Comparison ofultrasonography,computed tomography and …peritoneal collaterals from the azygos and henii-zzygos systems. pattenit umbilical vein. splenomegaly andascites (Fig. 4)

Gut, 1987, 28, 242-247

Comparison of ultrasonography, computedtomography and 99mTC liver scan in diagnosis of Budd-Chiari syndromeS GUPTA, S BARTER, G W L PHILLIPS, R N GIBSON,AND H J F HODGSON

Fromti the Departments ofMedicine anid Radiology, Hammersmnith Hospital, Royal Postgradluate MedicalSchool, Lonidoni

SUMMARY Ultrasonography, computed tomography and "" Tc liver scanning are all useful indiagnosis of patients with the Budd-Chiari syndrome. In a study to determine their comparativevalue characteristic findings were recorded in all nine patients at ultrasonography and in sevenpatients at computed tomography. In contrast ""Tc liver scan showed a characteristic pattern inonly one of eight patients. In our experience intrahepatic venous abnormalities were seen better atultrasonography than at computed tomography. In addition, abnormality in the direction of bloodflow could be detected by pulsed Doppler examination. Ultrasonography is relatively inexpensive,readily accessible, does not require administration of radiation or contrast agents and thereforeshould be the primary non-invasive investigation of patients with Budd-Chiari syndrome, or thoseat risk of developing it.

The diagnosis of Budd-Chilri syndrome is generallymade by either liver biopsy or hepatic venography.'-8Although it is often commented that liver biopsy maynot be possible in many of these paitients,2' this hasbeen only occasionally so in our experience and thatof others.' The major disadvantages of liver biopsyand venographic studies, however, are that they areinvasive, require expertise and may be available onlyin specialised centres. Furthermore, neither of thesetechniques is ideally suited for screening or monitor-ing the course of the illness. Budd-Chiari syndromemay be more common than is realised as a diagnosismay be made more often with greater awareness`7and therefore rapid, non-invasive means which mightsuggest the condition are desirable.

Traditionally, isotope liver scan has been utilisedfor this purpose and maty show a characteristicpattern for central uptake in patients with Budd-Chiari syndrome.'' ' 7"' Both computed tomo-graphy""" and ultrasound' '2-` have recently beenfound to be useful in the diagnosis of Budd-ChiariAdIdrcvss tor(orrespol(iI1LCc: I)r S (upti JF S Xvr l n It .K;iotho Anilgo sllo%piItI 7(01 1- Imperial fighwy. I)ow ny. ( A 9)0214 ti A.Recevdil t'or pulrllictior 10.uI y198I6.

syndrome. Of these two, ultrasound is advantageousbecause it is cheaper and widely available.Experience with ultrasonography in Budd-Chiarisyndrome, however, is limited.We have studied the diagnostic features at ultra-

sonography in nine patients of Budd-Chiari syn-drome. A comparison with computed tomographyand """'Tc liver scanning was carried out to determinewhich of the three should be the non-invasive investi-gation of choice.

Methods

PATIFNTSStudies were carried out on nine patients with Budd-Chiari syndrome. Four patients were of a recentonset, five had long standing disease. The diagnosiswas made in all patients by hepatic venography and inaddition liver histology in eight patients showed acharacteristic pattern of sinusoidal dilatation andcongestion along with perivenous hepatocellularatrophy.' The clinical details of these patients aresummarised in Table 1. There were four men and fivewomen in the age range of 27-48 years.

242

on June 29, 2021 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.28.3.242 on 1 March 1987. D

ownloaded from

http://gut.bmj.com/

Non invasive diagnosis of Budd-Chiari syndrome

Table 1 Clinical characteristics ofpatient.s

Diraiilotio ofs VtnptofnsAge Ser belforediagtiosix (notntlis) Associ(ate(tilltnesses AsciIes HepatOInfeg(ilt Sp)lnomfegaliI

1 40 F 6 Nil + +2 45 M 2) llctcrozygous Ct- azintitrypsin dcticicncy + + +3 33 F 4 Nil + + +4 43 M Systcmic vasculitis + + +5 29 F 44 Nil + +6 42 F 7 Nil +++ + +7 27 M 2 Paroxysmal nocturnal hacmoglobinuria + + + + + +8 48 M 3 Nil + +9 28 F Nil + +

The radiological investigations done in the fivepatients with chronic disease were reviewed.Ultrasonography, "'Tc liver scans, and venographicstudies were done in all patients at the time of theinitial referral and all but one patient also hadcomputed tomography, which was carried out simul-taneously in all patients. The duration of symptomsbefore the diagnosis was made is listed in Table 1.One patient in this group died (patient 2, Table 1),but the remaining four have been under regularfollow up and all had recent ultrasound scans to verifyprevious findings. Ultrasonography was the initialinvestigation in the four patients of recent onset andthe diagnosis was confirmed subsequently by angio-graphy in all. Three of these patients had '"Tc liverscans and all had computed tomography of theabdomen. Follow up ultrasound scans were carriedout within two to four weeks on all these patients.

EQUIPMENTUltrasonographyReal time ultrasound examination was carried outusing a Toshiba SAL 50A, 5 MHz linear arrayscanner or a Diasonics 400 CV 3 5 MHz mechanicalsector scanner. In addition the direction of flow invenous collaterals was determined in three patientsusing a Diasonics 3.5 MHz cardiac probe with 3.0KHz Doppler.

All examinations included assessment of the con-fluence of veins draining into the inferior vena cava,presence or absence of the hepatic veins, alterationsin the calibre or the shape of the intrahepatic veins,presence of intrahepatic collaterals, enlargement ofthe caudate lobe and narrowing of the inferior venacava. Presence of splenomegaly, ascites and extra-hepatic venous collaterals including azygos, hemi-azygos and umbilical vein were noted.

Comptied tomographyExaminations were carried out on a SiemensSomatom 2 scanner with contiguous slices of 8 mmthickness and a scan time of five seconds. Scans were

done before and after injection of contrast medium(100 ml Urografin 310 M injected intravenously at therate of 1 ml/sec). All patients also received 1 litre ofdiluted oral Gastrografin before the examination toopacify bowel loops.The presence or absence of hepatic veins was noted

and hypertrophy of caudate lobe was determined byusing the previously described caudate/right loberatio." In addition presence of parenchymal abnor-malities, distortion or narrowing of the inferior venacava, presence of dilated azygos vein or collaterals,ascites and splenomegaly were noted.

Tc liver scanningTechnetium -99m labelled sulphur colloid(Amersham International) was used for isotopescanning. Two millicuries (80 MBq) dose of theisotope was administered intravenously and scanningwas started 10 minutes later. An InternationalGeneral Electric Maxi camera 400A on line to aMedtronics Data Systems computer was used to takeviews in anteroposterior, posteroanterior, right andleft lateral positions. Tomographic sections were alsogenerated.

Presence of liver or splenic enlargement, homo-geneity of uptake of the isotope by the liver andcaudate lobe enlargement were determined.

Results

The findings at radiological investigations aresummarised in Table 2. There were abnormalities onultrasound examination in all nine patients. Thenormal confluence of hepatic veins into the inferiorvena cava could not be demonstrated in any patient,although it is generally an easily recognised feature innormal individuals (Fig. 1). Clearly defined, outflow-ing hepatic veins were absent although distortedmiddle and right hepatic veins were present in onepatient (Fig. 2) and a left hepatic vein with anirregular calibre could be seen in two patients. Inthese three patients pulsed Doppler examination

243


http://gut.bmj.com

/G


ownloaded from

http://gut.bmj.com/

Glupta, Barter, Phillips, Gibson, antd Hodgson

zz -~~~~~~~~~~~~~~~~~~~_J 7_ - - -j _

C.) C) C C C) C)

L LL LUI L L LU LU

2D _JZ C C

s *- C> >oC n -C >

< < < 0 LU 0 <

-J oC C.) :6 < < 6 < 6 < <

>-~- _Z ZJZ

C C

-J CJJ i

7 -C7

->

jz~ ~ ~ ~EE-.-rcz 'OJ Z O i

/Vo,i i/Zlo.i'(IS (Ii(Ii4llOSi. o/' Bl(I(I- ( 'lijari .s VEzls^ldrfl 4

.... ze~~~~~~~~~~~~~~~~~~~~~~~~~.....*'w -:.

~~~-4 .~~~~ _s-*|F *s _

Fiu d T1-lal-s (i%( IIl(1,o lid ,((iii of I/tu us rintral/ icll ivilli

('110 C ( //lr(i'(l

showed that the blood flow via tileir veins was in anabnormal direction, awaly from the hepatic vein ostia.The venous structures which could be denmonstratedon the region of the ostia were prominlenit collateralnetwork of small, tortuous vessels. This pattern waisseen in aill patients with long standing disease. Thecaudate lobe was enlarged in all cases (Fig. 3).

In three patients the inferior venat cava waisnarrowed and in two it was almost occiuded by theenlairged caudate lobe. Less specific findings as atresult of portal hypertensioni included enlarged retro-peritoneal collaterals from the azygos and henii-zzygos systems. pattenit umbilical vein. splenomegalyand ascites (Fig. 4). Five patients were known to havegaistro-oesophaigeeal varices at endoscopy bhit thesewere not detected by ultrasonography. known to beinferior to endoscopy in detecting varices.' Followup ultrasound scaniis in four patients with recent onsetof sym)ptoms showed developmieinit of abnormialintrahepatic venoLis structures and collaterals. It wasdifficult to quantitate rigorously tile extent ofcollateral flow or the dlegree of caiudate lobecilIalr,cninciit.

245

o~

-- ... 'I

-hk.

7

am 'Air Y.1%S4w.

.. OW .

Ish:--1b-

.= 'I.r -

- IRKil L_. -".'

. I 11 .1 r. . (t (} tiot-iti(it confilience ol nepauc veins tiilti Inferior rella cava


http://gut.bmj.com

/G


ownloaded from

http://gut.bmj.com/

Guipta, Barter, Phillips, Gibsoni, a1n1 Hodgson

Fig. 3 Longituidinal ultrasound scan in Budd-Chiari.si(tdrome shlowing prominent caudate lobe (CL) betweenMarkers (patient 's head end to lefi).

Computed tomography showed an enlargedcaudate lobe with low attenuation elsewhere in sevenpatients and equivocal enlargement in one patient. Infour patients the caudate lobe was enlarged withabnormal attenuation. Intrahepatic veins did not fillafter administration of contrast in seven patients. Inone patient postcontrast studies were inadequate to

Fig. 4 Longitudinail altrasound scan through left lobe ofliver in Budd-(Chiari s'vndrorne (patient's head end to left).Showing' abnormnal intraheptic venous channels cauidall-v,and tortuous e,niaZ 'gous ccrllaterals beneath the inJerior.slr/a(' f tlle livel

evaluate venous filling. Other abnormalities weredemonstration of narrowing of the inferior vena cavaby the enlarged caudate lobe in two patients, ascitesin four patients, enlarged azygos vein in two patients,and enlargement of spleen in five patients.Of the three investigations ""."Te liver scan was the

least specific. Central accumulation of the isotope asdescribed characteristically with caudate lobeenlargement''=" was seen in only one of eightpatients. Liver or splenic enlargement was identifiedand areas of patchy uptake of the isotope werepresent in five patients.

If demonstration of abnormalities in hepatic veinsor enlargement of caudate lobe are taken as specificdiagnostic features, although caudate lobe enlarge-ment is of course possible in other circumstances,ultrasonography and computed tomography weresuccessful in making a diagnosis in all patients, and'99"1Tc liver scanning only 12*5%,.

Discussion

Before the advent of ultrasonography' "'' andcomputed tomography,:'""' non-invasive screeningand monitoring of patients with Budd-Chiaii syn-drome was carried out by '""Tc liver scanning.'' ""7Gallium liver scanning has recently been reportedto be useful.s In our experience, however, thecharacteristic central accumulation of the isotope isseen rarely and liver scans are a relatively poor meansof making a diagnosis, perhaps because of variablepreservation of liver function in different lobes.

This study indicates that the diagnostic accuracy ofultrasound in Budd-Chiari syndrome is similar toangiography. Absence of hepatic veins at the normalvenous confluence at the inferior vena cava,abnormal intrahepatic collaterals and enlargedcauduate lobe are of maximum discriminant value.Computed tomography is similarly successful insuggesting the diagnosis.i" "' The overall pattern ofvenous abnormalities and intrahepatic collaterals,however, is relatively clearer at ultrasonography andin suitably trained hands a diagnosis can possibly bemade with greater conviction.There are a number of other reasons for which

ultrasound theoretically may prove to be advanta-geous in the diagnosis of Budd-Chiari syndrome. Ithas been recognised recently that in this conditionvenous occlusion may not be total and partial throm-bosis has been described both in patients"` and inexperimental conditions."' In these circumstances aninability to cannulate the venous ostia at venography,or non-filling of hepatic vein side branches, both ofwhich are taken as evidence of venous occlusion, maygenerate erroneous information which could becorrected by ultrasonography. Narrowing or dis-

246


http://gut.bmj.com

/G


ownloaded from

http://gut.bmj.com/

Non1 illvasil'e diagnosis oJ'Budd-Chitri syndronie 247

placement of the inferior vena cava by an enlargedcaudate lobe in some patients may simulate lesionssuch as tumours or abscesses. Venography alonedoes not differentiate between such lesions andfurther imaging by computed tomography or arterio-graphy is often necessary.' In the Far East venouswebs are relatively common cause of Budd-Chiarisyndrome. The ultrasonographic appearances ofwebs in the inferior vena cava are not known to us,and as webs are very thin," it is conceivable that theymight only be detected with difficulty, but proximaldilatation of the hepatic veins might be expected.Another advantage of ultrasound is the demonstra-tion of features of portal hypertension such asenlarged retroperitoneal veins or a patent umbilicalvein, which may reflect the severity of venousocclusion of liver damage, and also associated portalvein or splenic vein thrombosis are obviously notdemonstrable by hepatic venography alone.When ultrasound examination was repeated over

two to four weeks in patients with recent onset ofsymptoms, it was possible to see evolution of thelesions, notably with development of intrahepaticcollaterals. The direction of flow in collaterals wasalso demonstrable using pulsed Doppler ultrasound,which showed abnormal intrahepatic channels withblood flowing away from the site of the hepatic veinostia. Although we have no evidence from this study,it is possible that ultrasound monitoring of thedevelopment of collaterals might correlate well witha good outcome from conservative management, oralternatively define a group of individuals with poorcollateral development, in whom surgical interven-tion might be required. We hope that prospectivestudies will investigate this.The evidence in this study, however, shows that

ultrasound is an effective tool in the detection ofBudd-Chiari syndrome, clearly superior to isotopeliver scan, and detection of normal ultrasonic findingsin a patient in whom the diagnosis of Budd-Chiari hasbeen considered should obviate the need for angio-graphic demonstration of the hepatic veins.References

1 Tavill AS, Wood EJ, Kreel J, Jones EA, Gregory M,Sherlock S. The Budd-Chiari syndrome: correlationbetween hepatic scintigraphy and the clinical, radio-logical and pathological findings in nineteen cases ofhepatic venous outflow obstruction. Gastroenterology1975; 68: 509-18.

2 Mitchell MC, Boitnott JK, Kaufman S, Cameron JL,Maddrey WC. Budd-Chiari syndrome: etiology, diag-nosis and management. Medicine 1982; 61: 199-218.

3 Clain D, Freston J, Kreel L! Sherlock S. Clinicaldiagnosis of Budd-Chiari syndrome. A report of sixcases. Am J Med 1967; 43: 544-54.

4 Maddrey W. Hepatic vein thrombosis (Budd-Chiarisyndrome). Hepatologv 1984; 4: 445-65.

5 Baert AL, Fcvery J, Marchal G, et al. Early diagnosis ofBudd-Chiari syndrome by computed tomography andultrasonography: report of five cases. Gastroenterology1983; 84: 587-95.

6 Harbin WP, Robert NJ, Ferrucci JT. Diagnosis ofcirrhosis based on regional changes in hepaticmorphology. A radiological and pathological analysis.Radiology 1980; 135: 273-83.

7 Gupta S, Cuss FM, Crofton ME. Baron JH, HodgsonHJF. Budd-Chiari syndrome and alpha-I antitrypsindeficiency. Hepato-Gastroenterol 1985; 32: 253-4.

8 Meindok H, Langer B. Liver scan in Budd-Chiarisyndrome. J Nucl Med 1976; 17: 365-8.

9 Thijs LG, Heidendal GAK, Hujgens PC, Mol JJ,Verkuyl M. The use of nuclear medicine procedures inthe diagnosis of Budd-Chiari syndrome. Clin Nucl Med1978; 3: 389-92.

10 Staab EV, Hartman RC, Parrott JA. Liver imaging inthe diagnosis of hepatic venous thrombosis in paroxysmalnocturnal haemoglobin-uria. Radiology 1975; 117: 341-8.

11 Rossleigh MA, Uren RF, Morris JG. Radionuclide liverscan. A screening test for the Budd-Chiari syndrome.Med J Aust 1984; 140: 234-6.

12 Rossi P, Sposito M, Simonetti G, Sposato S, CusumanoG. CT diagnosis of Budd-Chiari syndrome. J CornpuitAssist Tomogr 1981; 5: 366-9.

13 Yang PJ, Glazer GM, Bowerman RA. Budd-Chiarisyndrome: computed tomography and ultrasonographicfindings. J Comput Assist Tomogr 1983; 7: 148-50.

14 Cooper RA, Picker RH, Fulton AJ, Lunzer MR. Theultrasonic appearance of the liver in hepatic venousoutflow obstruction (Budd-Chiari syndrome). J ClinUltrasound 1982; 10: 35-6.

15 Makuuchi M, Hasegawa H, Yamazaki S. Moriyama N,Takayasu K, Okazaki M. Primary Budd-Chiari syn-drome: Ultrasonic demonstration. Radiology 1984; 152:775-9.

16 Powell-Jackson P, Karani J, Ede R, Meire H, WilliamR. Diagnosis of Budd-Chiari syndrome by liver ultra-sound and colloid scintigraphy. [Abstracti. Gui 1985;26: A568.

17 Subramanyam BR, Balthazer EJ, Madamba MR,Raghavendra BN, Horri SC, Lefleur RS. Sonography ofportosystemic venous collaterals in portatl hypertension.Radiology 1983; 146: 16 1-6.

18 Garty I, Horovitz 1, Kenyan A. The use of gallium-67liver imaging for the early diagnosis of Budd-Chiarisyndrome. J Nucl Med 1984; 25: 320-2.

19 Saint-Marc Girardini MF, Zafrani ES, Prigent A, ei al.Unilobar small hepatic vein obstruction: possible role ofprogestogen given as orail contraiceptive. Gastro-enterology 1983; 84: 630-5.

20 Maguire R, Doppman JL. Angiographic abnormalitiesin partial Budd-Chiari syndrome. Radiology, 1977; 122:629-35.

21 Nakamura H, Tamaka J, Horii S, Yashioka H, KurodaC. Partial Budd-Chilri syndrome. J Compiut AssistTomogr 1982; 6: 833-5.

22 Takeuchi J. Takada A, Hasumura Y, Matsuda Y,Ikegami F. Budd-Chiari syndrome aissociated withobstruction of the inferior vena cava1: A report of sevencases. Am J Med 1971; 51: 11-20.


http://gut.bmj.com

/G


ownloaded from
http://gut.bmj.com/

Documents

Comparison ofultrasonography,computed tomography and …peritoneal collaterals from the azygos and henii-zzygos systems. pattenit umbilical vein. splenomegaly andascites (Fig. 4)