magn Reson Imaging 2005;21:29–36.

69. Kim TK, Choi BI, Han JK, et al. Hepatic tumors: con-trast agent-enhancement patterns with pulse-inversionharmonic US. Radiology 2000;216:411–7.

70. Lee JY, Choi BI, Han JK, et al. Improved sonographicimaging of hepatic hemangioma with contrast-enhanced coded harmonic angiography: comparisonwith MR imaging. Ultrasound Med Biol 2002;28:287–95.

71. Quaia E, Bertolotto M, Dalla Palma L. Characteriza-tion of liver hemangiomas with pulse inversion harmonic imaging. Eur Radiol 2002;12:537–44.

72. Ding H, Wang WP, Huang BJ, et al. Imaging of focalliver lesions: low-mechanical-index real-time ultra-sonography with SonoVue. J Ultrasound Med 2005;24:285–97.

73. Wanless IR, Mawdsley C, Adams R. On the patho-genesis of focal nodular hyperplasia of the liver.Hepatology 1985;5:1194–200.

74. Di Stasi M, Caturelli E, De Sio I, et al. Natural history of focal nodular hyperplasia of the liver: an ultrasound study. J Clin Ultrasound 1996;24:345–50.

75. Yen YH, Wang JH, Lu SN, et al. Contrast-enhancedultrasonographic spoke-wheel sign in hepatic focalnodular hyperplasia. Eur J Radiol 2006;60:439–44.

76. Ros PR, Menu Y, Vilgrain V, et al. Liver neoplasms andtumor-like conditions. Eur Radiol 2001;11(Suppl 2):S145–65.



77. Hung CH, Changchien CS, Lu SN, et al. Sonographicfeatures of hepatic adenomas with pathologic corre-lation. Abdom Imaging 2001;26:500–6.

78. Lim AK, Patel N, Gedroyc WM, et al. Hepatocellularadenoma: diagnostic difficulties and novel imagingtechniques. Br J Radiol 2002;75:695–9.

79. Solbiati L, Kirn V, Cova L, et al. Other benign lesionsand pseudolesions. In: Solbiati L, ed. Contrast-enhancedUltrasound of Liver Diseases. Rome: Springer-Verlag,2003:60–1.

80. Kojiro M, Yano H, Nakashima O. Pathology of earlyhepatocellular carcinoma: progression from early toadvanced. Semin Surg Oncol 1996;12:197–203.

81. Nakashima Y, Nakashima O, Hsia CC, et al.Vascularization of small hepatocellular carcinomas:correlation with differentiation. Liver 1999;19:12–8.

82. Bartolozzi C, Lencioni R, Ricci P, et al. Hepatocellularcarcinoma treatment with percutaneous ethanolinjection: evaluation with contrast-enhanced colorDoppler US. Radiology 1998;209:387–93.

83. Choi D, Lim HK, Kim SH, et al. Hepatocellular carci-noma treated with percutaneous radio-frequencyablation: usefulness of power Doppler US with amicrobubble contrast agent in evaluating therapeu-tic response—preliminary results. Radiology 2000;217:558–63.

84. Kim AY, Choi BI, Kim TK, et al. Hepatocellular car-cinoma: power Doppler US with a contrast agent—preliminary results. Radiology 1998;209:135–40.

85. Wang JH, Lu SN, Changchien CS, et al. Flash echogray scale imaging with subtraction in assessment ofsmall hepatocellular carcinoma treated with percu-taneous ethanol injection. J Ultrasound Med 2001;20:539–44.

86. Wang JH, Lu SN, Tung HD, et al. Flash-echo contrastsonography in the evaluation of response of smallhepatocellular carcinoma to percutaneous ablation.J Clin Ultrasound 2006;34:161–8.

87. Ding H, Kudo M, Onda H, et al. Contrast-enhancedsubtraction harmonic sonography for evaluatingtreatment response in patients with hepatocellularcarcinoma. AJR Am J Roentgenol 2001;176:661–6.

88. Meloni MF, Goldberg SN, Livraghi T, et al.Hepatocellular carcinoma treated with radiofrequencyablation: comparison of pulse inversion contrast-enhanced harmonic sonography, contrast-enhancedpower Doppler sonography, and helical CT. AJR Am JRoentgenol 2001;177:375–80.

89. Wen YL, Kudo M, Zheng RQ, et al. Radiofrequencyablation of hepatocellular carcinoma: therapeuticresponse using contrast-enhanced coded phase-inversion harmonic sonography. AJR Am J Roentgenol2003;181:57–63.

90. Youk JH, Lee JM, Kim CS. Therapeutic response evaluation of malignant hepatic masses treated byinterventional procedures with contrast-enhancedagent detection imaging. J Ultrasound Med 2003;22:911–20.

contrast-enhanced ultrasonography in small liver tumors ( 3cm) · 2017-01-28 · contrast-enhanced...

Documents