Andreas OchsLoretto-Krankenhaus,Freiburg/Germany

Ultrasonography of the Liver

Diagnostics series on liver diseases for the physician with a special interest in hepatology

• Viral Hepatitis Diagnostics (U41e)

J. Rasenack – Freiburg (Germany) (32 pages)

• Ultrasonography of the Liver (U42e)

A. Ochs – Freiburg (Germany) (126 pages)

7th edition 2012© 2012 Falk Foundation e.V.All rights reserved.

Publisher

www.falkfoundation.org

Ultrasonographyof the Liver

Andreas OchsLoretto-Krankenhaus, Freiburg/Germany

Author’s Address:

Prof. Dr. A. OchsInnere MedizinLoretto-KrankenhausMercystr. 6–1479100 FreiburgGermany

2

Indications for ultrasoundexamination of the liver 5

Technique of examination andnormal findings 7

Segmental anatomy of the liver 10

Liver size and variation 21

Diffuse liver disorders 22

“Diffuse liver parenchymal injury” 25

Fatty liver 25

Acute and chronic hepatitis 28

Hepatic congestion 29

Fibrosis and cirrhosis 30Irregular surface 33Portal hypertension 35Ultrasonography of ascites 40

Vascular alterations of the liver 42Budd-Chiari syndrome (BCS) 42Veno-occlusive disease (VOD) 45Portal vein thrombosis (PVT) 46Osler-Weber-Rendu disease 48

Focal liver lesions 50

Hemangioma 53Adenoma 55Focal nodular hyperplasia (FNH) 55Abscess 55Metastases 55

Pseudotumors 56

Hematomas 58

Rare examples 59

Cystic processes 60Liver cysts 60

Abscesses 67

3

Contents

Benign liver tumors 73Hemangioma 73Liver cell adenoma 79Focal nodular hyperplasia (FNH) 81Nodular regenerative hyperplasia 84Rare benign liver tumors 85

Malignant tumors 85Metastases 85Lymphomas 94

Primary malignant liver tumors 95Hepatocellular carcinoma (HCC) 95Cholangiocellular carcinoma(CCC) 102Rare primary tumors 102

The status of ultrasonography indiffuse and focal disorders 104

Diffuse parenchymal disorders 104Focal liver lesions 104Diagnostic approach incholestasis 110

Ultrasound-guided diagnostic andtherapeutic procedures 110

Diagnostic aspiration 110

Ultrasound-guided therapy 112Aspiration of ascites andpleural effusion 112Percutaneous cholangiography/drainage 114Abscess and cyst drainage,sclerotherapy 114Percutaneous therapy ofprimary tumors and metastases 116Guidance of TIPS 116

References 117

4

Ultrasound of the liver is painless, radiation-free, repeatable and widely available. It istherefore the investigation of first choicewhen liver disease is suspected. The disad-vantage is the necessity to establish thefindings during the examination, the highdependency on the experience of the exam-iner and the non-standardized image docu-mentation. Table 1 shows the type of ques-tions that can be elucidated by liver ultra-sonography. In combination with colourDoppler ultrasound and the numerous ul-trasound-guided diagnostic and therapeuticprocedures, the range of indications hasbroadened considerably. Great experience isrequired in order to make full use of theseadditional possibilities.If there are positive findings, the question ofappropriate and above all economical sup-plementary diagnostic investigation arises.This is where a high degree of knowledgeabout the accuracy and costs of the individ-

5

Indications for ultrasound examination of the liver

Diffuse liver disorders

Focal liver lesions

Biliary disease

Suspected vascular processes

Upper abdominal symptoms

TMN classification of malignanttumors

Orientation before surgery(anatomy, segment classification,variants of normal)

Monitoring and follow-up

Ultrasound-guided diagnostic andtherapeutic procedures

Table 1: Indications for ultrasoundexamination.

ual investigation methods is required of theultrasound examiner. In general, early biop-sy should be considered when an abnormal-ity is found on ultrasound, as it can shortenthe diagnostic process economically (Pashaet al., 1998).

6

“Curved array” transducers have the advan-tage of a fan-shaped echo propagation witha relatively small contact surface. For routinepurposes a 3.5 MHz transducer is used, andin very slender persons, a 5 MHz transducermay be useful. Modern machines have“tunable“ transducers that allow differentfrequencies. Both the depth of penetrationand the resolution can be improved by tech-nical refinements such as “harmonic imag-ing“ or the combination of fundamentalsound and the “compound scan“ method.Very good artefact suppression is alsoachieved with these techniques. The avail-ability of colour Doppler equipment is ideal;this has opened up a new domain in ultra-sonography of the liver (Seitz et al., 1997).The sector transducer, which offers a goodview into the liver with a minimal contactarea, especially transcostally, has also provenits worth. The sector transducer is essentialin patients after surgery with large dressings

or scars (e.g. during follow-up after livertransplantation).Ultrasound examination of the liver takesplace with the patient in supine position.Examination in the left lateral position canimprove visibility of the liver and especiallyof the hilum in the presence of ascites. Ex-amination of the seated patient from be-hind can also be useful when the anteriorview is restricted.During the examination, the size of the liver,which is usually measured in the midclavicu-lar line, is recorded. The examiner also assess-es the deformability of the liver during inspi-ration and expiration and during one-fingerpalpation. The shape of the liver, the caudalmargin of the lobes of the liver and the echotexture and sound transmission are analyzed.The branches of the portal vein, the arteries,the intra- and extrahepatic bile ducts, the he-patic veins and the gallbladder are also as-sessed. The area around the liver (in the re-

7

Technique of examination and normal findings

gion of the hepatoduodenal ligament) isexamined for lymph nodes, ascites and col-lateral veins. The size of the spleen is also de-termined. Measurement of the diameter ofthe hilum has proved useful, where valuesunder 5 cm are usually considered normal.The liver is examined systematically in differ-ent planes as shown in figure 1. Startingwith a longitudinal scan to the left of thexiphoid process (A in figure 1), the transduc-er is then moved further to the right parallelto the long axis of the body. The patient isasked to inspire as deeply as possible in casethe left or right lobe of the liver is initiallybeneath the costal arch. When the trans-ducer is moved further to the right as far asthe right flank, where the size, shape, inter-nal structure and mobility of the right he-patic lobe can be assessed; furthermore, thegallbladder, porta hepatis with portal vein,the common bile duct and hepatic arterybecome visible, along with the kidney at the

extreme right. The echo texture of the liverand kidney can now be compared, which isimportant in diagnosing fatty liver and cir-rhosis.This is followed by transverse scans in theepigastrium (B), when the transducer ismoved parallel to the costal margin. Using acurved array transducer, it is possible to ex-amine high into the dome of the diaphragmby tilting it in a cranial direction with gentlepressure under the costal margin and withdeep inspiration. By tilting it in the postero-inferior direction, the stellate hepatic veinconfluence, the right branch of the portalvein and the gallbladder become visible insuccession. If this tilt is applied at the ex-treme right, the right kidney can usually beimaged in cross section with the liver asacoustic window.The intercostal scans (C in figure 1) are usedto show the subphrenic areas of the righthepatic lobe. When the diaphragm is ele-

8

9

Fig. 1:Scanning sequence in liver examination.

Inspiration and expiration Inspiration and expiration Inspiration and expiration

vated, e.g. in the case of meteorism, ascitesand/or atelectasis of the right lung, the livercan be displaced cranially under the rightcostal margin, so that it is difficult to imageit using the subcostal scan. In these cases,the intercostal scans offer a way of assess-ing larger areas of the right lobe of the liverand the gallbladder.It is also important to examine dynamicallyin different intercostal spaces both from cra-nial to caudal and from ventral to dorsalduring inspiration and expiration. In somecases, an incomplete left lateral decubitusposition may be helpful; this causes the liverto “drop“ medial and caudal to the ribs.

Segmental anatomy of the liver

The Couinaud system of liver segments (fig-ure 2), which can be removed deliberatelysurgically, is used predominantly today. Lo-calizing space-occupying lesions accordingto their segmental anatomy affects thesurgical procedure. The lines of separationare formed by the sites of intersection ofbranches of the portal and hepatic veins.Because of the strictly horizontal scanning,segmental classification is easier on comput-ed tomography. A detailed description ofthe segmental anatomy can be found inFrank (1992) and Takayasu and Okuda(1997).Typical scans are shown below in the otherillustrations. In each case, diagrams explainthe represented structures and the segments.Figure 3 shows a longitudinal scan over theinferior vena cava. The caudate lobe is sepa-rated from the left lobe of the liver by a fine

10

septum. The size of the caudate lobe (seg-ment I) is assessed when cirrhosis or Budd-Chiari syndrome is suspected. The ligamen-tum teres (LT) is normally the obliteratedremnant of the umbilical vein, which canbe reopened in portal hypertension (seebelow).Figure 4 shows a longitudinal scan in theaxis of the left main branch of the portalvein. Portal veins are easy to distinguishfrom hepatic veins, as they have strong wallreflections which derive from the accompa-nying bile ducts and arteries. High-resolu-tion transducers separate these wall reflec-tions and must not be mistaken for the“double barrelled phenomenon” found inmechanical cholestasis.Figure 5 shows an almost strict longitudinalscan over the gallbladder. Here, the degreeof filling, the wall thickness, the contents(stones, polyps, tumor) and possible tender-ness are examined.

11

Fig. 2:Liver segments after Couinaud. Thesegments result from the intersections ofhepatic veins and branches of the portalvein. Segment I corresponds to the caudatelobe; segment IV to the quadrate lobe.The latter is bounded on the right by thegallbladder (not shown here) and on theleft by the left main portal vein.

12

Fig. 3:VCI = IVCARD = RRALongitudinal scan over the inferior vena cava (IVC) which appears dorsally at the level ofthe right renal artery (RRA). I = segment I, corresponding to the caudate lobe. The left mainbranch of the portal vein is continued in the ligamentum teres (LT). PV designates the extra-hepatic portal vein, P the partially shown junction of the head and body of the pancreas.

13

Fig. 4:Longitudinal scan over the left main portal vein (LP). It separates segments II and III of theleft lobe of the liver on the left. IVC (VCI) = inferior vena cava, LHV (LLV) = left hepatic vein.

14

Fig. 5:Longitudinal scan over the gallbladder (GB). RPV = right main portal vein,VCI = inferior vena cava.

Figure 6 shows a transverse section at thelevel of the right main portal vein, whichusually follows a horizontal course. Undergood acoustic conditions, this scan offers awealth of information. It can show seg-ments IV, V and VI and a cross section of theright kidney.Figure 7 shows a view of the costal marginin deep inspiration. This shows the “hepaticvenous star”. The width of the hepatic veins,which can be identified easily by the ab-sence of the reflection from the wall, variesgreatly. Nevertheless, their width in the vicin-ity of the inferior vena cava is seldom greaterthan 1 cm. They usually have an extendedcourse, and the vena cava demonstratesdouble pulsations. In cirrhosis (see below)and severe fatty liver, the hepatic veins arecompressed, and in Budd-Chiari syndromethey are occluded by thrombosis.

Figure 8 shows a slightly oblique scan, todemonstrate the gallbladder bed and theleft main portal vein. Segment IV (quadratelobe) is bounded by these structures. A re-duction in the size of the segment is said tobe a sign of cirrhosis. The transverse diame-ter is measured at the point where the leftportal vein is joined by the branch from seg-ment IV.Figure 9 shows an intercostal scan takenrelatively cranially. The white line passesthrough the transversely encountered righthepatic vein (RH) and the branching of theright portal vein (RP). This marks the bound-ary of segments V and VIII.Figure 10 also shows an intercostal scan,which demonstrates the right kidney andsegment VI.

15

16

Fig. 6:Transverse scan in the axis of the right portal vein (RPV). LL = left lobe of the liver,PS (PT) = tail of pancreas, SMA (AMS) = superior mesenteric artery, Ao = aorta,Sp (WS) = spinal column, RK (RN) = right kidney, VCI = inferior vena cava,segments IV (quadrate lobe), V and VI.

17

Fig. 7:Transverse scan along the right costal margin to show the “hepatic venous star”, whichdetermines the corresponding segments. Dorsal to the right hepatic vein lies segment VII,segment VIII lies between the right and middle hepatic veins, and segment IV betweenthe two branches of the middle hepatic vein. VCI = inferior vena cava, H (C) = heart.

18

Fig. 8:Demonstration of segment IV, which is bounded by the gallbladder bed (arrow) and theleft portal vein. The measurement line runs at the level of segmental branch IV and isnormal at 5 cm (RK [RN] = right kidney).

19

Fig. 9:Cranially positioned intercostal scan. Branching of the right portal vein (RP), right hepaticvein (RH) [RL], VCI = inferior vena cava.

20

Fig. 10:Intercostal scan over the right kidney (RK) [RN]. Parts of segment VI are becoming visible.

Liver size and variation

Measurement of the vertical and deep di-ameters in the right midclavicular line (MCL)is usually adequate to determine the size ofthe liver. The vertical diameter is normally12 ± 3 cm, and the deep diameter 9 ± 1 cm.However, marked individual variation can beobserved. For instance, the vertical diameterof the liver in the MCL can be greater than15 cm in a person of asthenic habitus; if it isa slender organ with a small deep diameter,

the total size of the liver is still entirely nor-mal. The Riedel lobe is a normal variant.This can cause the right lobe of the liver toextend far in a caudal direction. When theliver is up to 13 cm in the MCL, about 93%of those examined have a normal liver. Onlywith a length greater than 15 cm in theMCL is there definite hepatomegaly, usuallyas a result of liver damage (Gosink and Ley-master, 1981). The shape of the liver is alsovariable.

21

The size and shape, consistency, surface, in-ternal structure and the intrahepatic vesselsand bile ducts are used in the investigationof diffuse liver disorders. Determination ofthe size is of little value as it depends on themeasurement parameters and, for instance,on the patient’s height. These may be sup-plemented by findings outside the liver, es-pecially signs of portal hypertension whencirrhosis is suspected or the syndrome of hy-persplenism or lymph nodes in the regionof the hepatoduodenal ligament in viralhepatitis (Schwerk, 1987; Forsberg et al.,1987).Table 2 and figure 11 list syndromes associ-ated with diffuse liver changes and the ex-pected findings on ultrasonography.

The sensitivity of ultrasound in diffuse liverdisease depends greatly on the gold stan-dard employed. Biopsy has a sensitivity ofonly 60–70% in cirrhosis compared to la-paroscopy with targeted tissue sampling(Cardi et al., 1997). Reasons for the inferior-ity of biopsy alone include aspiration fromareas which are only slightly altered and in-sufficiently large biopsies, which do notallow fibrosis to be distinguished from com-plete cirrhotic transformation. A biopsy op-timally should contain ten portal areas.

22

Diffuse liver disorders

23

Diffuse disorder Comment

Fatty liver Characteristically echogenic liver with at least 30% fatty infiltra-tion of the hepatocytes. Increasing dorsal echo attenuation.

Acute and No definite sonographic correlate, but acute tends to bechronic hepatitis hypoechoic, and signs of portal hypertension are possible in ful-

minant hepatitis. There is an increased incidence of detection oflymph nodes in the hepatoduodenal ligament andfatty infiltration in hepatitis C.

Acute and chronic Ascites and distended hepatic veins are characteristic, andhepatic congestion fibrosis or cirrhosis develops when long-standing.

Vascular disorders Nonhomogeneous and more echogenic areas and enlargementof the caudate lobe in Budd-Chiari syndrome, ectatic vesselsin Osler’s disease.

Fibrosis Uncharacteristically coarse and uneven internal echoes are possible.Distinction from fatty infiltration and cirrhosis difficult. (Fibroscan®?)

Cirrhosis Irregular surface and signs of portal hypertension are verysensitive and specific, and liver size correlates directly withthe prognosis.

Table 2: “Diffuse” parenchymal liver disorders.

24

Fig. 11:Diagram of the changes in diffuse liver parenchymal alterations which can be observedsonographically.

Normal liver:Sharp edges, fine homogeneousinternal echoes, 12–15 cm in themidclavicular line (MCL)

Fatty liver:Blunt edges, enlarged up to 20 cm inthe MCL, echogenic internal structure,dorsal echo attenuation, hepatic veinsbarely distinguishable

Cirrhosis:In the initial stage enlarged, finely to coarselynodular surface, coarse internal echoes,poorly distinguished hepatic veins

“Diffuse liver parenchymal injury”

Ultrasound reports very often contain diag-noses such as “diffuse liver parenchymaldamage” or descriptions such as “roundedinferior margin”. There are hardly any nor-mal reports among beginners. Our owninvestigations showed that alterations ofshape alone, increased echogenicity or theuse of the concept of “diffuse liver parenchy-mal damage” are seldom associated withgenuine liver pathology, when biopsy orthe transaminases are used for comparison(Ochs et al., 1994). Only when a definitivediagnosis such as fatty liver, hepatic conges-tion or cirrhosis is made sonographicallydoes this correlate with real pathologicalchanges in more than 70% of cases.

Fatty liver

Fatty infiltration of the liver can be foundwhen there is at least 30% fatty infiltrationof hepatocytes (Joseph et al., 1991; Savery-muttu et al., 1986). A fatty liver is usuallyenlarged with blunt edges and increasedechogenicity. There is posterior echo atten-uation, so that the diaphragm cannot bevisualized with older equipment even withthe gain turned up. The hepatic veins aremore difficult to identify than in the normalliver (figures 12 and 13). With severe fattyliver hepatitis, portal hypertension can occur.Regionally, there can be reversed flow in theintrahepatic portal veins. A fatty liver can becompressed by finger palpation under vi-sion. Ultrasound is superior to enzyme mea-surement in diagnosing fatty liver (Petritschet al., 1987; Steinmaurer et al., 1984). Thedegree of fatty infiltration can be estimatedvery well on ultrasound (Ricci et al., 1997).

25

26

Fig. 12:Transverse scan through a markedly fattyliver. The echogenic internal hepaticstructure, the increasing posterior echoattenuation and the poorly demarcatedhepatic veins are characteristic.

Fig. 13:Transverse scan of a fatty liver showingthe right kidney (N = K) and the gallbladder(GB). Markedly increased echogenicityof the liver parenchyma compared tothe kidney.

Sonography is also an important buildingblock in establishing the diagnosis of non-alcoholic fatty liver (Ahima, 2007).Nonhomogeneous greater or lesser degreesof fatty infiltration can be mistaken for focallesions and may induce complex investiga-tions for diagnosis. Focal disorders of fatdistribution are found in metabolic disorders(diabetes mellitus, obesity, alcohol abuse),after chemotherapy and with vascular alter-ations in the liver. Echo-poor or echogenicareas are usually not sharply demarcatedand are localized around the blood vesselsor gallbladder (figure 14). They have a dy-namic course. Occasionally computed to-mography or biopsy is necessary if a tumoris suspected (see also pseudotumors of theliver).

27

Fig. 14:Fatty liver with regional variation in thedegree of fat infiltration in a patient withobesity and diabetes mellitus.

Acute and chronic hepatitis

In acute hepatitis, the ultrasound appear-ance is uncharacteristic. The importance ofultrasonography lies in excluding mechani-cal cholestasis or a chronic liver disorder(tumor, cirrhosis). The liver is enlarged andthe edge is blunt. The internal structure ishomogeneous and shows fine loosely dis-tributed individual reflections. The echogenic-ity is normal or somewhat reduced. It maybe accompanied by moderate splenomegalyand a contracted gallbladder with thickenedwall. Ascites can occur when the disease issevere, indicating incipient portal hyperten-sion. Lymph nodes in the region of the

hepatoduodenal ligament (Forsberg et al.,1987) are said to correlate with the activityof chronic hepatitis C (Dietrich et al., 1997).In chronic hepatitis, the individual reflec-tions are coarser, leading to an increase inechogenicity. In chronic hepatitis C, increasedechogenicity on ultrasound corresponds tofatty infiltration on biopsy. There is no sono-graphic substrate for the extent of the in-flammatory activity or fibrosis (Dietrich etal., 1998). Nonhomogeneity can occur sub-sequently, which can extend to nodularchanges. Analyses of texture, even comput-er-aided, have not hitherto become clinicalroutine (Taylor et al., 1986).

28

Hepatic congestion

Acute and chronic hepatic congestion can becharacterized well on ultrasound. Figure 15shows widely congested hepatic veins andascites. When it is long-standing, fibrosis orcirrhosis can develop (cardiac cirrhosis). As-cites is nearly always present. Protein mea-surement in ascites gives levels of 4 g/100 mlor more. Colour Doppler ultrasound showsslow but undulating flow in the portal vein.Figure 15 shows acute hepatic congestionwith distended hepatic veins. Figure 16 showschronic hepatic congestion, early cirrhosis ispossible.

29

Fig. 15:Acute hepatic congestion with distendedhepatic veins and noncollapsing inferiorvena cava (VCI).A = ascites, HV (LV) = hepatic veins

Fibrosis and cirrhosis

The sonographic appearance in fibrosis andcirrhosis is variable depending on the de-gree of the pathological anatomical changes.Early cirrhosis may not be apparent on ultra-sound. However, it is also possible that he-patomegaly, a rounded edge and increasedechogenicity with coarsening of the individ-ual reflections are present, giving a sono-graphic appearance similar to that foundwith fatty infiltration. The transition from fi-brosis (reversible) to cirrhosis (irreversible) iscontinuous. When several non-invasive pa-rameters are included (ultrasound findings,serum fibrosis markers), a sensitivity of 87%for diagnosing cirrhosis can be achieved(Oberti et al., 1997; Gaiani et al., 1997; DiLelio et al., 1989). The severity of the fibro-sis or cirrhosis can be determined with amethod of determining liver stiffness knownas transient elastography. The correlation of

30

Fig. 16:Chronic hepatic congestion in cardiomy-opathy due to secondary hemochromatosis.The inferior vena cava does not collapse,the inferior edge of the liver is blunted,and there is ascites (A). The cirrhosis haspossibly developed because of the conges-tion and hemochromatosis; however,there is obvious undulation of flow in theportal vein on colour Doppler ultrasound.

Fibroscan® with liver histology is particularlygood in homogeneous conditions such ashepatitis C; even the degree of portal hy-pertension can be deduced (Vizzutti et al.,2007).Table 3 and figure 17 reproduce importantdirect and indirect signs. Each sign on itsown usually has low sensitivity and specifici-ty. Considered overall, an experienced ex-aminer assisted by colour Doppler sonogra-phy can make diagnoses which exceed thequality of biopsy. The gold standard in thediagnosis of cirrhosis continues to be la-paroscopy. In cirrhosis induced by hepatitisC, the diagnosis is made by biopsy alone inonly 60–70% of cases. Further studies of Fi-broscan® must be conducted before biopsycan be omitted. Obesity and ascites limitthis method (Friedrich-Rust and Zeuzem,2007).

31

Sonographic criteria of cirrhosis:

Liver size variable

Enlargement: early cirrhosis

“Shrunken liver”: end stage

Irregular surface

Nonhomogeneous internal structure

Coarse echo texture

Compressed hepatic veins

Irregular calibre of the intrahepaticportal vein branches

Pseudo double-barrelled appearance(prominent arteries)

Narrowing of segment IV

Enlargement of the caudate lobe

Thickening of the wall of the gall-bladder

Lax gallbladder (often with stones)

Table 3

32

Fig. 17:Findings in cirrhosis which can be made with conventional ultrasound.

Ascites

Thickened gallbladder wall,sludge, gallstones

Pseudo double-barrelledappearance (prominentarteries)

Coarse internal echoes

Nodular surface Compressedhepatic veins Collaterals

Splenomegaly

*enlarged caudate lobe

Distended splenic vein orportal vein without respira-tory fluctuations in diameter

Irregular surface

The irregularity of the hepatic surface canbe demonstrated (with high-resolution trans-ducers) even without ascites (Di Lelio et al.,1989). This irregularity has a high positivepredictive value of up to 84% in viral cirrho-sis (Oberti et al., 1997; Gaiani et al., 1997).Ultrasound parameters can be combinedwith serum parameters. Here, measurementof the hyaluronic acid appears to give addi-tional information with regard to fibrosingactivity. The diagnosis of cirrhosis is easywhen there is ascites and a grossly alteredsurface (figure 18).Coarse internal echoes are an expression offatty infiltration, fibrosis or regenerative nod-ules. Various liver segment diameters havealso been utilized. For instance, the trans-verse diameter of segment IV (quadrate lobe,figure 8) is thought to be only 28 ± 9 mm incirrhosis, and 43 ± 8 mm in normal persons

33

Fig. 18:Decompensated liver cirrhosis. The irregularsurface and nonhomogeneous internalstructure are striking.

(Lafortune et al., 1998). The relationshipof the caudate lobe to the left lobe of theliver is also regarded as a sign of cirrhosis(figure 19) (Giorgio et al., 1986). An en-larged cirrhotic liver has a favourable prog-nosis in contrast to the small coarsely nodu-lar liver (Zoli et al., 1990).Thickening of the gallbladder wall is ob-served with and without ascites (figure 20).A wall thickness greater than 4 mm is foundin more than half of patients with cirrhosis.The wall thickness correlates with the Child-Pugh class and with the presence of ascites(Wang et al., 1997). On colour Doppler ul-trasound, gallbladder varices can sometimesbe detected as the cause of the wall thick-ening.

34

Fig. 19:Markedly enlarged caudate lobe (LC)(segment I) in cirrhosis.

Portal hypertension

The detection of signs of portal hyperten-sion increases the accuracy of the diagnosisof cirrhosis on ultrasound (Vilgrain et al.,1990). Figure 17 and table 4 give a summa-ry of the signs and parameters of portalhypertension. Only some cases of cirrhosisare associated with portal hypertension;the reasons for this are unclear. If colourDoppler ultrasonography is available, portalhypertension can be best quantified by thecongestion index, which includes the velo-city of flow in the portal vein and the cross-sectional area of the portal vein (Moriyasuet al., 1986). However, this has not becomeaccepted as routine.Portal hypertension is often accompaniedby splenomegaly (figure 21). However,splenomegaly is not essential. A normal-sized or only slightly enlarged spleen does

35

Fig. 20:Markedly thickened gallbladder wall incirrhosis and ascites. Collateral vesselswere found on colour Doppler sonography.

36

Splenomegaly

Ascites

Distension of the portal vein (> 1.3 cm) and veins draining into it(splenic vein, superior mesenteric vein)

Pathological congestion index (CDU)

Absence of respiratory fluctuations in diameter (< 10%)

Absence of compressibility of these vessels

Increase in diameter of portal vein at the hilum of the liver

Portosystemic collaterals (ventricular coronary vein)

Patent umbilical vein

Thickening of gastric wall

Thickening of gallbladder wall

Spontaneous splenorenal shunt

Table 4: Sonographically detectable signs of portal hypertension(CDU = colour Doppler ultrasound).

not exclude hypersplenism or portal hyper-tension. Particularly marked splenomegalyof over 8 cm in the region of the hilumshould suggest splenic vein thrombosis, anadditional hematological disorder, idiopathicnon-cirrhotic portal hypertension or schisto-somiasis.In portal hypertension, the portal vein isusually distended to over 1.3 cm. The splenicvein (figure 22) or superior mesenteric veinno longer demonstrate compressibility andthe fluctuation in size with respiration is lessthan 15%.

37

Fig. 21:Indirect sign of cirrhosis: markedsplenomegaly with a diameter at thehilum of 7.3 cm (normal up to 4.5 cm).

38

Fig. 23:Large-diameter collaterals projected ontothe head of the pancreas. Duodenal variceswere seen endoscopically. Con = conflu-ence, C = collaterals, LL = left hepatic lobe.

Fig. 22:Indirect sign of cirrhosis due to finding ofportal hypertension with splenic vein upto 1.6 cm in diameter projected onto thepancreas.

Collateral vessels can also be demonstratedif sought precisely even without colourDoppler ultrasound. At the splenic hilum,they can be equivalent to a spontaneoussplenorenal shunt or gastric varices. Theventricular coronary vein or other collateralvessels can sometimes be demonstratedoriginating from the confluence (figure 23).The finding of a patent umbilical vein (Cru-veilhier-Baumgarten syndrome, figure 24) orof paraumbilical veins is proof of portalhypertension. Grape-like collaterals goingin the retroperitoneal direction can be thesource of severe intra-abdominal hemor-rhage (figure 25).A fresh portal vein thrombosis shows an ex-tension of the vessel, filled with echogenicmaterial, while the so-called cavernomatoustransformation develops subsequently (seevascular alterations).

39

Fig. 24:Patent umbilical vein in cirrhosis(Cruveilhier-Baumgarten syndrome).LL = left hepatic lobe.

Ultrasonography of ascites

Small volumes of ascites are identified as anarrow perihepatic or perisplenic echo-freerim. When ascites is first detected and com-plications are suspected, diagnostic para-centesis should be performed.Sonographic detection of ascites is greatlysuperior to percussion. In 80% of cases, as-cites has a hepatic cause. In 10% of cases,there is an underlying malignancy. The re-maining 10% are distributed through renal,pancreatic, tuberculous, cardiac and otherrare causes. Fibrin threads indicate an in-flammatory process. They are also observedin patients on peritoneal dialysis. Sedimen-tations can be found following hemorrhageafter ruptured spleen, ascites aspiration andbursting of intra-abdominal collaterals orspontaneously ruptured hepatocellular car-cinoma. Diagnostic paracentesis gives im-portant information.

40

Fig. 25:Large collaterals projected on to the rightrenal hilum. Flow in the dorsal directionwas found on colour Doppler ultrasound.In addition, ascites is present.

Spontaneous bacterial peritonitis (SBP) is acomplication of ascites. The diagnosis is alsomade by diagnostic paracentesis. The ascitescan appear sonographically as non-echofree(figure 26).After frequent paracentesis or after opera-tions with peritonitis, compartments can formwhich make further paracentesis impossible(figure 27).

41

Fig. 26:Non-echofree ascites as expression of SBP.The patient complained acutely of abdomi-nal pain. Aspiration yielded a leucocytecount of 6000/µl.DS = BL = bowel loop.

Vascular alterations of the liver

Budd-Chiari syndrome (BCS)

This syndrome consists of postsinusoidalportal hypertension, due to occlusion or dis-placement of the hepatic veins. The princi-pal pathological mechanism in our latitudesis thrombotic occlusion due to hematologi-cal disorders or genetic coagulation defects(Menon et al., 2004). There is also an in-creased incidence in female smokers takingcontraceptives. Obstruction of hepatic veinsby membranes (“webs”) is seen rarely inWestern Europe. Other causes are compres-sion due to primary tumors or metastases.Rarely, there is a primary sarcoma of the he-patic veins or inferior vena cava.

42

Fig. 27:Loculated ascites in a cirrhotic patientas a consequence of a complicated chole-cystectomy. There are honeycombedcompartments under the right diaphragm.

The clinical syndrome is fulminant in theform of hepatic and renal failure with as-cites or subacute or chronic. In the lattercase, cirrhosis is often diagnosed incorrectly.Complete and incomplete causes of BCSare being diagnosed increasingly frequent-ly by ultrasound and especially by colourDoppler sonography.BCS can be identified on conventional ultra-sound also (Braun et al. 1983). The criteriaare summarized in table 5. Brancatelli et al.have compiled a collection of characteristicimaging findings (2007).Figures 28–30 show examples of BCS. Thecaudate lobe can appear hypoechoic like atumor and compress the inferior vena cava,leading to thrombosis in this region.

43

Ascites (always present)

Enlargement of the liver(acutely painful)

Absence of imaging ofhepatic veins

Enlarged caudate lobe(with compression of theinferior vena cava)

Intrahepatic portal venous shunts

Tumor-like parenchymalnonhomogeneity(regeneration nodules, nodularregenerative hyperplasia)

Table 5:Ultrasound findings in BCS.

44

Fig. 28:Subacute BCS with ascites, narrowedechogenic hepatic veins (LV = HV) andthickened gallbladder wall (GB).

Fig. 29:Acute BCS. Longitudinal scan over theinferior vena cava (VCI) which is greatlycompressed by the acutely enlargedcaudate lobe. This led to obstruction offlow beneath it. The problem was relievedfully by a transjugular portosystemic stentshunt (TIPS).

Veno-occlusive disease (VOD)

VOD has a similar clinical appearance toBudd-Chiari syndrome. However, the largehepatic veins are patent. Toxic and autoim-mune processes lead to fibrotic and inflam-matory occlusion of small hepatic veins.There is usually icterus and ascites clinically.Multiorgan failure can develop rapidly. VODis found in up to 50% of patients after allo-genic bone marrow transplantation. The di-agnosis is difficult on conventional sonogra-phy. Hepatosplenomegaly, thickening of thegallbladder wall, a wide portal vein, smallhepatic veins, ascites and a patent umbilicalvein are typical. Colour Doppler ultrasoundcan detect the portal hypertension and, inparticular, the reversed intrahepatic portalvein flow. Flow in the portal vein is demodu-lated and the RI of the hepatic arteries isover 0.75 (Lassau et al., 1997). The diagno-sis is made by biopsy. A transjugular biopsy

45

Fig. 30:Longitudinal scan over the inferior venacava (VCI = IVC). BCS for six months. Thecaudate lobe (LC = CL) has large veins andthus partially compensates the abnormalvenous outflow. (Con = confluence).

can be taken in liver failure and if therapeu-tically indicated. An important differentialdiagnosis is toxic cardiac damage. Raisedtransaminases, ascites and hepatomegalyare found in this situation also. However,the hepatic veins and the vena cava arecongested. The portal vein demonstrates re-duced but pulsatile flow indicating unim-peded sinusoidal perfusion.

Portal vein thrombosis (PVT)

A complete or incomplete occlusion can befound acutely or subclinically in patientswith and without cirrhosis (figure 31). In pa-tients with cirrhosis, acute portal vein throm-bosis is always suspicious for the presenceof a hepatocellular carcinoma (HCC) untilproven otherwise. If vessels are found inthe thrombus by Doppler ultrasound, thediagnosis of HCC can be regarded as con-firmed.

46

Fig. 31:Fresh incomplete portal vein thrombosis.Even without colour Doppler ultrasound,echogenic areas can be found in theportal vein.

In non-cirrhotic PVT, there is an increased in-cidence of genetic coagulation factors (e.g.factor V disease, see also Budd-Chiari syn-drome). The extent of the thrombosis is di-vided into four stages, depending on thedegree of thrombosis. In stage IV, the portalvein, splenic vein and superior mesentericvein are occluded. There is little informationabout the natural course of PVT. Smallerintrahepatic thromboses can resolve or canbecome the origin of periportal fibrosis. Thisis how non-cirrhotic portal hypertension isbelieved to develop. Larger thromboses canprogress to cavernomatous transformationwithin a few weeks (figure 32, Braun et al.,

1984). If this has occurred, a liver transplantor surgical shunt procedure can becometechnically impossible. The liver then re-ceives its portal supply through collaterals(e.g. gallbladder wall) or capsular vessels.The main blood supply then comes from thearteries. The transformation can involve thesplenic and mesenteric veins. There is then adanger of fundal variceal bleeding. Collater-als in the region of the splenic hilum in theform of a gastroportal or gastrocaval shuntor a spontaneous splenorenal shunt cansometimes be shown with conventional ul-trasound and particularly well with colourDoppler sonography (Seitz et al., 1997).

47

Osler-Weber-Rendu disease

This is a rare hereditary disorder in whichmultiple angiodysplasias occur, which usu-ally become apparent in the gastrointestinaltract. Corkscrew-like vascular malformationscan be detectable in the liver (figure 33).Large arteriovenous shunts can develop,leading to heart failure. Cirrhosis can devel-op as a result of this disorder. The syndromeof hypersplenism can produce severe hem-orrhage, including intracranial bleeding, dueto thrombocytopenia. Hepatic involvementcan be diagnosed when the diameter of thecommon hepatic artery is more than 1 cm(Caselitz et al., 2003).

48

Fig. 32:Cavernomatous transformation of theportal vein. This oblique scan over the he-patoduodenal ligament does not show anyecho-free vascular structure. The portal veinrather is replaced by echogenic material.On colour Doppler sonography, numerouscollaterals could be seen over the livercapsule and around the former portal vein.

49

Fig. 33:Typical intrahepatic vascular malformations in Osler’s disease. Colour-coded ultrasonographyshows numerous arteriovenous shunts.

The search for focal lesions and the inter-pretation of focal abnormalities with theirdiagnostic and therapeutic consequencesrepresent the main task of liver ultrasonog-raphy. Figure 34 shows the variety of possiblefocal lesions. Focal lesions can be demon-strated according to their frequency, theirultrasonographic appearance or their histo-logical classification. No system entirely meetsthe real situation. In the concrete case, theexaminer is faced with two situations:• the search for a liver focus• finding a liver focus.

Algorithms are given below for both sets ofproblems. The basic question in looking fora liver lesion is whether there is a metastasisfrom a known primary tumor. A hepaticspace-occupying lesion is also sought inconditions which suggest a tumor or in thecase of recent cholestasis. If a lesion isfound, investigation can often be shortened

by ultrasound-guided biopsy. If no lesion isfound when there is a high degree of suspi-cion, the examiner must initiate the nextuseful investigations.The incidental finding of a hepatic lesion isusually followed by a “wait and see” atti-tude if the focus lacks signs of malignancy.Thus, the examiner must consider whetherhe can give a firm diagnosis using ultra-sound. This decision depends greatly on theexaminer’s experience. As figure 34 shows,multiple histological conditions (Altmann,1994) contrast with only few unequivocalultrasonographic appearances in differentialdiagnosis.The finding of a focal hepatic lesion in prin-ciple always opens up the entire range ofdifferential diagnosis. In practice, however,the findings are limited to a few differentialdiagnoses. Table 6 shows that the majorityof detected focal lesions are metastases orcysts, which do not cause problems in the

50

Focal liver lesions

51

Fig. 34:Tumors and tumor-like lesions of the liver, classified according to histological criteria.The findings which are encountered most frequently are shaded in pink.

Benign non-epithelial tumors:

AngiomyolipomaHamartomaInfantile hemangioendothelioma,other

Hemangioma, lipoma

Malignant epithelial tumors:

Biliary cystadenocarcinomaHepatoblastoma, mixedother carcinomas, carcinoid tumor

MetastasesHepatocellular carcinoma(fibrolamellar type)Cholangiocellular carcinoma

Malignant non-epithelial tumors:

Embryonal and rhabdomyosarcoma

MetastasesLymphomas, epithelioidHemangioendothelioma

Benign epithelial tumors

Bile duct adenomaBiliary cystadenomaPapillomatosis of the bile duct

Tumors and tumor-like lesions of the liver

Hepatocellular adenoma

Dysontogenetic cystsTraumatic cystsEchinococcal cystsAbscesses (pyogenic, amoebic)

Peliosis hepatis, Osler’s disease,Caroli’s syndrome

Focal nodular hyperplasiaNodular regenerative hyperplasiaRegenerative nodulesInfarct, hematoma, calcificationFocal fatty infiltration, major or minorExtrahepatic tumors

AdenomaMetastasis

Hemangioma

Cyst

clinical context (Weiss 1987). 180 hepaticlesions were unclear initially. Further investi-gation then resulted predominantly in metas-tases, cysts or hemangiomas. The finding ofhepatic lesions is also greatly dependenton the patient groups investigated. In livercentres, hepatocellular carcinomas are seenvery frequently, while they are a rarity in ageneral practice.The introduction of echo contrast enhancershas brought about a revolution in ultra-sonography, especially in the area of theliver (Wermcke, 2006).Great progress was achieved with echo con-trast or echo signal enhancers (CEUS) in theimproved detection and characterization ofspace-occupying lesions. There are two echosignal enhancer products on the market,Levovist® und Sonovue®, and others are inthe pipeline. With corresponding hardware,which works with phase-inverted soundapplication, the reflections are extinguished

52

Total 75,840

Liver focal lesions 180

No abnormality onrepeat examination 35 20%

Metastases 59 33%

Hemangiomas 24 14%

Cysts 26 14%

HCC 8 4.4%

Adenomas 5 2.8%

Echinococcus/abscesses 3 1.5%

Other lesions 20 11%

Table 6:Frequency of causes of 180 initiallyunclarified focal lesions (Weiss 1987) in75,840 ultrasound scans.

on linear reflectors such as liver cells andgive a virtually absent signal. The gas bub-bles of the echo signal enhancers are non-linear reflectors that are stimulated to vi-brate by the sound waves. They reflect non-linearly as the compression and expressionhave slight differences. Because of this, thereflections are not extinguished and give asignal. Levovist® can be sonicated only withhigh sound energy (high mechanical index,MI); when the transducer is swivelled, thebubbles burst, producing what is known asa burst. An examination with low MI is pos-sible with Sonovue®; the bubbles only vi-brate and remain detectable for minutes.Thus, observation of the perfusion of boththe liver and the space-occupying lesion ispossible. Triple-phase examination is alsoused in computed tomography, with a so-called spiral “driven“ in each phase withcorresponding radiation burden. Figure 35 a

shows these phases. The arteries appearabout 15 sec after injection of an echosignal enhancer. This phase is important incirrhosis, adenoma and in metastases andhepatocellular carcinoma (HCC). Metastasescan “wash out“ the contrast in the portalvenous phase, which occurs in the latephase or parenchymal phase at the latest.In hepatic tissue there are sinusoids in whichthe bubbles can persist for a long time.Non-hepatic tissue does not have thesesinusoids so that these lesions are apparentas spaces in the late phase.Figure 35 b shows in diagram form the fol-lowing echo signal enhancers.

Hemangioma: There is so-called globularenhancement in the early arterial phase.Later the lesion “flows“ from without in-wards (iris phenomenon). Non-stained partscan correspond to scars or thrombi.

53

54

Fig. 35 a:Dynamics in triple-phase sonography withecho contrast enhancers. As in computedtomography, three phases of contrast up-take can be distinguished. Examples ofcharacterization of space-occupying lesionsare shown.

0 15 30 60 > 120 sec

HCCFNH

FNH

Arterial Late phase

AdenomaCirrhosis

Metastasis

CirrhosisThrombosisMetastasis

Hemangioma

Metastasis

Injection

Inte

nsity

Echo contrast enhancer

Portal venous

Fig. 35 b:Behavior of different lesions after echosignal enhancers. Native or fundamentalmode on the left, then the arterial, portalvenous, and late phase (see also figure 35 a).

Adenoma: the domain of CEUS. It is im-portant that the lesions are sometimes visi-ble only for seconds in the early arterial con-trast phase and later become isoechogenicwith the surrounding liver tissue. They thusbecome invisible .

Focal nodular hyperplasia (FNH): Spoke-like arteries are strongly enhanced in theearly arterial phase. The lesion fills earlywith contrast and usually remains some-what more echogenic than the surroundingparenchyma in the late phase.

Abscess: The native scan shows hypoechoicto anechoic cystic areas. After echo signalenhancers the margin can take up contrast,while the lacunar areas inside the abscessalways remain anechoic and appear punchedout.

Metastases: They have only one arterial in-flow and one venous outflow. They cantake up contrast arterially. In the portalvenous phase and at the latest in theparenchymal phase, the metastases “washout“ the contrast and become anechoic.The metastases of neuroendocrine tumorsstain particularly highly arterially. If there is ahigh chromogranin A in the blood, the diag-nosis of such a tumor is highly likely.

Focal lesions appear ultrasonographicallywhen they are distinguished from the sur-rounding hepatic tissue by their echo tex-ture. Space-occupying lesions can becomeapparent indirectly by alterations in liversize, compression of blood vessels or biliaryducts or irregularity of the surface. Calcifi-cations are apparent above about 3 mm,cysts above 3–5 mm, solid lesions above5 mm. However, identification depends notonly on the size of the lesion but also on

55

the equipment employed, the investigator’sexperience and the location of the lesion.Lesions that are directly lateral or subdi-aphragmatic or located at the caudal mar-gin of the liver are difficult to detect. Thedetection of focal lesions through echo sig-nal enhancers is now comparable to com-puted tomography, and the two methodsare partially complementary.

Pseudotumors

With the introduction of imaging, localizedparenchymal alterations which sometimeshave no anatomical pathological correlatebecame visible. Table 7 gives a summary ofultrasonographically detectable phenomenawhich can look like focal lesions or imitatemalignant processes. Occasionally, calcifica-tions are found in the liver, which are distin-guished from echogenic hepatic lesions bytheir posterior echo extinction. Such calcifi-

cations are often then found in other or-gans also, especially in the spleen. They canusually be attributed to previous infectionssuch as tuberculosis. Aerobilia or hepa-tolithiasis must be considered in the differ-ential diagnosis. Calcification of regressedmetastases, hepatocellular carcinoma or inthe wall of echinococcal cysts is also foundoccasionally.Focal major or minor fatty infiltration issometimes mistaken for a tumor. Increasedfocal fatty infiltration is often found particu-larly around the falciform ligament oraround branches of the portal vein; focalminor fatty infiltration is also associatedwith certain vascular areas and with thecaudate lobe (figure 14). Nonhomogeneousfat distribution is found in patients withalcoholic disease, diabetes mellitus, non-alcoholic steatohepatitis (NASH) or duringchemo- or hormone therapy. On computedtomography, areas with different Hounsfield

56

57

Table 7: Alterations which can cause problems in differential diagnosis, appearing onultrasound as circumscribed space-occupying lesions.

In metabolic disorders, chemotherapyor hormone therapy:Nonhomogeneous fatty infiltrationFocal lesions in Wilson’s diseaseAlpha-1-antitrypsin deficiencyType I tyrosinemiaGaucher’s diseasePorphyrias

Anatomical structures:Ligamentum teres, Zahn’s grooves

Postoperative conditions, trauma:Needle biopsy, endoscopy,spontaneousScars after cholecystectomyDefect after resectionClipsAerobiliaSeromaBiliomaGas in the portal veinHematoma/bleeding into lesion

Vascular alterations:Nodular regenerative hyperplasia inBudd-Chiari syndrome (BCS)Nonhomogeneous enlargement of thecaudate lobe in BCSPseudoaneurysm of the portal veinAneurysm of the hepatic arteriesIntrahepatic shuntsFocal lesions in Osler-Weber-RendudiseaseMap-like changes in HELLP syndromePeliosis hepatisLiver infarct (very rare)

Fibrosis/cirrhosis:Macronodular regenerative nodulesConfluent focal fibrosisIrregular surface

Biliary tract abnormalities:HepatolithiasisCaroli’s syndromeBile duct abscesses

Tumors:Extrahepatic tumors

58

units are found. If biopsy is performed,areas of variable echogenicity should be as-pirated. Ideally, a margin between majorand minor fatty infiltration will then be visi-ble histologically. Observation is feasible inthe case of incidental findings. The nonho-mogeneous areas are dynamic, for instance,with better control of diabetes mellitus or ab-stinence from alcohol (see also p. 27, fig. 14).It can be shown with echo signal enhancersthat these areas take up contrast identicallyin the parenchymal phase so that they donot correspond to metastases.

Hematomas

Hematomas in the liver are usually echo-free initially and when they liquefy subse-quently.They usually follow the anatomy and there-fore have irregular outlines. Typically, theyoccur following trauma (figure 36) or after

Fig. 36:Subcapsular, relatively fresh hematomaafter diagnostic laparotomy and cholecys-tectomy. At the same time, there is cav-ernous transformation (CVT) of the portalvein, which had the appearance of a biliarytract tumor on CT, leading to laparotomy.

59

invasive diagnostic procedures. However,spontaneous bleeding occurs with coagula-tion disorders or in the presence of pre-ex-isting liver lesions. The lesions become moreechogenic and better demarcated at firstand can be demonstrable for weeks afterthe trauma. When they liquefy, there isagain a change in the pattern of echogenic-ity. Residual conditions have the appearanceof cysts.

Rare examples

An appearance similar to metastases canoccur in Wilson’s disease (figure 37). Thesechanges correspond histologically to fattyinfiltration and fibrosis, which make the liverappear echogenic. There are hypoechoic is-lands of normal hepatic tissue in between.This appearance is believed to indicate afavourable prognosis for treated Wilson’sdisease (Vogel et al., 1988).

Fig. 37:Multiple hypoechoic lesions in a liverwith Wilson’s disease. No evidence ofmalignancy was found on biopsy.

Cystic processes

Liver cysts

Dysontogenetic liver cysts are found at upto 0.15% of autopsies and on up to 3% ofliver ultrasound scans. The typical liver cyst(figure 38) has a soft edge, a typical dorsalecho enhancement and a completely echo-free interior. If these criteria are met and thisis a coincidental finding, no further investi-gations are necessary. It is called a polycysticliver when there are more than 10–15 cysts.Hepatic metabolic functions are preservedintact for a long time even with multiplecysts. The sensitivity of ultrasound is com-parable to that of CT and MRI. The findingof hemorrhage in polycystic syndromes isbest made with MRI. Liver cysts are oftenassociated with cysts in other organs (kid-ney, pancreas, spleen). This combination isobserved in the Hippel-Lindau syndrome.Larger cysts over approx. 5–10 cm can cause

60

Fig. 38:Typical cyst of the liver. The interior isecho-free, the wall is thin, and there isposterior echo enhancement.

symptoms and lead to slightly raised para-meters of cholestasis (figure 38 a + b). Theycan be sclerosed under ultrasound controlwith ethanol or saline (see interventionaltherapy). The differential diagnosis of cystsincludes Echinococcus cysticus, early, almostecho-free metastases (melanoma, lymphoma,ovarian and esophageal carcinoma, carci-noid, cystadenoma, cystadenocarcinoma),Caroli’s syndrome, liquefied hematomas andabscesses. Diagnostic aspiration is sometimesrequired to aid diagnosis.

61

Fig. 38 a:Very large liver cyst with a diameter of17 cm which gave rise to upper abdominalsymptoms (N = right kidney).

Echinococcosis:Echinococcus cysticus is relatively rare inGermany. In the typical case, several daugh-ter cysts are found in a cyst bordered by arelatively echogenic rim. Echinococcal cystsare occasionally found in other organs.Classification of stage can be deduced fromthe sonomorphology (Lewall, 1998; Strohmand Weimer, 1997; Di Matteo et al., 1996;Caremani et al., 1996; Richter et al., 2003).Sonomorphology is relatively typical so thatin experienced hands the correct diagnosiscan be made rapidly together with theserology. The WHO has proposed a classifi-cation for cystic lesions (figure 39). In thisoverview, the typical cyst (CL) is seen onthe left. This is followed by the stages ofEchinococcus cysticus:CE1 Ovular cysts with thicker wall than

dysontogenetic cysts, but at this ini-tial stage, they can be mistaken forliver cysts (figure 39).

62

Fig. 38 b:Large liver cyst showing sludge-like material.Bleeding into the cyst, infection or bile-likesecretion can cause this phenomenon.

CE2 Cysts with daughter cysts (figure 39and 40). CE1 and CE2 are regardedas the active form.

CE3 Honeycomb structure or pathogno-monic wheel spoke structure (figure39, 41 and 42). In this case, the abili-ty to maintain the pressure in thecysts was lost; as a result, the mem-branes are floating in the main cyst.The activity is transitional.

CE4 Solid tumor with possible calcifica-tions. Degenerative late stage, wherethe solid parts are possibly due toearlier bacterial superinfections. Cystwalls lying close together give theappearance of double membranes.

The serological test (an IgG ELISA) reaches asensitivity of 85–94% and a specificity of99%.

Up to a few years ago, surgery was regardedas the treatment of choice of cysts. Today, lessinvasive therapies such as (ultrasound-guided)puncture, aspiration, injection and reaspira-tion (PAIR, McManus et al., 2003) are pre-ferred in the cystic stages. The cysts are scle-rosed with 95% ethanol. Solid manifestationscan be converted to cystic forms by medica-tions in some cases and then treated by PAIR.Radiofrequency thermoablation has also beenused.Echinococcus alveolaris must be distinguishedfrom the cystic form; it appears more like adiffusely infiltrating tumor and is not anechoic(figure 43).The region of the hilum is usually affected,and E. alveolaris cannot be distinguished inappearance from a central bile duct carcino-ma. The diagnosis is often made only atdiagnostic laparotomy. If the diagnosis ismade in time, treatment is medical in orderto limit spread as much as possible. The

63

64

Fig. 39:Summary of the cystic lesions according to the WHO. CL = cystic lesion. CE1 and CE2correspond to the active form of Echinococcus cysticus. Large solitary cysts can be mistakenfor harmless dysontogenetic cysts. Stage CE3 is designated as transitional. CE4 and CE5 areregarded as inactive.

65

Fig. 40:Echinococcus cysticus, stage CE2. There are daughter cysts in the main cyst.

66

Fig. 41:Stage CE3: The honeycomb structurepredominates.

Fig. 42:The honeycomb structure of stage CE3 isalso shown well by computed tomography.

most radical surgery possible is accompa-nied by two-year follow-up treatment withalbendazole (McManus et al., 2003).

Abscesses

Two large groups can be distinguished: pyo-genic and parasitic abscesses. There is an in-creased incidence of pyogenic abscesses inthe 5th to 7th decades, especially in immuno-suppressed (AIDS, chemotherapy) or oper-ated patients (biliary procedures, biliarymalignancies, interventions). Diverticulitis orother foci can be the source. The ascendingbiliary form predominates (up to 70%),compared to hematogenous dissemination(up to 30%). Pyogenic abscesses are solitaryin only half of cases (figure 44) and 70% ofthem are found in the right lobe. Aspirationprovides a microbiological result in 80%of cases (usually E. coli, Klebsiellas, also en-terococci, streptococci). Mortality is up to

67

Fig. 43:Echinococcus alveolaris. The disease oftenappears at the liver hilum and leads toobstruction of the bile ducts. E. alveolariscan therefore not be distinguished initiallyfrom a hilar bile duct carcinoma. There isnonhomogeneous echogenic infiltrationof hepatic tissue.

11–25% and is particularly high in the caseof multiple abscesses (figure 45). Pyrexiaand abdominal pain are almost always pre-sent.Table 8 shows the approach with pyogenicabscesses. In about 10–20% of cases, theorigin of the infection is not found. Thereare several kinds of sonographic appear-ance. At an early phase, no capsule has yetformed and the abscess can be overlooked.The abscesses are usually echo-poor andshow dorsal echo enhancement (figure 44).A capsule with dorsal echo enhancementappear later (figure 45), and the internalechoes are variable. On contrast-enhancedcomputed tomography, colouring of thecapsule is pathognomonic. However, thiseffect can be absent. Abscesses can beshown excellently with echo signal en-hancers. The hyperemic margin can be seenin the capillary or portal venous phase (Bau-ditz et al., 2007).

68

Fig. 44:Liver abscess. There is a hypoechoic space-occupying lesion with a suggestion of dor-sal acoustic enhancement. The origin wassevere infectious colitis. Treatment was withdrainage and antibiotics.

69

Fig. 45:Multiple abscesses in a patient with a bile duct carcinoma. Placement of a stent in thecommon bile duct brought temporary improvement.

70

Table 8:Algorithm for pyogenic abscess.

Therapy

Antibiotics Drainage Operation

Small abscesses a) abscess < 4 cm: when drainage hassingle aspiration and been unsuccessfulirrigation with saline,

antibiotics for three weeks

b) abscess > 4 cm:drain through

pigtailed catheter

With mixed infections and air inclusions,there is a complex picture, and the abscessmay be mistaken for a tumor, especiallywhen the classical clinical signs with pyrexiaare absent.

Younger patients who give a history ofspending time in the tropics often sufferfrom an amebic abscess. Pyrexia and upperabdominal symptoms can occur weeks afterthe return from tropical regions. Diarrheadue to the amebas can have a relativelybland clinical course. Solitary abscesses withhomogeneous contents predominate (figure46). Sedimentation can appear in this casealso; there is dorsal echo enhancement. Ab-scesses which are superinfected have com-plex internal echoes (figure 47).

The diagnosis can be made from the historyand the typical ultrasound appearance. Thehemagglutination test is nearly always posi-tive. Blood cultures typically are sterile. Diag-nostic aspiration (usually not necessary)yields an odourless brownish cocoa-like liq-uid. Treatment is mainly conservative. It cantake months before the abscesses disap-pear. If there is a risk of rupture or vicinity tothe pericardium, drainage or surgical evacu-ation can be necessary. Since there are usu-ally no severe concomitant illnesses in thisgroup of patients, the prognosis is good.

71

72

Fig. 46:Amebic abscess in the right lobe of the liver10 cm in diameter after a sojourn in thetropics. There are fine internal echoeswhich appeared with a change of patientposition.

Fig. 47:No longer fresh amebic abscess 4 weeksafter antibiotic therapy. The patient hadonly become ill with pyrexia 8 weeks afterher return from India. The complex internalstructure and the initial absence of aresponse to metronidazole suggestedbacterial superinfection.

Benign liver tumors

Hemangioma

The commonest benign liver tumor is thecavernous hemangioma. It is found at 1–7%of autopsies. Women are affected five timesmore often. A growth-promoting effect ofestrogens and progesterone is suspected.The ratio of solitary to multiple heman-giomas is 4–9:1. Complications are rare.Calcifications are found in 10% of cases.A hemangioma consists histologically oflacuna-like distended capillaries, throughwhich flow is slow. A reticuloendothelialsystem is absent, so that the hemangiomaremains unenhanced on colloid isotopescintigraphy. With increasing size, there isa tendency to thrombosis, fibrosis, cystic de-generation or calcification.The typical ultrasound appearance is of astrongly echogenic tumor with relativelysmooth margins (snowball). The numerous

impedance leaps of blood and vessel wallsgive rise to strong echoes, which can leadto repeat echoes posterior to the lesion (re-verberations). There is usually an inflow oroutflow vessel directly connected to the he-mangioma (figure 48). The listed character-istics are typical of the capillary heman-gioma, which seldom becomes bigger than3 cm. The sonographic appearance can beextraordinarily varied in the case of heman-giomas greater than 3 cm (figure 49, 51).Most are of the cavernous type with some-times wide vessel lacune. Hypoechoic struc-tures can correspond to a thrombosis or fi-brosis. In case of doubt, further investiga-tion is performed, usually a spiral CT withbiphasic contrast injection (table 9). On con-trast CT, the surrounding parenchyma firstbecomes coloured, while the hemangiomaappears as a gap. Small spots of contrastcan be visible at this early phase in the re-gion of the margin. Because of the slow

73

74

Fig. 48:Typical hemangiomas. Echogenic, relativelysharply demarcated lesions in the rightlobe of the liver. A vessel can be seen inthe upper part of the larger hemangioma.

Fig. 49:Atypical hemangioma 4.2 x 3 cm in diame-ter, with rather few internal echoes. How-ever, on CT the lesion demonstrates atypical iris phenomenon. The echo-poorinterior possibly corresponds to thrombosis.

75

Table 9:Findings and accuracy of various imaging techniques for hemangioma.

Sensitivity Specificity Comment

Ultrasound 80% up to Problems in the event of thrombosis100% and with giant hemangiomas, and

with echo-poor hemangiomas in afatty liver. In nearly 80% the irisphenomenon can be demonstratedwith echo signal enhancers.

CT 85–95% Native: hypodense, after contrast irisphenomenon.

MRI 95% T1: hypointense, T2: native hyperin-tense, T2 with contrast iris phenome-non. MRI pattern little disturbed byfibrosis and thrombosis

blood flow, the hemangioma fills slowlyfrom the margin towards the centre (irisphenomenon), while the contrast agent hasalready flowed out of the rest of the liver(figure 50). However, this behaviour of thecontrast is also seen with metastases from amammary carcinoma.Alternatively, MRI can be performed. Bloodpool scintigraphy has become less impor-tant. On colour Doppler sonography, he-mangiomas under 3 cm diameter do notgive any colour signals in the interior or pe-riphery. If colour signals are obtained and ifa pulsatile signal is found particularly in theinterior of the lesion, a hepatocellular carci-noma or metastasis should be suspected(vide infra). In 3% of small hemangiomas,nevertheless, arterial signals are found inthe interior of the lesion with modern colourDoppler ultrasound equipment (Wachsbergand Jilani, 1999). Contrast-enhanced colourDoppler sonography can also demonstrate

the phases of so-called globular enhance-ment and the iris phenomenon (Wermkeand Gassmann, 1998; Dietrich et al., 2007).The iris phenomenon occurs within 180 secin 78% of cases of biopsied hemangiomas(Dietrich et al., 2007) and is therefore highlyreliable (figure 51 left).Hemangiomas can be aspirated (Brambsand Spamer, 1985; Dietrich et al., 2007). Indoubtful cases, diagnosis can be hastenedin this way and costs saved. As with all le-sions, there should be a rim of normal he-patic tissue between the liver surface andthe presumed hemangioma.Differential diagnosis: echogenic metastases(colon carcinoma, neuroendocrine tumors)(Zimmer et al., 1995), focal nodular hyper-plasia, lipomas, (rarely angiomyolipomas,e.g, in tuberous sclerosis) (Nonomura et al.,1996). In case of doubt, it is very helpful ifcomparison with earlier scans is possible.

76

77

Fig. 50:Contrast behaviour on CT. The native hemangioma is hypo- or isodense (arrow).After injection of contrast, the hemangioma fills from outside inwards (iris phenomenon,arrow, right side), while in the rest of the liver the parenchyma has become decoloured.

78

Fig. 51:On the left is a hemangioma after injection of an echo signal enhancer. The border of thehemangioma shows nodular enhancement (red circle). In the right-hand picture, a typicalhemangioma can be seen with a blood vessel in a congested liver in a patient with globalcardiac failure. (A = ascites).

79

Progress: There is no tendency to malignanttransformation. Growth in size can occurunder hormonal influence. At sizes over15 cm, a consumption coagulopathy is pos-sible. Rupture is very rare and seen onlywith very large hemangiomas. An operationmay be necessary in this situation. Asympto-matic giant hemangiomas are observed.Symptomatic hemangiomas can be operat-ed, and no recurrence has been described.Individual case reports also recommendtransarterial embolization.

Liver cell adenoma

The liver cell adenoma is a genuine epithe-lial tumor which has increased in incidence.It is found in fewer than 0.5% of autopsies.Women are affected nine times as oftenas men. Growth occurs with steroid treat-ment. Men should be asked about anabolicsteroids. It also occurs more frequently in

the glycogenoses. The typical age for thedisorder is 15–45 years. In 30% it is a coin-cidental finding, and in up to 50% it isfound because of abdominal symptoms.There are complications in up to 30% ofcases, such as bleeding into the tumor orrupture, so that an emergency operation isrequired. In 90% of cases, the tumor is soli-tary (Bleck et al., 1997).Histologically, it consists of hepatocytes, por-tal areas and bile ducts, and Kupffer cellsare usually but not always absent. The ade-noma is often surrounded by a pseudocap-sule. The adenoma can be difficult to distin-guish histologically from normal hepatic tis-sue or from a highly differentiated hepato-cellular carcinoma.The ultrasound appearance is variable andnot diagnostic. It is usually echo-poor, butit can also be echogenic, have complexechoes (figure 52) or demarcation from sur-rounding tissue may not be possible. With

echo signal enhancers imaging is most suc-cessful in the early arterial phase. CEUSeven appears to be superior to other meth-ods. Occasionally, only a projection of theliver surface suggests the diagnosis. Since amalignancy cannot be excluded by any formof imaging, biopsy is required. The otherimaging methods are also non-specific. Theappearance is variable, and partly depen-dent on the fat content. Since malignantdegeneration occurs, the current opinion isthat adenomas should be removed surgical-ly. Even histologically, distinction from ahighly differentiated hepatocellular carcino-ma can be difficult (De Carlis et al., 1997).Differential diagnosis: hepatocellular carci-noma, fibrolamellar carcinoma, metastases,focal nodular hyperplasia.

80

Fig. 52:Adenoma in segment IV confirmed bybiopsy. The internal structure has complexechoes and does not allow precise diagno-sis. At the same time, the patient whorefused operation has a cavernomatoustransformation of the portal vein with acongenital abnormality of coagulation.GB = gallbladder

Focal nodular hyperplasia (FNH)

Focal nodular hyperplasia (FNH) is a benignhepatic tumor and is found in 0.1–1% ofautopsies. Women are affected twice asoften. Steroid hormones are not regardedas causes but appear to promote growthso that it has hitherto been considered cor-rect to discontinue them following diagno-sis of FNH. This view has recently been chal-lenged in a large series of 216 women withFNH (Mathieu et al., 2000). In this study thesize and number of FNH was not influencedby use of oral contraceptives; in addition,pregnancy was not associated with FNHchanges or complications. The typical ageis 20–50 years. In 80% there is only onefocus. Ruptures occur much more seldomthan in the case of adenoma. In 45–80% itis a coincidental finding. One third of pa-tients complain of abdominal pain (Kojiro etal., 1996; Bartlett et al., 1996).

All of the liver’s cell types are representedhistologically. However, the bile ducts arerarefied and impaired in function. A fewpathologists regard FNH as a hamartoma inassociation with vascular malformations(Takayasu and Okuda, 1997; Wanless,1990). Others describe FNH as localized cir-rhosis or regenerative nodules. The ultra-sound appearance can be highly varied (fig-ure 53 and 54). Small lesions are homoge-neous, but over 3 cm a stellate scar canusually be detected. Connective tissue septawith arteries are arranged in the form of astar, thus producing the characteristic wheelspoke pattern on arteriography and colourDoppler sonography. The wheel spoke struc-ture can be demonstrated well using echocontrast or in power mode (figure 55). Afterecho signal enhancement, FNH “stains“strongly in the early arterial and portal ve-nous phases (figure 55). In the late phase,FNH often remains more highly stained with

81

82

Fig. 53:FNH nearly 10 cm in diameter confirmedby biopsy. A definite scar star is not visible.

Fig. 54:FNH over 6 cm of the right lobe. A wheelspoke pattern is suggested.

echo signal enhancers than the surroundingtissue, which distinguishes them frommetastases (Wermke and Gassmann, 1998;Wermke, 2006).Until recently, hepatobiliary sequence scintig-raphy was regarded as the gold standardin imaging FNH. Since FNH contains allcell types but the correct architecture ofthe liver lobe is lost, there is delayed up-take and delayed biliary excretion of thetracer (Sciuk and Schober, 1997). If theperfusion, parenchymal and late phases areevaluated adequately in this investigation,cholescintigraphy achieves a sensitivity of90% and a specificity of 100% in differen-tiating it from hepatocellular carcinoma(Gratz and Weimann, 1998). On CT, there isbrief massive enhancement and delayedcontrast equilibrium with the surroundinghepatic tissue. The stellate scar is not alwayspresent (Schild et al., 1987).

83

Fig. 55:FNH approx. 30 sec after injection of anecho signal enhancer. The FNH is alreadytaking it up while the surrounding paren-chymas does not yet appear completelyechogenic.

Ultrasound-guided needle biopsy shortens thediagnostic process. It is usually done to differ-entiate from adenoma. Asymptomatic FNH isobserved. Malignant transformation has notbeen described (Bennett and Bova, 1990).When there is clear evidence of FNH on CEUS,biopsy can be omitted.

Nodular regenerative hyperplasia

Multiple benign lesions in the form of nodu-lar regenerative hyperplasia (NRH) oftenpose a considerable problem in differentialdiagnosis (figure 56). One theory statesthat hyperplastic or hepatocellular lesionsform as after chemical carcinogens. Wanless(1990) considers NRH to be a compensationmechanism after local atrophy of liver tis-sue. Micronodular nodules remain invisibleon ultrasound. However, they become visi-ble when they are the size of the regenera-tive nodules of cirrhosis. NRH is found in

84

Fig. 56:Nodular regenerative hyperplasia in a42 year old patient with an unclarifiedsystemic illness (eosinophilia). Multipleechogenic nodules with a diameter ofup to 1.5 cm are seen throughout theentire liver (confirmed by biopsy).

systemic illness (vasculitis), and during hor-mone and chemotherapy and in myelo- andlymphoproliferative disorders. NRH can beaccompanied by portal hypertension. Theultrasound appearance is variable, and onlybiopsy allows the diagnosis to be made.

Rare benign liver tumors

Lipoma, angiomyolipoma, hamartoma andmany other benign tumors all have nopathognomonic appearance. Lipomas areechogenic and can be mistaken for a he-mangioma.

Malignant tumors

Metastases

Metastases are the commonest cause offocal lesions seen on ultrasound.Table 10 lists the possible questions thatultrasonography can answer when lookingfor and evaluating metastases.Metastases are very variable in appearance.Table 11 and figures 57–60 represent possi-ble imaging forms.Focal metastases over a diameter of 5 to10 mm can usually be identified. However,identification depends on the echogenicity,the site of the lesion, the quality of theequipment and the experience of the exam-iner. In unfavourable sites (metastases insubphrenic area, superior left lobe, extremelateral right lobe), CT is superior. With thenewest generation equipment, metastases5 to 10 mm in size can be detected (See-mann et al. 1998). Only limited conclusions

85

86

Table 10:Questions answered by ultrasound in the case of metastases.

Lesions typical of metastases detected:• Secure diagnosis• Possibility of biopsy• Number, site (segment classification) and size with reference to resection• Lymph node involvement or infiltration of neighbouring organs• Number, site, size and ultrasound appearance with regard to follow-up during treat-ment (reference lesions)

• Recommendation of other necessary investigations (with limited echo quality, smalllesions)

No lesion detected:• How accurately could the liver be examined?• What other investigations are recommended?

87

Nodular micrometastases – invisible on ultrasoundDiffuse infiltration – invisible on ultrasoundDiffuse infiltration with hepatomegalyNodular metastases with the same echogenicity as hepatic tissue:Visible – due to irregularity of liver surface

– displacement or compression of vessels– changes in echogenicity subsequently

Map-like infiltration

Focal appearance: Anechoic with dorsal echo enhancementHypoechoic with and without a rimEchogenic with and without a rimEcho-complex with echo-free or echogenic internal interior(spontaneous or as a result of therapy)Regressive changes (scar, calcification, necrosis)

Table 11:Ultrasound appearance of liver metastases.

about the primary tumor can be drawnfrom the ultrasound appearance of the le-sions. The sensitivity of ultrasound for thedetection of metastases is approx. 60–70%.When highly suspected but ultrasound isnegative, the next useful investigation iscontrast-enhanced ultrasound (if available)or computed tomography (Bidlingsmaier etal., 1999; Fröhlich et al., 1997). Compara-tive studies show that a sensitivity of over90% can be achieved with modern high-end devices. Computed tomography can beregarded as complementary as differentfocal lesions can be shown in both. In CEUSmetastases “wash out“ in the late phase asthey are not hepatic tissue and thus have nosinusoids and therefore they cannot “store“the echo signal enhancer (figure 58). This“washout“ in the portal venous phase re-sults in a sensitivity of 100% for malignancy(Mörk et al., 2007). In a multicenter study,basic ultrasound had a sensitivity of 84.6%

and specificity of 78%. The values were in-creased by CEUS to 88.5% and 94% re-spectively compared to contrast-supportedspiral computed tomography of 92.3% and89.2% respectively (Dietrich et al., 2006).However, caution is warranted in the caseof metastases of neuroendocrine tumors.They have very high arterial perfusion butcan also acquire the characteristics of he-mangiomas.Metastases can also be monitored ultra-sonographically. Early metastases tend to beecho-poor and are therefore easily identifi-able only in a fatty liver. In the case of lym-phomas or melanoma, the metastases canbe mistaken for cysts (figure 57).

88

89

Fig. 57:Appearance of metastases. Early metastases appear echo-poor or anechoic, and theycan demonstrate dorsal echo enhancement (malignant melanoma, diameter < 1 cm, 1).Somewhat larger metastases tend to be echo-poor, and a rim is absent in this case (breastcancer, 2). Typical echo-poor rim, target type (small cell bronchial carcinoma, 1.5 cmdiameter, 3). Echogenic metastasis, typical of gastrointestinal primary tumors, can bemistaken for a hemangioma (here colorectal carcinoma, 4). Larger echogenic lesion of acolorectal carcinoma, echo-poor rim, echogenic ring, echo-poor interior, “bull’s eye” (5).

The detection of an echo-poor rim arounda slightly more echogenic centre (target)has a sensitivity and specificity for the pres-ence of metastases of 84 to 90%. The rimis thus highly suspicious but no proof. Itis not yet clear whether this is tumor tissueor a reaction by the area surrounding themetastasis.Very small metastases not detectable ultra-sonographically are found particularly oftenwith small cell bronchial carcinoma (Schöl-merich et al., 1984 and 1987). Laparoscopyshould be considered when therapeuticallyindicated. Breast cancer and melanoma alsooften have small hypoechoic or ultrasono-graphically undetectable metastases. Thediagnosis becomes easier as the disease pro-gresses. Figure 59 shows diffuse intersper-sion with small hypoechoic metastases (someof target type) in malignant melanoma.

90

Fig. 58:Metastases with a punched-out appearancein the late phase of CEUS.

In the clinical course metastases may be-come confluent and will show a map-typephenotype (figure 59).

Ultrasound is used widely today to monitortumor disease. There are two situations:monitoring in patients initially free frommetastases, and monitoring during treat-ment. Depending on the indication, a ratio-nal examination interval and actual thera-peutic consequences in discovering focallesions should be ensured. The high psycho-logical stress on patients of excessively fre-quent follow-up examinations should beconsidered along with the possibility of anunsatisfactory result when repeat examina-tions are performed too frequently duringtherapy. In monitoring progress, planimetryor even volumetry of space-occupying le-sions is superior to measuring the transverseand longitudinal diameters. On the otherhand, this is more time consuming andthe accuracy of reproducibility of the samemeasurement direction is questionable. In-vestigations have shown that progression orregression can be reported only when the

91

Fig. 59:Confluent metastases of map type,also in malignant melanoma.

change is greater than 30% (Schölmerichand Gross 1997). Measurements using spi-ral CT are more precise, but the method ismore complex.Not only the size but also the appearance ofthe metastases can alter spontaneously orduring treatment. As the diameter increas-es, areas of spontaneous necrosis develop.This can be regarded as success duringchemotherapy, and necrosis occurs evenwith small diameter lesions. Metastases candisappear completely on treatment or canundergo fibrous or calcified degeneration(figure 60).If primary resection is possible, exact diag-nosis of the number and site of the metas-tases is necessary. CT arterioportography(CTAP) can be used preoperatively. An arte-rial catheter is inserted and the behaviourof the contrast agent during the arterial(through the hepatic or superior mesentericartery) or portal venous phase (through the

splenic artery) is studied by computed to-mography. The disadvantage of the methodis its invasiveness and the necessity of usingcontrast. However, the gold standard is in-traoperative ultrasound. The operative planstill has to be changed or the operationhalted in up to 40% of cases nowadays be-cause of the intraoperative detection of fur-ther metastases or other sites. Contrast-en-hanced MRI is being improved constantly,but the selection of examination parametersand interpretation of the findings is consid-erably more complex.

92

93

Fig. 60:Progress of echogenic metastases in colon carcinoma. (1) Large echogenic lesion initially. (2)Necrotic areas appear subsequently during chemotherapy. (3) Regressively altered areas withcalcification after chemotherapy for colorectal carcinoma.

Lymphomas

Lymphomas of the liver can produce dif-fusely infiltrating and focal changes in theliver (figure 61 and 62). The ultrasound ap-pearance is not pathognomonic. Focal le-sions are usually echo-poor or almost ane-choic, so that it is possible to mistake themfor cysts. Extrahepatic manifestations oflymphoma (splenomegaly, mesenteric lym-phomas) support the suspected diagnosis oflymphomatous infiltration of the liver, butonly histology is confirmative.

94

Fig. 61:Focal manifestation of a high-grade non-Hodgkin’s lymphoma of the liver. The focuscannot be distinguished from a metastasisfrom an epithelial tumor.

Primary malignant liver tumors:

Hepatocellular carcinoma (HCC)

The hepatocellular carcinoma is the com-monest malignant tumor worldwide, due tothe high prevalence of hepatitis B in coun-tries of the Third World. In Germany, theincidence is comparatively lower. Cirrhosisis present in 80% of cases. The annual inci-dence in cirrhosis is 3–6%. In particular, pa-tients with hepatitis B and C, alcoholic cir-rhosis and genetic hemochromatosis are athigh risk and require close screening, con-sisting of liver ultrasound and measurementof the alpha-1-fetoprotein every 6 months.It is still unclear whether this improves theprognosis or whether the tumor is only dis-covered sooner.

95

Fig. 62:B-cell non-Hodgkin’s lymphoma. Apartfrom a lesion typical of metastasis, thereare further echo-poor lesions and oneechogenic focus.

The manifestation of a HCC can be diffuse,focal or pedunculated. The ultrasound ap-pearance is very variable, and several echoqualities, i.e. echo-poor and echo-rich areas,can occur in a tumor nodule (mosaic pat-tern, figure 63; Takayasu and Okuda, 1997).A connective tissue capsule is believed to in-dicate a better prognosis.On colour Doppler sonography, lesions lessthan 3 cm give signals at the periphery (bas-ket sign) or in the interior (tumor vessel,spot sign, figure 64) in up to 80% of cases.The detection of internal flow signals increas-es in power mode (Lin et al. 1997). If thesehave a pulsatile spectrum, a hepatocellularcarcinoma is present with a high degree ofprobability, especially in the presence of cir-rhosis, when metastases are rare. A centralcontinuous spectrum is believed to indicateregenerative nodules or adenomatous hy-perplasia. On CEUS 90% of HCCs showring enhancement, but only 30% of

96

Fig. 63:Large HCC in a patient with cirrhosis afterhepatitis B. The inside of the space-occupy-ing lesion is echo-complex (mosaic pattern)and thus typical of a HCC.

dysplastic regenerative nodules. 100% ofHCCs but only 10% of regenerative nod-ules demonstrate chaotic perfusion andcentripetal uptake of echo signal enhancers(Rickes et al., 2002). Washout in theparenchymal phase can be observed withecho signal enhancers as a criterion ofmalignancy, and this is more marked withLevovist® than with Sonovue® (von Herbayet al., 2007). However, in the individualcase, distinguishing a HCC from a regenera-tive nodule can be very difficult even histo-logically. In general, biopsy is recommendedin the case of doubt unless transplantationis planned anyway. Whether a biopsy canthen be omitted must be agreed with therespective center.HCC can be divided into stages togetherwith clinical parameters, signs of portal hy-pertension and taking liver function into ac-count. There are different staging systems,and the Barcelona Clinic Liver Cancer stag-

97

Fig. 64:Diagram of the findings which can bemade with colour Doppler ultrasound(see text).

Basket sign

Tumor vessel

“spot”

ing system (BCLC) comes close to clinical re-quirements. It takes into account liver func-tion, portal hypertension, the extent of theHCC and any vascular invasion. Therapy al-location is according to stage and extendsfrom resection and transplantation in earlystages to palliative procedures such asethanol injection (PEI) or radiofrequencythermoablation (RFTA) (Schacherer et al.,2007).The CEUS does not show a clear picture.Distinction from regenerative nodules isdifficult. Table 12 lists comparative studies.With the current spiral CT a sensitivity of54–70% and specificity of 70–93% can beachieved. For MRI, the sensitivity is between57 and 85% and the specificity is between21 and 82% (Schacherer et al., 2007).An increased alpha-1-fetoprotein of morethan 500 ng/ml and the finding of a corre-sponding lesion on ultrasound are regardedtoday as adequate proof of a HCC. Typical

signs also include tumor thrombosis of theportal vein and its branches. If arterial ves-sels are found in the thrombus, malignancyis confirmed (figure 65).

98

Fig. 65:Tumor thrombosis of the portal vein in al-coholic cirrhosis. The right lobe of the liveris diffusely infiltrated and there is perihep-atic ascites.

In contrast to earlier pessimism, a range ofpossible treatments is available today, whichmake a differentiated procedure necessary(see algorithm). The aim is the earliest possi-ble diagnosis of the HCC by the screeningof high-risk groups described above (fig-ure 71). Table 12 shows the usefulness ofdifferent imaging methods. Ultimately, nomethod is capable of distinguishing be-tween a hyperplastic regenerative noduleand a small HCC. On contrast-enhancedspiral CT, HCC can be seen particularly inthe early arterial phase. The value of Lipi-odol® CT is controversial. The selectiveaccumulation of Lipiodol® in the HCC isutilised also in transarterial chemoemboliza-tion (TACE), when Lipiodol® can be mixedwith chemotherapeutic agents.

The diagnostic and therapeutic procedurealso depends on hepatic function. This istaken into account in Okuda’s classification(Takayasu and Okuda, 1997).Ultrasound-guided diagnostic and thera-peutic aspiration is used in palliative proce-dures (see therapeutic interventions). CTshows the relationship to adjacent organsand the size of the lesion better (figure 66).

99

100

US CT Lipiodol®- MRI DSA Intra-CT operative

US

Nagasue et al. (1989) 91 92 93 98

Choi et al. (1991) 73 82 96

Takayasu et al. (1990)* 84 84 93 81 96

Shibata et al. (1991) 84

De Santis et al. (1992)* 67 50 93 63 74

Table 12:Sensitivity (%) of imaging of hepatocellular carcinoma. *Small carcinomas. (US = ultrasound,CT = computed tomography, MRI = magnetic resonance imaging, DSA = digital subtractionangiography).

101

Fig. 66:Large HCC in segment III with cirrhosis due to hepatitis C. Left, CT, right, ultrasound.The patient recovered well from resection of this initially solitary lesion. However, severalnew foci of HCC appeared in the remaining liver eight months later.

Cholangiocellular carcinoma (CCC)

These rare tumors are divided into peripher-al and hilar types (Klatskin tumors). The pe-ripheral carcinomas cannot be distinguishedultrasonographically from other primary orsecondary carcinomas (figure 67). Centralhilar tumors often remain invisible onultrasound and are apparent only indirect-ly through congestion of the bile ducts (fig-ure 68). Again it is CEUS that can show thetumor spaces in the late phase in the caseof Klatskin tumors (Bauditz et al., 2007).ERCP is performed today for diagnosis andplacement of a palliative stent. WhetherMRCP (magnetic resonance cholangiopan-creatography) will improve early diagnosisremains to be investigated.In the past, the thorium used as the X-raycontrast agent Thorotrast® represented animportant risk factor for a CCC. It has notbeen used since 1950. The last patient in

Freiburg with Thorotrast®-induced CCCdied in 1989.A risk group for the development of a CCCtoday is represented particularly by patientswith primary sclerosing cholangitis. There isno useful screening test. The prognosis isgood only when the CCC is discovered asan incidental finding during transplantation.Echinococcus alveolaris is an important dif-ferential diagnosis in CCC, which can alsoshow as diffuse hilar infiltration with block-age of the bile ducts (cf. figure 43).

Rare primary tumors

Hemangioendotheliosarcoma, angiosarco-ma, fibrolamellar carcinoma and hepatomain childhood do not have a typical ultrasoundappearance and can only be confirmed byhistology.

102

103

Fig. 67:CCC of peripheral type. Locally blockedbile ducts and ascites (A) can be seen inaddition to echogenic, poorly demarcatedareas of tumor (T).

Fig. 68:CCC of central type (Klatskin tumor). Thebile ducts are clearly distended. A diffuselyinfiltrating process is suspected in the areaof the hilum.

The following pages contain suggestions forthe procedure in certain situations usingtables and algorithms.

Diffuse parenchymal disorders

Fatty liver, cardiac hepatic congestion andcirrhosis can be diagnosed well today byultrasound, if several signs coincide and theexaminer has the appropriate experience.Beginners often diagnose “diffuse parenchy-mal damage” which is associated withpathological liver tests in no more than 30%of cases (Ochs et al., 1994). Using colourDoppler sonography, portal vein thrombo-sis, Budd-Chiari syndrome and portal hyper-tension can also be diagnosed (Haag et al.,1999).However, the gold standard is still biopsyand laparoscopy. Biopsy will always be at-tempted in the case of treatable causes of

liver disease (hepatitis C, Wilson’s disease,hemochromatosis, alpha-1-antitrypsin defi-ciency, autoimmune hepatitis, primary bil-iary cirrhosis etc.). In ascites or severe coag-ulation disorders, transjugular biopsy canalso be attempted.

Focal liver lesions

Innumerable algorithms have been pro-posed to investigate focal lesions of the liver(Brambs, 1996; Feuerbach et al., 1997;Fröhlich et al., 1997; Gerok, 1988; Helm-berger et al., 1998; Linhart et al., 1998; Liseet al., 1996; Lock et al., 1997; Rettenmaier,1988; Strohm and Weimer, 1997; Weimannet al., 1997; Weiss, 1991). The complexityof the possible differential diagnoses con-trasts with the frequency distribution of thefindings.

104

The status of ultrasonography in diffuse and focal disorders

Metastases account for up to 70% of cases,which appear as such and can be confirmedrapidly by needle aspiration, if a more ac-cessible tumor site is not available. Table 13gives a synopsis of current imaging meth-ods with the typical findings and their accu-racy.As discussed, two situations arise:• the procedure with an incidentally detect-ed lesion

• the procedure when tumor is suspected

Before working through each of these algo-rithms, an attempt should be made to ob-tain a second opinion from an experiencedexaminer (Fröhlich et al., 1997).If a focal lesion is found on ultrasound andclear classification is not possible, biopsy isan economical way of shortening furtherprocedure. No other imaging method willgive a histological diagnosis.

Figure 69 shows an algorithm for the inci-dental finding of a focal lesion. Five ultra-sound diagnoses can be regarded as defi-nite, if they meet classic criteria:• cysts• hemangioma• mild fatty infiltration• classical Echinococcus cysticus• metastases

With CEUS, FNH and adenoma can be diag-nosed highly visibly along with metastases.

Figure 70 shows the procedure when look-ing for hepatic lesions. These are usuallypatients with a known or highly suspectedmalignancy on initial investigation or follow-up. This algorithm differs from figure 69in that unclear findings are biopsied muchsooner and a hemangioma or cyst is keptunder observation only with quite unequiv-ocal evidence.

105

Focal lesion Gold standard Ultrasound Colour Doppler Spiral CT(fundamental) ultrasound (CEUS)

Hemangioma Ultrasound Echogenic, Iris phenomenon Iris phenomenonin typical feeding vessel and nodular and nodularhemangiomas enhancement enhancement

Cysts Ultrasound Anechoic, No additional Hypodenseposterior acoustic informationenhancement

Adenoma Biopsy, CEUS Hypoechoic or Readily demonstrable Hypodense withslightly more early arterial hyperdense areasechogenic than enhancement of bleedingthe surroundings,uncharacteristic

Focal nodular CEUS, CT, MRI, Sharp margins, Wheel spoke Hypodense withhyperplasia (biopsy) usually hypoechoic, appearance of the centrifugal enhance-

wheel spoke structure vessels (with or ment (“blush“)sometimes visible without CEUS)

Metastases Spiral CT with Hypoechoic halo CEUS: washout in Hypodense,contrast, CEUS, typical but not the late phase nonhomogeneousMRI, biopsy obligatory, due to necrosis or

hypoechoic, scarring or bleedingechogenic

Hepato- Biopsy? AFP and Signs of cirrhosis in Evidence of tumor Early arterial margincellular detection of 80%, hypoechoic thrombi, arterial enhancement,carcinoma focal lesion in and echogenic, fistulas, pathological irregularly hypodense,

cirrhosis (refer “mosaic pattern“, arterial flow profile (Lipiodol® storagethere, contrast tumor thrombus 14 days afterCT, MRI, CEUS) administration)

106

Table 13:Different focal lesions and their typical appearance in different imaging procedures(modified from Bidlingsmaier et al., 1999 and Fröhlich et al., 1997). Nuclear medicineand angiographic methods have lost much of their importance.

107

Focal lesion Magnetic resonance Nuclear medicine Angiographyimaging

Hemangioma T1: hypointense Blood pool scintigraphy Contrast poolingT2: hyperintense has become less important ”cotton wool“ sign

Cysts T1: hypointense No importance No importanceT2: hyperintense

Adenoma T1: hyperintense Sulfur colloid: Hypervascular,T2: iso-hypointense, 80% defect, non-specificnon-specific 20% staining,

has become less important

Focal nodular T1: isointense Hepatobida scan: Hypervascular,hyperplasia T2: iso-hyperintense, delayed staining, wheel spoke

centrally hyperintense delayed biliary outflow structure

Metastases T1: usually hypointense Somatostatin receptor Hypovascular(hemorrhage, necrosis) scintigraphy in GEP-NET. apart fromT2: usually hyperintense Positron emission GEP-NET(hemorrhage, necrosis) tomography

Hepato- T1: usually hypointense Similar staining as in Hypervascular,cellular (hemorrhage, necrosis) FNH possible on pathological arteries,carcinoma T2: usually hyperintense hepatobida scan, arterioportal or

(hemorrhage, necrosis) arteriovenous shunts arteriovenous fistulas

CEUS = contrast enhanced ultrasound, GEP-NET = gastroenteral-pancreatic neuroendocrinetumors, MRI = magnetic resonance imaging, grey type: low clinical significance

108

Focal lesion as incidental ultrasound finding

Cyst

appearance

Monitoring or treatment

yes

no

Suspectedmetastasis?Adenoma?

HCC, CCC?

yes

Search for primary tumorBiopsy

Benignappearanceyes

nounclear

Hemangioma

BenignFNH, abscess Spiral CT with

(MRI)contrast

CEUS

Fig. 69:Procedure for the incidental finding of a focal lesion. HCC = hepatocellular carcinoma,CCC = cholangiocellular carcinoma.

109

Fig. 70:Procedure when searching for a hepatic focus, usually in patients with a known primarytumor. This also allows for the situation when no hepatic foci are shown despite high clini-cal suspicion. A biopsy should always be performed always if a better extrahepatic biopsysite does not offer itself.

Cyst, hemangiomaFNH, abscess

Benign appearanceMonitoring or treatment

no

Suspected metastasis?Adenoma?

HCC, CCC?

yes

Search for primary tumorBiopsy

CEUSSpiral CT with contrast

(MRI)

Benignappearance

no

Focallesion

Search for a lesion

yes

no

yesyes

110

A particular situation arises when looking fora HCC. Cirrhosis is usually present, so thatthe important differential diagnosis is regen-erative nodules. The therapeutic procedurewill depend on the extent of hepatocellularinsufficiency and tumor size (figure 71).Monitoring of cirrhosis patients is also in-cluded (Ruelas-Villavicencio and Vargas-Vorácková, 2004).

Diagnostic approach in cholestasis

No algorithm is given for this situation. Dis-tended bile ducts can usually be detectedeasily on ultrasound, but detection of thecause of the blockage and its precise siteis not always successful. The sensitivity of ul-trasound for choledocholithiasis is thus nogreater than 70%. It can be supplementedby endosonography, ERCP or MRCP whenthere are grounds for suspecting me-chanical cholestasis. ERCP and transcuta-

neous transhepatic bile duct puncture anddrainage are available for palliative treat-ment. The diagnosis of Klatskin tumors ismade possible by CEUS. In the late phasethe extent of the tumor infiltration sur-rounding the bile ducts can be demonstrat-ed well (Bauditz et al., 2007).

Ultrasound-guided diagnostic andtherapeutic procedures

Diagnostic aspiration

Hepatic lesions over 1–2 cm in size can usu-ally be aspirated under sonographic guid-ance. Aspiration of a liver lesion is indicatedwhen knowledge of its significance willhave consequences for therapy. Ultrasound-guided fine needle aspiration allows tissueto be obtained with little burden for the pa-tient and usually saves further investigation.Cytological opinion only is possible with

111

Fig. 71:Procedure with hepatic lesions suspicious for HCC.

Focallesion

Ultrasound and AFPevery 6 monthsChild-Pugh A

yes

no

AFP normal US, AFP every 6 months

AFP raised Spiral CT US, AFP every 6 months

Lesion < 1 cm US every 3 months

No HCC US, AFP every 6 months

HCC

Lesion > 1 cm

HCCLesion > 2 cm AFP > 400 ng/ml, CT, MRI

Lesion 2 cm Biopsy HCC

very fine needles. Tissue cores obtainedwith cutting biopsy needles (e.g. Truecut®)can be assessed histologically. Contraindica-tions must be noted (table 14). Figure 72shows the various possible aspiration meth-ods. In Freiburg, ultrasound-guided aspira-tion through a perforated curved arraytransducer is preferred.Serious complications after ultrasound-guid-ed needle biopsy of hepatic lesions are rare.The mortality is 0.08%. Complications suchas bleeding, bile peritonitis, dissemination oftumor cells, hematoma or peritonitis ofother origin was described in 0.5% of allaspirations. The overall rate, including minorcomplications, is 5% (Spamer et al., 1986).Dissemination of tumor cells along theneedle track occurs very rarely (0.05%) anddoes not lead to decreased life expectancyof the patient. Echinococcus should beexcluded serologically before aspiration, asanaphylactic reactions are possible. How-

ever, the risk of such a reaction is overesti-mated. In many cases, ultrasound-guidedPAIR is also used today as treatment (referthere).

Ultrasound-guided therapy

Aspiration of ascites and pleural effusion

Even small volumes of ascites or pleural ef-fusions can be aspirated diagnostically afterultrasound localization.

112

Severe coagulation disorders(Echinicoccus cysticus)

Ascites (relative)

Table 14:Contraindications to ultrasound-guidedaspiration of focal hepatic lesions.

113

Fig. 72:Ultrasound-guided biopsy. Left, needle aspiration through a perforated transducer.The probable route of aspiration is shown on the monitor by guidelines (A), which allowsexact adjustment of the focus (B). On the right, the needle is guided “freehand” in thecase of large lesions (C).

Adjusting the lesion Biopsy with cutting needle Freehand needle aspirationof a large lesion

114

Percutaneous cholangiography/drainage

If percutaneous transhepatic cholangiogra-phy and drainage (PTCD) proves necessary,puncture especially of the left bile duct sys-tem can be undertaken under ultrasoundcontrol. Imaging with contrast and the in-sertion of guidewires then takes place incombination with radiology.

Abscess and cyst drainage, sclerotherapy

Figure 73 shows the procedure of cyst scle-rotherapy. Symptomatic cysts can be scle-rosed with ethoxysclerol, 10% saline or 95%ethanol. Ethoxysclerol has the advantage ofan additional local anesthetic effect. A sin-gle aspiration will not be successful, sincethese are genuine cysts with a secreting ep-ithelial covering. The cyst is first aspiratedunder local anesthesia and ultrasound con-trol. The cyst contents can be withdrawn for

diagnostic purposes. Using the Seldingertechnique, a guide wire is inserted. This isfollowed by dilatation and the insertion of a5–7F pigtailed catheter. The cyst is drainedfor at least 24 h and during this time 1000to 2000 ml of fluid can be secreted. Thecyst wall is then sclerosed by one of thesubstances listed above with approx. 10%of the cyst contents. Permanent successis achieved in 70%, and a 50% reductionis size is usually obtained in the remaining30%. Echinococcus cysts are treated simi-larly (PAIR).Larger abscesses can also be drained by thesame technique. Often the reduction in sizeby short-term drainage is enough to allowrapid healing with antibiotic therapy.

115

Fig. 73:Procedure of cyst sclerotherapy. Ultrasound-guided aspiration (A) of the cyst and insertionof a pigtailed catheter (B). Drainage of the cyst for at least 24h. The cyst wall is thensclerosed by a concentrated solution (see text). After 8 weeks, cyst volume is markedlyreduced.

US-guided puncture Pigtailed catheter in place Status 8 weeks after therapy

116

Percutaneous therapy of primarytumors and metastases

Primary and secondary malignancies of theliver are being treated increasingly underultrasound control. The size and number ofthe lesions are important for the success oftherapy. Lesions up to 3 cm in size can betreated by targeting them with percuta-neous ethanol injection (PEI) (Allgaier et al.,1996). 3–4 sessions are usually needed. Ra-diofrequency thermoablation (RFTA) repre-sents another method. A needle is insertedinto the tumor under ultrasound control, atthe tip of which curved metal wires are ex-truded in which local heat is produced bya high frequency generator. As a rule, onlyone session is necessary. Lesions up to over4 cm can be treated this way. The complica-tion rate is higher than with PEI. A life-pro-longing effect has not been confirmed hith-erto.

Guidance of TIPS

Placing a transjugular intrahepatic portosys-temic stent shunt without ultrasound guid-ance when the anatomy is complicated canno longer be imagined today. The aspiratingneedle can be guided from the hepatic veininto the portal vein. In particular, the anteri-or or posterior direction of aspiration, whichis not visible on fluoroscopy, can be predict-ed by ultrasound.

Ahima RS.Insulin resistance: cause or consequence ofnonalcoholic steatohepatitis?Gastroenterology. 2007;132:444–6.

Allgaier HP, Blum U, Deibert P, Spangenberg HC,Blum HE.Diagnostik des hepatozellulären Karzinoms.Schweiz Med Wochenschr. 1996;126:1975–83.

Altmann HW.Hepatic neoformations.Pathol Res Pract. 1994;190:513–77.

Bartlett DL, Schwartz LH, Klimstra D, Blumgart LH.Diagnosis and treatment of focal nodularhyperplasia.J Hepatobiliary Pancreat Surg. 1996;3:111–21.

Bauditz J, Schade T, Wermke W.Sonographische Diagnostik des hilären cholangio-zellulären Karzinoms mittels Echosignalverstärkern.Ultraschall Med. 2007;28:161–7.

Bennett WF, Bova JG.Review of hepatic imaging and a problem-orientedapproach to liver masses.Hepatology. 1990;12:761–75.

Bidlingsmaier J, Barkhausen J, Müller RD.Moderne Diagnostik fokaler Leberläsionen:Methoden, Indikationen und Befundmuster.Leber Magen Darm. 1999;29:11–7.

Bleck JS, Wagner S, Manns MP.Nicht-chirurgische Therapie benigner Lebertumore.Internist (Berl). 1997;38: 937–43.

Brambs HJ.Raumforderungen in der Leber – rationale Stufen-diagnostik.Schweiz Rundsch Med Prax. 1996;85:1518–23.

Brambs HJ, Spamer C.Diagnostik von Leberhämangiomen mittels sono-graphisch gezielter Punktion.Schweiz Rundsch Med Prax. 1985;74:1113–4.

Brancatelli G, Vilgrain V, Federle MP,Hakime A, Lagalla R, Iannaccone R, et al. Budd-Chiari syndrome: spectrum of imaging findings.AJR Am J Roentgenol. 2007;188:W168–76.

Braun B, Rückel E, Rückel A, Börner N, Pfeiffer J,Lang H.Sonographische Diagnose eines Budd-Chiari-Syndroms.Dtsch Med Wochenschr. 1983;108:700–2.

117

References

Braun B, Börner N, Majdandzic J,Dormeyer HH, Reuss J, Schild H.Diagnostik der kavernomatösen Pfortadertrans-formation.Z Gastroenterol. 1984;22:244–9.

Cardi M, Muttillo IA, Amadori L, Petroni R,Mingazzini P, Barillari P, et al.Superiority of laparoscopy compared to ultrasonog-raphy in diagnosis of widespread liver diseases.Dig Dis Sci. 1997;42:546–8.

Caremani M, Benci A, Maestrini R, Rossi G,Menchetti D.Abdominal cystic hydatid disease (CHD): classificati-on of sonographic appearance and response totreatment.J Clin Ultrasound. 1996;24:491–500.

Caselitz M, Bahr MJ, Bleck JS, Chavan A, Manns MP,Wagner S, et al.Sonographic criteria for the diagnosis of hepaticinvolment in hereditary hemorrhagic telangiectasia(HHT).Hepatology. 2003;37:1139–46.

Choi BI, Han JK, Song IS, Kim CW, Han MC,Kim ST, et al.Intraoperative sonography of hepatocellularcarcinoma: detection of lesions and validity insurgical resection.Gastrointest Radiol. 1991;16:329–33.

De Carlis L, Pirotta V, Rondinara GF, Sansalone CV,Colella G, Maione G, et al.Hepatic adenoma and focal nodular hyperplasia:diagnosis and criteria for treatment.Liver Transpl Surg. 1997;3:160–5.

De Santis M, Romagnoli R, Cristani A, Cioni G,Casolo A, Vici FF, et al.MRI of small hepatocellular carcinoma: comparisonwith US, CT, DSA, and Lipiodol-CT.J Comput Assist Tomogr. 1992;16:189–97.

Di Lelio A, Cestari C, Lomazzi A, Beretta L.Cirrhosis: diagnosis with sonographic study of theliver surface.Radiology. 1989;172:389–92.

118

Di Matteo G, Bove A, Chiarini S, Capuano LG,De Antoni E, Lanzi G, et al.Hepatic echinococcus disease: our experience over22 years.Hepatogastroenterology. 1996; 43:1562–5.

Dietrich CF, Lee JH, Herrmann G, Teuber G, Roth WK,Caspary WF, et al.Enlargement of perihepatic lymph nodes in relationto liver histology and viremia in patients withchronic hepatitis C.Hepatology. 1997;26:467–72.

Dietrich CF, Wehrmann T, Zeuzem S, Braden B,Caspary WF, Lembcke B.Analyse des Echomusters der Leber bei chronischerHepatitis C.Ultraschall Med. 1998;19:9–14.

Dietrich CF, Kratzer W, Strobe D, Danse E, Fessl R,Bunk A, et al.Assessment of metastatic liver disease in patientswith primary extrahepatic tumors by contrast-enhanced sonography versus CT and MRI.World J Gastroenterol. 2006;12:1699–705.

Dietrich CF, Mertens JC, Braden B, Schuessler G,Ott M, Ignee A.Contrast-enhanced ultrasound of histologicallyproven liver hemangiomas.Hepatology. 2007;45:1139–45.

Feuerbach S, Geissler A, Seitz J, Gmeinwieser J.Radiologische Diagnostik von Raumforderungen derLeber. Computertomographie, Magnetresonanz-tomographie, Angiographie.Internist (Berl). 1997;38:908–16.

Forsberg L, Florén CH, Hederström E, Prytz H.Ultrasound examination in diffuse liverdisease. Clinical significance of enlarged lymphnodes in the hepato-duodenal ligament.Acta Radiol. 1987;28:281–4.

Frank W.Die Darstellung der Lebersegmente im Ultraschall.Eine praktische Anleitung.Radiologe. 1992;32:189–97.

Friedrich-Rust M, Zeuzem S.Nicht invasive Beurteilung des Leberfibrosestadiumsmit der transienten Elastographie (Fibroscan):Aktueller Stand und Perspektiven.Z Gastroenterol. 2007;45:387–94.

119

Fröhlich E, Gerlach A, Gölkel H, Strunk H.Fokaler Leberherd - Stufendiagnostik heute.Leber Magen Darm. 1997;27:209–16.

Gaiani S, Gramantieri L, Venturoli N, Piscaglia F,Siringo S, D'Errico A, et al.What is the criterion for differentiating chronic he-patitis from compensated cirrhosis? A prospectivestudy comparing ultrasonography and percutaneousliver biopsy.J Hepatol. 1997;27:979–85.

Gerok W.Diagnostik der umschriebenen Leberveränderungen.Synopsis: diagnostische Sequenz.Z Gastroenterol. 1988;23:31–6.

Giorgio A, Amoroso P, Lettieri G, Fico P,de Stefano G, Finelli L, et al.Cirrhosis: value of caudate to right lobe ratio indiagnosis with US.Radiology. 1986;161:443–5.

Gosink BB, Leymaster CE.Ultrasonic determination of hepatomegaly.J Clin Ultrasound. 1981;9:37–44.

Gratz KF, Weimann A.Diagnostik von Lebertumoren – Wann ist die Szinti-graphie von Bedeutung?Zentralbl Chir. 1998;123:111–8.

Haag K, Rössle M, Ochs A, Huber M, Siegerstetter V,Olschewski M, et al.Correlation of duplex sonography findings andportal pressure in 375 patients with portalhypertension.Am J Roentgenol. 1999;172:631–5.

Helmberger T, Holzknecht N, Gregor M, Gauger J,Helmberger R, Reiser M.Fokale Lebererkrankungen.Radiologe. 1998,38:263–9.

Joseph AE, Saverymuttu SH, al-Sam S, Cook MG,Maxwell JD.Comparison of liver histology with ultrasonographyin assessing diffuse parenchymal liver disease.Clin Radiol. 1991;43:26–31.

Kojiro M, Okudaira S, Nakashima O.Clinicopathologic characteristics of focal nodular hy-perplasia (FNH) of the liver.J Hepatobiliary Pancreat Surg. 1996;3:106–10.

120

Krause T, Hauenstein K, Studier-Fischer B,Schuemichen C, Moser E.Improved evaluation of technetium-99m-red bloodcell SPECT in hemangioma of the liver.J Nucl Med. 1993;34:375–80.

Lafortune M, Matricardi L, Denys A, Favret M,Déry R, Pomier-Layrargues G.Segment 4 (the quadrate lobe): a barometer ofcirrhotic liver disease at US.Radiology. 1998;206:157–60.

Lassau N, Leclère J, Auperin A, Bourhis JH,Hartmann O, Valteau-Couanet D, et al.Hepatic veno-occlusive disease after myeloablativetreatment and bone marrow transplantation: valueof gray-scale and Doppler US in 100 patients.Radiology. 1997;204:545–52.

Lewall DB.Hydatid disease: biology, pathology, imaging andclassification.Clin Radiol. 1998;53:863–74.

Lin ZY, Wang LY, Wang JH, Lu SN, Chen SC,Chuang WL, et al.Clinical utility of color Doppler sonography in thedifferentiation of hepatocellular carcinoma frommetastases and hemangioma.J Ultrasound Med. 1997;16:51–8.

Linhart P, Bönhof JA, Baqué PE, Pering C.Sonographie bei der Diagnostik benigner undmaligner Lebertumoren.Zentralbl Chir. 1998;123:119–23.

Lise M, Da Pian PP, Nitti D, Miotto D, Balduino M,Bacchetti S.Benign liver tumors in adults: diagnosis andmanagement.J Hepatobiliary Pancreat Surg. 1996;3:89–97.

Lock G, Schölmerich J.Vorgehen bei sonographisch nachgewiesenerRaumforderung der Leber.Internist (Berl). 1997;38:924–7.

Mathieu D, Kobeiter H, Maison P, Rahmouni A,Cherqui D, Zafrani ES, et al.Oral contraceptive use and focal nodular hyperplasiaof the liver.Gastroenterology. 2000;118:560–4.

121

McManus DP, Zhang W, Li J, Bartley PB.Echinococcosis.Lancet. 2003;362:1295-304.

Menon KV , Shah V, Kamath PS.The Budd-Chiari syndrome.N Engl J Med. 2004:350:578–85.

Mörk H, Ignee A, Schuessler G, Ott M, Dietrich CF.Analysis of neuroendocrine tumour metastases inthe liver using contrast enhanced ultrasonography.Scand J Gastroenterol. 2007;42:652–62.

Moriyasu F, Nishida O, Ban N, Nakamura T, Sakai M,Miyake T, et al.“Congestion index“ of the portal vein.Am J Roentgenol. 1986;146:735–9.

Nagasue N, Kohno H, Chang YC, Galizia G,Hayashi T, Yukaya H., et al.Intraoperative ultrasonography in resection of smallhepatocellular carcinoma associated with cirrhosis.Am J Surg. 1989;158:40–2.

Nonomura A, Mizukami Y, Kadoya M, Matsui O,Shimizu K, Izumi R.Angiomyolipoma of the liver: its clinical andpathological diversity.J Hepatobiliary Pancreat Surg. 1996;3:122–32.

Oberti F, Valsesia E, Pilette C, Rousselet MC,Bedossa P, Aubé C, et al.Noninvasive diagnosis of hepatic fibrosis or cirrhosis.Gastroenterology. 1997;113:1609–16.

Ochs A, Rottler RM, Schölmerich J.Clinical relevance of abnormal liver findings with ul-trasound.Hepatogastroenterology. 1994;41:9–15.

Pasha T, Gabriel S, Therneau T, Dickson ER,Lindor KD.Cost-effectiveness of ultrasound-guided liver biopsy.Hepatology. 1998;27:1220–6.

Petritsch W, Pristautz H, Eber B.Ist die Sonographie der Enzymdiagnostik bei derErkennung der Steatosis hepatis überlegen?Wien Klin Wochenschr. 1987;99:153–6.

Rettenmaier G.Diagnostik der umschriebenen Leberveränderungen.UltrasonographieZ Gastroenterol. 1988;23:11–7.

122

Ricci C, Longo R, Gioulis E, Bosco M, Pollesello P,Masutti F, et al.Noninvasive in vivo quantitative assessment of fatcontent in human liver.J Hepatol. 1997;27:108–13.

Richter J, Hatz C, Häussinger D.Ultrasound in tropical and parasitic diseases.Lancet. 2003;362:900–2.

Rickes S, Ocran K, Schulze S, Wermke W.Evaluierung Doppler-sonographischer Kriterien zurDifferenzialdiagnostik von hepatozellulären Karzino-men und Regeneratknoten bei Leberzirrhose.Ultraschall Med. 2002;23:83–90.

Ruelas-Villavicencio AL, Vargas-Vorácková F.In whom, how and how often is surveillance forhepatocellular carcinoma cost-effective?Ann Hepatol. 2004;3:152–9.

Saverymuttu SH, Joseph AE, Maxwell JD.Ultrasound scanning in the detection of hepaticfibrosis and steatosis.Br Med J (Clin Res Ed). 1986;292:13–5.

Schacherer D, Schölmerich J, Zuber-Jerger I.Diagnostik des hepatozellulären Karzinoms.Z Gastroenterol. 2007;45:1067–74.

Schild H, Kreitner KF, Thelen M,Grönninger J, Weber M, Börner N, et al.Fokal-noduläre Hyperplasie der Leber bei 930Patienten.Röfo. 1987;147:612–8.

Schölmerich J, Volk BA, Neuner C, Fröhlich J,Gerok W.Aussagefähigkeit der Sonographie bei Leber-metastasen unter besonderer Berücksichtigungdes Primärtumors.Dtsch Med Wochenschr. 1984;109:326–9.

Schölmerich J, Volk BA, Gerok W.Value and limitations of abdominal ultrasound intumour staging – liver metastasis and lymphoma.Eur J Radiol. 1987;7:243–5.

Schölmerich J, Groß V.Bildgebende Diagnostik in der Hepatologie.Klinik der Gegenwart 1997;IV:1–30.

123

Schwerk WB.Sonographie in der Diagnose der Leberzirrhose undihrer Komplikationen.Internist (Berl). 1987;28:477–88.

Sciuk J, Schober O.Nuklearmedizinische Diagnostik von Raumforderun-gen der Leber.Internist (Berl). 1997;38:917–23.

Seemann MD, Bonél H, Wintersperger B,Herrmann K, Sittek H, Reiser MF.Vergleich eines High-End-Ultraschallgerätes miteinem Spiral-CT-Scanner beim Screening von Leber-metastasen.Ultraschall Med. 1998;19:164–7.

Seitz K, Wermke W, Haag K.Sonographie bei portaler Hypertension und TIPS.Falk Foundation e.V. Freiburg. 1997.

Shibata T, Kubo S, Itoh K, Sagoh T, Nishimura K, Na-kano Y, et al.Recurrent hepatocellular carcinoma: usefulness ofultrasonography compared with computed tomog-raphy and AFP assay.J Clin Ultrasound. 1991;19:463–9.

Spamer C, Brambs HJ, Koch HK, Gerok W.Benign circumscribed lesions of the liver diagnosedby ultrasonically guided fine-needle biopsy.J Clin Ultrasound. 1986;14:83–8.

Steinmaurer HJ, Jirak P, Walchshofer J, Clodi PH.Treffsicherheit der Sonographie bei der Diagnosediffuser Leberparenchymerkrankungen – Vergleichzwischen Sonographie und LeberhistologieUltraschall Med. 1984;5:98–103.

Strohm WD, Weimer J.Sonographische Kriterien benigner Leberherde.Internist Prax. 1997;37:279–306.

Takayasu K, Moriyama N, Muramatsu Y,Makuuchi M, Hasegawa H, Okazaki N, et al.The diagnosis of small hepatocellular carcinomas:efficacy of various imaging procedures in 100patients.Am J Roentgenol. 1990;155:49–54.

Takayasu K, Okuda K.Imaging in Liver Disease. From Diagnosis to Treat-ment.Oxford University Press, Oxford, New York, Tokyo.1st edition. 1997.

124

Taylor KJ, Riely CA, Hammers L, Flax S, Weltin G,Garcia-Tsao G, et al.Quantitative US attenuation in normal liver and inpatients with diffuse liver disease: importance of fat.Radiology. 1986;160:65–71.

Vilgrain V, Lebrec D, Menu Y, Scherrer A, Nahum H.Comparison between ultrasonographic signs andthe degree of portal hypertension in patients withcirrhosis.Gastrointest Radiol. 1990;15:218–22.

Vizzutti F, Arena U, Romanelli RG, Rega L, Foschi M,Colagrande S, et al.Liver stiffness measurement predicts severe portalhypertension in patients with HCV-related cirrhosis.Hepatology. 2007;45:1290–7.

von Herbay A, Häussinger D, Gregor M, Vogt C.Charakterisierung und Detektion des hepatozellulä-ren Karzinoms (HCC): Vergleich der Ultraschallkon-trastmittel Sonovue (BR 1) und Levovist (SH U 508A).Ultraschall Med. 2007;28:168–75.

Vogel W, Kathrein H, Dietze O, Judmaier G.Zur Sonographie der Leber bei Morbus Wilson.Sonographische Beobachtungen an der Leber beiMorbus Wilson – Bedeutung für die Beurteilung derPrognose?Ultraschall Med. 1988;9:270–3.

Wachsberg RH, Jilani M.Duplex Doppler sonography of small (< 3 cmdiameter) liver tumours: intralesional arterial flowdoes not exclude cavernous haemangioma.Clin Radiol. 1999;54:103–6.

Wang TF, Hwang SJ, Lee EY, Tsai YT, Lin HC, Li CP,et al.Gall-bladder wall thickening in patients with livercirrhosis.J Gastroenterol Hepatol. 1997;12:445–9.

Wanless IR.Micronodular transformation (nodular regenerativehyperplasia) of the liver: a report of 64 cases among2,500 autopsies and a new classification of benignhepatocellular nodules.Hepatology. 1990;11:787–97.

125

Weimann A, Ringe B, Klempnauer J, Lamesch P,Gratz KF, Prokop M, et al.Benign liver tumors: differential diagnosis and indi-cations for surgery.World J Surg. 1997;21:983–90.

Weiss H.Probleme bei der Diagnostik von Raumforderungender Leber. In: Gebel J, Majewski A, Brunkhorst R.(Hrsg.). Sonographie in der Gastroenterologie.Springer-Verlag Berlin, New York 1987.

Weiss H.Sonographische Artdiagnostik tumoröser Leberver-änderungen.Ultraschall Klin Prax. 1991;6:257–71.

Wermke W.Sonographische Differentialdiagnose Leberkrank-heiten.Deutscher Ärzteverlag Köln 2006.

Wermke W, Gaßmann B.Tumordiagnostik der Leber mit Echosignalverstärker –Farbatlas.Springer-Verlag Berlin, Heidelberg, New York, 1998.

Zimmer T, Faiss S, Buhr HJ, Hamm B, Wiedenmann B.Bildgebende Verfahren in der Diagnostik neuroen-dokriner Tumoren des Gastrointestinaltraktes.Bildgebung. 1995;62:5–13.

Zoli M, Cordiani MR, Marchesini G, Abbati S,Bianchi G, Pisi E.Ultrasonographic follow-up of liver cirrhosis.J Clin Ultrasound. 1990;18:91–6.

126

Ursofalk® 500mg film-coated tablets, Ursofalk® 250mg capsules, Ursofalk® 250mg/5ml suspension. Active ingredient: ursodeoxycholic acid. Composition: One film-coated tablet contains: active ingredient: 500 mg of ursodeoxycholic acid. Other ingredients: magnesium stearate, polysorbate 80, povidone K25, microcrystalline cellulose, silica colloidal anhydrous, crospovidone (type A),talc, hypromellose, macrogol 6000. One hard capsule or 5 ml of suspension contains: active ingredient: 250 mg of ursodeoxycholic acid. Other ingredients: hard capsules: magnesium stearate (Ph.Eur.), titanium dioxide (E171), maize starch, colloidal silica anhydrous, gelatin, sodium dodecyl sulfate, purified water. Suspension: benzoic acid, citric acid, glycerol, microcrystalline cellulose, carboxymethylcellulose sodium, sodium chloride, sodium citrate 2H2O, sodium cyclamate, propylene glycol, purified water, xylitol, lemon flavouring. Indications: 1. Symptomatic treatment of primary biliary cirrhosis provided no decompensated liver cirrhosis is present. 2. Dissolution of cholesterol gallstones in the gallbladder. The gallstones must not be larger than 15 mm, must be radiolucent (not show up on x-ray) and the gallbladder must be functioning despite of (a) gallstone(s). 3. Biliary reflux gastritis (only Ursofalk® 250mg capsules). Dosage: For 1: 14 ± 2 mg/kg body weight daily. For 2: approx. 10 mg/kg body weight daily before going to bed. For 3: 1 hard capsule daily before going to bed. Contraindications: acute inflammation of the gallbladder and bile ducts; obstruction of the biliary tract (common bile duct or cystic duct), frequent biliary colic, radio-opaque calcified gallstones, disturbed contractility of the gallbladder, , hypersensitivity to bile acids or to any other ingredient, pregnancy, breast-feeding. Side effects: Pasty stools or diarrhea are common. Very rarely: severe right-side upper abdominal pain, calcification of gallstones, urticaria. During therapy of advanced stages of primary biliary cirrhosis, in very rare cases decompensation of hepatic cirrhosis (reversible). Ursofalk® 250mg/5ml suspension contains benzoic acid. Interactions: see patient information leaflet. Available on prescription only. Date of information: 05/2011.export

1) Parés A et al. Gastroenterology. 2006;130(3):715-20. 2) Shi J et al. Am J Gastroenterol. 2006;101(7):1529-38 („especially when given at early stages“). 3) Lindor KD et al. Mayo Clinic Proc 1997; 72:1137-40.

Ursofalk®

U r s o d e o x y c h o l i c a c i d

G o l d s t a n d a r d i n p r i m a r y b i l i a r y c i r r h o s i s

I m p r o v e s s u r v i v a l 1

E a s e s s y m p t o m s 2

P r o t e c t s a g a i n s t c o m p l i c a t i o n s 3

B r i n g i n g h o p e t o l i v e r p a t i e n t s

5 0 Y E A R S

www.drfalkpharma.com

Ursofalk_e_180x210_5-11_50 Years.indd 1 01.06.2011 11:04:33 Uhr

U42

e 7-

11/2

012/

3.00

0 PO

P