Etsuo Keiko

Inoue, MD Kuriyama,

a

MD

Shinichi Hon1 a Tsuyoshi

Portal-Systemic Presence of Basal with High Signal

Sixteen patients with cirrhosis of the liver underwent cranial magnetic resonance (MR) imaging and transarterial portography to evaluate the relationship between basal ganglia lesions and portal-systemic collateral vessels. No neuropsychiatric disturbance was observed in any of the patients at the time of the MR examination, but four patients with portal-systemic encephalopathy were included in the study. Basal ganglia lesions, characterized by increased signal intensity on Tiweighted MR images, were observed in nine of the 16 patients, including the four with portal-systemic encephalopathy. These nine patients had large portal-systemic collateral vessels that were more than 10 mm in diameter. These collateral vessels were receiving blood from the superior mesenteric vein (SMV) in all nine patients. The lesions involved the globus pallidus and portions of internal capsules in a bilateral and symmetric fashion and did not exhibit mass effect. The authors conclude that there may be a significant relationship between high-intensity basal ganglia lesions and large portal-systemic collateral vessels receiving blood from the SMV. Index

terms:

142.1214 vein,

Basal

#{149} Liver,

flow

dynamics.

Radiology

1991;

the

Department

Center

for

Japan. bly.

the

Received October

accepted

1990

December

Diseases,

RadiolOsaka,

received 26.

Address

3

Osaka

scientific

1, 1990;

revision

537,

assem-

revision

request-

December reprint

17; re-

quests to El. 2

Osaka

Current

address:

University

pan. ,

RSNA,

1991

Department Medical

School.

Yoshifumi Narumi, MD a Chikazumi

of Radiology, Osaka,

Ja-

MD Kuroda,

#{149} Makoto

P

(1,2) is clinically different from the hepatic encephabopathy that occurs most commonly in patients with decompensated cirrhosis. Chronic neuropsychiatnic

states

in patients with temic collateral

chronic vessels.

psychiatric states with irreversible

may brain

authors

many

tomographic of patients and chronic

exist,

usually

portal-sysThese neurobe associated damage (3).

of the brain of been described (4-9).

indicating

rophy

cirrhosis

in liver

in a previous prevalence of

cerebral

patients

encephalopathy

(ii).

We have recently studied patients with cirrhosis of the liver and portalsystemic

encephalopathy

in whom

Ti-weighted MR imaging unusual lesions with high tensity

in the

basal

ganglia.

has shown signal inThese

basal ganglia lesions were also observed in subclinical hepatic encephalopathy after portal-systemic shunt-

Computed

brain atrophy (10). To our knowledge, the appearance of this condition on magnetic resonance (MR) images has not been previously reported except for structural MR imaging abnormalities

a

ing.

(CI) scans of the brains with cirrhosis of the liver persistent encephabopa-

thy have been obtained study to evaluate the

MD

Images’

without

encephalopathy

ORTAL-SYSTEMIC

Fujita,

MD

Encephalopathy: Ganglia Lesions Intensity on MR

by

RSNA

a

Pathologic changes these patients have

1, 957.79

Higashinari-ku,

October 29;

#{149} Portal

of Diagnostic

Adult

1-chome

From

studies,

179:551-555

the

Nakamichi

MR 761.794

957.71

From ogy,

ed

ganglia,

cirrhosis,

MD2 Kadota,

at-

MATERIALS

AND

METHODS

Sixteen patients with cirrhosis of the liver (12 men and four women aged 50-76 years), in whom direction of portal blood

flow and size of collateral vessels were confirmed by means of angiography, were studied with cranial MR imaging our

hospital

Abbreviation:

between

SMV

December

=

superior

1989

in and

mesenteric

vein.

551

a.

b.

d.

C.

Figure

1.

Images

strates

a large

of patients

splenorenal

with shunt

changes

creased

are

signal

image

disorders because

tients, lopathy.

four In

had those

globus

Patients were of possible

with

excluded

that

could Patients

Among

the

portal-systemic four patients,

16 pa-

panaumbilical

of the

pallidus

not show at the time

(a) Arterial vein.

frontal

and

portions

signal

by

and

(b)

temporal

of the

intensity

of patient

Ti-weighted

lobes.

internal

alterations

in

mood

and

(d)

1 obtained during the venous MR image (500/ 15) of same patient bilaterally (arrows). (c) Corresponding

genu

16 patients, the

and we studied

between

basal

unusual

ganglia

intensity

on Ti-weighted findings vessels.

in

MR imof the

hepatic encephalopathy. had normal physical

and

All four patients mental activities

To examine portal-systemic collateral vessels, transarterial portography was performed in all patients. For transarterial portography, 20 tg of alprostadil alfa-

at the

of cranial

MR

examination,

dex (Prostandin;

later

recurrent

findings

were

patients

but

with

time

would

attacks

552

different decompensate

have

of encephabopathy

a

from

Radiology

those

of the

cirrhosis

and

episodic

characterized

Osaka, nor

Japan) mesentenic

Ono was

Pharmaceutical,

injected

artery

into

approximately

the

are

no

seconds

the cornela-

supe30

capsule.

(500/

(arrows).

There

personality,

signal

image

bilaterally

distribution.

confusion and drowsiness, flapping tremor, and hyperammonemia. They recovered from encephalopathy in response to treatment. The duration of encephalopathic disturbances in these patients ranged from 3 days to 2 weeks. Transarterial portography was performed in all tion

of the internal

Ti-weighted

capsules

in a similar

ages and angiographic portal-systemic collateral

encephaclinical

portograph

and portions of the internal capsules signal intensity in globus pallidus and

decreased

and alcoholism were Wilson disease and would compli-

MR findings.

a small

matter

shows

in brain parenchyma brain imaging findings.

the

white

patient

patients did disturbance

with Wilson disease also excluded because Wernicke encephalopathy

cate

deep

MR examination.

neuropsychiatric from our study

and

globus pallidus shows decreased

involving

of same

July 1990. These neuropsychiatric

changes confuse

in the

intensity

(3,000/90)

of cranial

seen

e.

encephalopathy.

(arrows)

creased signal intensity involving image (3,000/90) of same patient emic

hepatic

phase demonshows in12-weighted

Multiple

15) of patient

small

(e) Corresponding lesions

in the

before

the

in-

12-weighted

white

matter.

injection

medium (iopamidol ing, Osaka, Japan];

isch-

3 shows

of contrast

[lopamiron; Scher370 mg of iodine per

milliliter). Portognaphs obtained in the venous phase were studied. The venous phase of celiac arteniography was also studied to evaluate portal-systemic collatenal vessels. The diameter of the collateral vessel was measured at its origin from the portal axis. All cerebral with a 1.5-T

MR images superconducting

(Magnetom;

Siemens,

using echo sequences 90 (repetition ny)

The and

by

Erlangen,

Ti-

sequential there was

tions. All transaxial

were

images plane.

and T2-weighted of 500/15 and time msec/echo sections a 2-mm were

obtained system Germaspin3,000/15, time msec).

were 8 mm gap between obtained

in

thick, secthe

May

1991

b.

a.

e.

d.

C.

Figure 2. Images of patient 5, a 64-year-old woman without clinical signs or symptoms of hepatic encephalopathy. obtained during the venous phase shows a lange paraumbilical vein. (b) Ti-weighted MR image (500/ 15) shows involving globus pallidus and portions of the internal capsules bilaterally (arrows). (c) Corresponding 12-weighted demonstrates decreased intensity involving a similar distribution. (d) Ti-weighted MR image (500/15) obtained shows foci of increased intensity within the cerebral peduncules (arrows). (e) Corresponding 12-weighted image intensity within the cerebral peduncules.

in the

RESULTS Angiographic summarized creased signal

and MR findings in the Table. Areas intensity relative

are of into

cerebral white matter were visualized in the basal ganglia on Tiweighted images in nine of the 16 patients (MR-positive finding). Four of those nine patients had portal-systemic tients

high temic

encephalopathy. with basal

signal

Ihe ganglia

intensity

collateral

had

vessels

nine lesions than

tients,

Volume

no

increased

179

Number

#{149}

signal

2

intensity

tive

finding).

of these

SMV

relative

Angiognaphc

seven

and

region

to

matter was visualized images (MR-negapatients

splenic

findings

were

as fol-

had small portalvessels less than 5 originating from the

vein,

three

had

two 10

patients

had

no collateral

pon-

yes-

sels.

MR images involved the portions bilateral

showed globus

of the internal and symmetric

that the lesions pallidus and capsules fashion

1,2).

Images

ative findings Six patients

tab-systemic collateral vessels oniginating from the splenic vein (flow from the SMV was not detected), and

portal-sys-

larger

ganglia

lows: two patients systemic collateral mm in diameter

paof

mm in diameter that were receiving blood from the superior mesentenic vein (SMV). In the other seven pa-

basal

cerebral white on Ti-weighted

in a (Figs

tensity cubes. signal

(a)

Arterial

portograph

increased signal intensity MR image (3,000/90) at a lower level than c (3,000/90) shows decreased

of patients

are had

with

MR-neg-

shown in Figure 3. foci of increased in-

within the cerebral pedunThese lesions had increased intensity on Ti-weighted im-

ages,

but had decreased signal intenon corresponding 12-weighted images. The area involving abnormal sity

signal ginated

intensity without

was smoothly mass effect

with all pulse sequences. two patients underwent MR examinations, there nite evidence of change abnormalities. glummne was

tened

to one

Although subsequent was no defiin the signal

Gadopentetate intravenously

patient;

maror edema

dimeadminis-

no evidence Radiology

of 553

#{149}

b.

a.

Figure

3. Images of patient i2, a 53-year-old ganglia. (a) Arterial portograph obtained

basal sels

are

seen

splenorenal (500/15)

from

the

shunts shows

no

SMV.

(b)

(arrows)

Celiac

arteriograph

originating

increased

signal

from intensity

abnormal enhancement related to these basal ganglia signal abnormalities was noted on the MR images. DISCUSSION It is noteworthy that basal lesions with increased signal ty on

11-weighted

images

ganglia intensiat high

field strength were observed in nine of the 16 patients with cirrhosis of the liver. These nine patients had large portal-systemic collateral yessels that were receiving blood from the SMV. Four of those nine patients had portal-systemic encephalopathy. All four patients with portal-systemic encephabopathy had SMV collateral vessels larger than 10 mm in diameten, and it is significant that another five patients without neuropsychiatnc disturbance also had collateral vessels larger than 10 mm in diameten. Although the basal ganglia lesions

did

not

correlate

well

with

neu-

robogic symptoms and mental status, there may be a significant relationship between high-intensity basal ganglia lesions and large portal-systemic collateral vessels filled via the SMV. Sherlock et al (1) reported that chronic disturbance of neuropsychiatric states usually exists in patients with chronic portal-systemic collateral vessels, and they called this condition portal-systemic encephabopathy. Recently, it has been recognized that the spectrum of portal-systemic encephabopathy includes not only the traditional clinical stages, but also a latent on subclinical stage that is associated with permanent structural cer554

a

Radiology

man with no increased signal intensity during the venous phase. No collateral obtained

the splenic in the

basal

during

vein.

the

venous

(c) Ti-weighted

in the yes-

phase

shows

MR image

ganglia.

ebral changes (12-14). This milder form of portal-systemic encephalopathy is often difficult to detect by means of ordinary clinical evaluation, but MR imaging may help detect its presence. In this series, we observed basal ganglia lesions on Tiweighted images in patients with cirrhosis of the liver and large portalsystemic collateral vessels, including five patients without any clinical signs and symptoms of encephabopathy. These five patients had basal ganglia lesions of high signal intensity on Ti-weighted images; they did not show signs of neuropsychiatric disturbances at 4-6 months followup, but they may develop them in the future. The observation of lesions with increased signal intensity on Tiweighted images is unusual, and such signal intensity has been associated with relatively few tissues. It occurs in lipids and with paramagnetic species of iron such as methemoglobin (hemorrhage) or melanin (melanoma) and with manganese (which has a paramagnetic effect and shows a preferential affinity for the gbobus pallidus in experimental studies) (15). The presence of ectopic Schwann cells in neurofibromatosis may also cause an increase in signal intensity (16). While a histopathobogic examination was not performed in our

study,

a review

of the

basic

pathologic aspects of hepatic abopathy provides significant mation on the unusual signal ty of the lesions. It is not known why basal

ganglia

lesions

intensity

had

increased

signal

encephinfonintensi-

C.

on Ti-weighted field in the portal-systemic

filled

via

nine

the

MR images at high patients with large collateral vessels

SMV.

However,

it is

generally believed that gut-derived toxins are responsible for the genesis of neurobogic disorders, although their nature and mechanism of toxicity still have not been clearly defined (4,5).

Nonenbeng

with

portacaval

en resin

an

(6) reported

shunts

ammoniated by

means

cation of

gavage

that

that

were

rats

giv-

exchange feedings

developed encephalopathy, and that astrocytic alterations were seen in the brain tissue. It has been reported that in patients with hepatic encephabopathy, type II Alzheimen cells-which are astrocytes with large, pale nuclei, margination of chromatin, and prominent nucleoli-are more numerous in the cerebral cortex, basal ganglia, cerebellum, and brain stem (5-8). In patients with a more protracted, undulating course and evidence of portab-systemic

collateral

severe and progressive abnormalities develop (8,9). Nonenberg (5,18)

vessels,

more

microscopic with time and Voorhies

May

1991

et al (17) found proliferation of mitochondnia, rough endoplasmic reticubum, and vacuoles (lipofuscin pigment) in the cytoplasm of astrocytes by means of electron microscopy. These ultrastructural changes in the astrocyte are thought to be the cause of Ti shortening. These astrocytic alterations are seen to a variable degree in the cerebral cortex, basal ganglia, cerebellum, and brain stem, but only basal ganglia lesions and cerebral peduncules demonstrate increased signal intensity on Ti-weighted images. Although discrepancies between the location of astrocytic alterations and that of the increased signal intensity were considered, Norenberg noted that astrocytes in different parts of the brain may be fundamentally different and may, therefore, respond differently in different states (18). It is believed that the concentration of these astrocytes increases markedly in the regions of Ti shortening. In conclusion, basal ganglia lesions with increased signal intensity on Ti-weighted images at high field strength were observed in nine patients with cirrhosis of the liver and large portal-systemic collateral yessels that were receiving blood from the SMV, including four patients with portal-systemic encephabopathy.

Although

by proved,

Volume

not

our

179

a

histopathobogical-

results

Number

suggest

2

that

large sels

portal-systemic originating

from

collateral

yes-

the

may

SMV

cause increased signal intensity on Ti-weighted MR images, and that brain MR imaging may be valuable in helping to detect portal-systemic encephabopathy, even in the subclinical stage. U References 1.

2.

3.

4. 5.

6.

7.

Sherlock 5, Summenskill WHJ, White LP, Phean EA. Portal-systemic encephalopathy: neurological complications of liver disease. Lancet 1954; 267:453-457. Summerskill WHJ, Davidson EA, Sherlock S, Steiner RE. The neuropsychiatnic syndrome associated with hepatic cirrhosis and an extensive collateral circulation. Q Med 1956; 25:245-266. Sherlock S. Diseases of the liver and biliary system. 7th ed. Oxford: Blackwell, 1985. Zieve L. The mechanism of hepatic coma. Hepatology 1981; 1:360-365. Norenberg MD. The role of astrocytes in hepatic encephalopathy. Neurochem Pathol 1987; 6:13-29. Norenbeng MD. A light and electron microscopic study of experimental portalsystemic (ammonia) encephalopathy. Lab Invest 1977; 36:618-627. Adams RD. Foley JM. The neurological disorder associated with liver disease. Assoc Res Nerv Ment Dis Proc 1953; 32:198-

10.

11.

12.

13.

rhosis.

14.

15.

9.

Waggoner RW, Malamud N. Wilson’s disease in the light of cerebral changes following ordinary acquired liver disorders. J Nenv Ment Dis 1942; 96:410-423. Victor M, Adams RD. Cole M. The acquired (non Wilsonian) type of chronic hepatocerebral degeneration. Medicine 1965; 44:345-396.

Lancet

1986;

2:893-895.

Bernthal P, Hays A, Tanter RE, Ihiel DV, Lecky J, Hegedus A. Cerebral CT scan abnormalities in cholestatic and hepatocellular disease and their relationship to neunopsychologic test performance. Hepatology 1987; 7:107-114. Newland MC, Ceckler IL, Kordower JH, Weiss B. Visualizing manganese in the primate basal ganglia with magnetic resonance imaging. Exp Neurol 1989; 106:251258.

16.

Mirowitz tensity weighted sis. AJNR

17.

Voorhies

SA, Sartor basal ganglia MR images 1989;

TM,

tito CK, Plum in the young

237. 8.

Zeneroli ML, Cioni G, Vezzelli C, et al. Prevalence of brain atrophy in liver cirrhosis patients with chronic persistent encephalopathy: evaluation by computed tomography. J Hepatol 1987; 4:283-292. Moore JW, Dunk AA, Crawford JR. et al. Neuropsychological deficits and morphological MRI brain scan abnormalities in apparently healthy non-encephalopathic patients with cirrhosis. J Hepatol 1989; 9:319-325. Gitlin N. Subclinical portal-systemic encephalopathy. Am J Gastroenterol 1988; 83:8-11. Tarter RE, Hays AL, Sandford 55, Van Thiel D. Cerebral morphological abnormalities associated with nonalcoholic cir-

18.

K, Gado M. High-inlesions on 11in neurofibromato-

10:1159-1163.

Ehrlich

ME,

Duffy

TE, Pe-

F. Acute hyperammonemia primate: physiologic and

neuropathologic correlates. Pediatr Res 1983; 17:970-975. Norenberg MD. The astrocyte in liver disease. Adv Cell Neurobiol 1981; 2:303352.

Radiology

a

555