Mario Marina

Savoiamdo, Grisoli,

MD MD

#{149} Liliana #{149}

Daniela

Strada, MD #{149} Floriano Girotti, MD Testa, MD #{149} Raffaele Petrillo, MD

Olivopontocerebellar and Relationship

Atrophy: to Multisystem

Clinical diagnosis of olivopontocerebellar atrophy (OPCA) must be confirmed by radiologic demonstration of atrophy in an appropriate distribution. OPCA may be associated with degeneration of other systems in multisystem atrophy (MSA). The authors report 23 cases of OPCA, eight of which were associated with MSA. Atrophy involved the cerebellum, pons, and middle cerebellar peduncles in all cases. On intermediate and T2-weighted magnetic resonance (MR) images, abnormal signal intensity was always observed in the transverse pontine fibers, middle cerebellar peduncles, and cerebellum, structures known from pathologic study to degenerate in OPCA. Pyramidal tracts and superior cerebellar peduncles stood out because of their normal signal intensity. Of the eight patients with MSA, four also had variable abnormal signal intensities in the putamen. The authors believe that the combination of atrophy and abnormal signal intensity in the appropniate distribution strongly supports the diagnosis of OPCA. In some cases, MR imaging may demonstrate involvement of different systems, thus confirming the diagnosis of MSA. Index terms: Basal ganglia, MR studies, 15.1214 #{149} Brain, atrophy, 15.83 #{149} Brain, diseases, 15.83 #{149} Brain, MR studies, 10.1214 #{149} Brain stem, MR studies, 15.1214 Radiology

1

1990;

From

(MS., Istituto Celoria

the

Departments

L.S., MG.) Nazionale 11, 20133

of Radiology,

174:693-696

of Neuroradiology

and Neurology Neurologico Milan, Italy;

Hospital

(F.G., D.T.), “C. Besta,” Via the Department

of the University

Pennsylvania, Philadelphia (R.A.Z.); Department of Diagnostic Radiology Unit), Istituto Nazionale dei Tumori, (R.P.). Received July 31, 1989; revision

of and the E (MR Milan request-

atrophy disof the

(OPCA) is a degenerative characterized by atrophy

ease pons,

middle

cerebellar

peduncles,

and cerebellam hemispheres (1). OPCA may present a spectrum of clinical features. It can occur in both familial and sporadic cases. Some nondistinctive enzymatic defects have been reported (2,3). Attempts have been made to organize this nosologic entity by separation into five groups (4). Moreover, transitional and mixed forms exist, in which involvement of the extrapyramidal system and autonomic failure are present in addition to the cenebellar disorder; these forms have been labeled multisystem atrophy (MSA) (1). Because differences between the clinical varieties of OPCA may be subtle and because the clinical symptoms and signs may overlap those of other disorders, most authors now agree that madiologic demonstration of atrophy of the proper structures is essential in establishing sis of OPCA (4-6).

the

diagno-

in the

pons,

middle

cerebellar

tracts

and

the

pe-

tegmentum

spared (1); in the cerebellum, a variable loss of Pumkinje cells may be seen. Themefore, we looked for magnetic resonance (MR) signal abnormalities that correspond to these histologic features in a series of patients with the tentative clinical diagnosis of OPCA or other cemebellar degenerative

MD

#{149}

diseases.

AND

METHODS

Of 2,200 0.5-T MR studies

obtained

within a 12-month period at the Istituto Nazionale Neurologico “C. Besta” in Milan, 18 were characterized by a peculiar pattern of atrophy and abnormal signal intensity in the posterior fossa structures. The clinical diagnosis was OPCA in eight cases, MSA in eight cases, a progressive

cerebellar disorder in ebellar and brain stem case. On review of the cal picture in all cases OPCA, although eight

one case, and a cersyndrome in one records, the cliniwas consistent with patients had addi-

tional signs of an extrapyramidal or of autonomic failure, pointing diagnosis

disorder to the

of MSA.

In the same period, three with the tentative diagnosis

derwent

MR imaging.

other patients of OPCA un-

The first

patient

had signs of multiple areas of demyelination and was given a final diagnosis of multiple sclerosis. The second patient had

cerebellar

atrophy,

dle cerebellar

The third ings.

but the pons

peduncles

patient

had normal

No definitive

clinical

reached in the last Of the 18 patients had

a familial

cases

were

MR finddiagnosis

patients. in this series,

was

two

history

were

and midnormal.

sporadic.

only

of OPCA; There

the

were

12

male and six female patients, aged 33-64 years (mean, 52.7 years). The duration of disease was 1-10 years

duncles, and cerebellum. The fibers affected in the pons are the transverse pontine fibers, while the pynamidal

PATIENTS

two other

In addition to gross atrophy, we know from pathologic studies that theme are characteristic histologic changes, such as the loss of specific fiber tracts and the presence of gliosis

A. Zimmerman,

MR Diagnosis Atrophy’

(LIVOPONTOCEREBELLAR I_,

#{149} Robert

are

(mean, 4.2 years). The MR studies were obtained on a 0.5-T Gyroscan unit (Philips Medical Systems International, Eindhoven, The Netherlands) and included spin-echo intermediate

(SE) transverse sections with and T2-weighted images,

generally

with

a repetition

time

(TR) of

1,900-2,000 msec and echo times (TEs) of 50 and 100 msec, respectively. Section thickness was usually 6 mm (in rare instances, 8 mm). Sagittal Ti-weighted im-

ages were

obtained with either SE fast field echo (FFE) sequences. Coronal SE images with the same TR and TE and/or coronal FFE Tior, more

always

frequently,

weighted images were obtained two patients. In eight patients,

in all but a study

was also obtained

Magnetom

with

a i.5-T

ed September 7; revision received October 20; accepted November 1 . Address reprint requests C

to MS. RSNA, 1990

Abbreviations: atrophy, SE

=

spin

FFE fast field echo, MSA echo, TE echo time, TR

multisystem repetition

atrophy, time.

OPCA

olivopontocerebellar

693

(Siemens,

unit lic

Erlangen,

of Germany).

ously

In

obtained

0.15-T

available. Five patients from another

Federal

two

a previ-

examination

was

(one man, four women) institution (Hospital of the

University

of Pennsylvania,

phia)

with

identical

ings

were

clinical

Repub-

patients,

also

Philadel-

posterior

included.

diagnosis

fossa

They

of OPCA

find-

all had

or

a

of cerebel-

lar degenerative disease, and their ages ranged from 20 to 65 years (mean, 40.4 years). They were all examined with a i.5-T

Signa

unit

Milwaukee). two

(GE

Medical

Sagittal,

cases,

coronal

sections

The section thickness weighted images and diate

and

Systems,

transverse,

and,

were

in

obtained.

was 3 mm for Ti5 mm

T2-weighted

for

interme-

a.

images.

The total number of OPCA cases reviewed was, therefore, 23; eight of these patients ter

had

to cases

MSA.

We shall

of “OPCA

patients) and cases (eight patients).

refer

without

hereaf-

MSA”

of “OPCA

with

(15

b.

Figure

1.

Midline

with OPCA. of its inferior pons.

sagittal

Ti-weighted

images

(FFE

378/16,

part.

atrophy is seen in the cerebellum (b) In this patient, the atrophy is more

90#{176} flip

(a) Moderate

Supratentorial

atrophy

is also

and severe

angle)

of two

in the pons, and

with

involves

the

patients

flattening whole

present.

MSA”

RESULTS All 23 patients had atrophy of the brain stem and cerebellum. This was more marked than the atrophic changes in the supratentonial brain, which were present in only 12 cases. Ti-weighted images best showed the distribution of atrophy; the midline sagiftal

section

showed

that

the

atmo-

phy selectively involved the pons with flattening of its inferior pant and with loss, therefore, of the normal bulge of the pons above the profile of the medulla oblongata (Fig i). On coronal Ti-weighted images, besides the cerebellar atrophy, which was more marked in the hemispheres, atrophy of the middle cerebellar peduncles gave a characteristic pointed appearance to the lateral aspect of the pons. This resulted in a loss

of the

normal

rounded

Figure 2. Coronal Ti-weighted hemispheres (a) (SE 550/30) and pointed appearance to the lateral

images show that atrophy parts of the

that cerebellar of the middle pons (b) (FFE

atrophy cerebellar 370/16,

is most marked in the peduncles gives a 90#{176} flip angle).

appear-

ance of the middle cerebellar pedundes at their origin from the pons (Fig 2). Atrophy was also well shown, within the same distribution, on the transverse sections. No signal inten-

the cerebellum was sometimes difficult to appreciate on transverse sections where supratentonial brain was not available for direct comparison.

than in the pallidum. Of the eight patients with OPCA with MSA, seven also underwent T studies. In four of the eight who

However,

sity

signal

underwent 0.5-T studies, creased signal intensity

abnormalities

were

seen

on

Ti-

weighted images. Definite slight hyperintensity was always seen on intermediate images and was seen to a lesser extent on T2weighted images. Signal abnormality involved the transverse pontine fibers ventral to the tegmentum, on the raphe and on the anterior and antemolateral contours of the pons where the transverse pontine fibers merged with the abnormal signal of the

middle

cemebellar

peduncles.

Be-

cause of their normal signal intensity, the pyramidal tracts stood out (Fig 3). Diffuse

694

slight

Radiology

#{149}

hypenintensity

in

peared

the intensity

obvious

abnormal

increased

in the

cerebellum

when

the

ap-

transverse

sections included parts of the temporal or occipital lobes or when coronal sections made the comparison easy (Fig 4). In the supratentonial compartment, particular attention was paid to changes in the basal ganglia. Of the i5 cases of OPCA

without

MSA,

abnormal-

none

showed

signal

ities in the basal ganglia on the 0.5-T study (iO studies). Six patients underwent i.5-T examination; only one had evidence of moderate signal loss on T2-weighted images in the putamma, which was slightly more evident

1.5-

slightly inwas observed

in one or both of the putamina on intermediate and T2-weighted images (Fig 5). The findings were normal in three of the seven 1 .5-T studies. In two the

cases signal loss was depicted putamina on T2-weighted im-

ages;

the

signal

loss

was

equal

in

to or

more marked than that in the pallidum. In one case the signal intensity in the putamina was increased, and in another both decreased and increased in

the

signal same

normal i.5-T in the patients examinations.

intensity structure.

studies with These

were The

present

four

ab-

were obtained abnormal 0.5-T patients had the

March

1990

Figure 3. verse sections bers between

(a) Transverse

section (SE 1,933/50, 0.5 T) shows high signal intensity mainly in the middle cerebellar peduncles. (b, c) Transpatient (SE 1,933/50 and 1,933/100, respectively; 0.5 T) show high signal intensity in the transverse pontine fiand the base of the pons, on the anterior and lateral aspects of the pons, and on the raphe, therefore delimiting the normal pyramidal tracts. Abnormal signal intensity is more evident on intermediate-weighted image (b) than on T2-weighted image (c). In C, the high signal intensity of the ventral transverse pontine fibers becomes confused with the signal of cerebrospinal fluid. (d) High signal intensity at a more cranial level in transverse pontine fibers is demonstrated with a i.5-T unit (SE 2,000/90). Signal abnormalities were

usually

in a different tegmentum

slightly

more

...,

a. Figure

evident

on 0.5-T studies

than

on l.5-T

studies.

.

.

b.

in addition

4.

Transverse (a) (SE 3,000/35, 1.5 T) and coronal (b) (SE 1,933/100, 0.5 T) T2weighted images of two patients show hyperintensity of the cerebellum. The middle cerebellar peduncles merging into the white matter of the cerebellar hemispheres are particular-

ly hyperintense normal arrow),

most

(open

signal with

arrow),

while

intensity. The normal or nearly

inferior normal

marked

On the decrease

extrapyramidal

cerebellar signal

signs.

i.5-T studies, a fairly marked in signal intensity on T2-

weighted

images

was

observed

in the

substantia

Smudging in three

the superior

nigra.

and in two cases MSA. Ten of the

was with

of OPCA 23 patients

present MSA

without had one

or more small areas of increased signal intensity in the peniventnicular or subcontical white matter of the cerebral hemispheres on T2-weighted images. These abnormalities were more frequent in older patients.

DISCUSSION Since Thomas

1900, first

Volume

174

when D#{233}jerine and introduced OPCA, the

Number

#{149}

3

concept of the become

types

occasionally

of its borders cases of OPCA

cerebellar

peduncles intensity.

peduncles are

also

(arrowheads) recognizable

been

created

and

to the abnormalities cerebellum

(not

in the brain

shown).

have (small

solid

of this degenerative disease cerebellum and brain stem has more variegated (1). Sub-

have

stem

to account

for different clinical aspects and inhenitance patterns (4). Even the name has been questioned; since degeneration of the inferior olives is secondany to cortical cerebellar lesions, the disease could simply be termed “pontocerebellar atrophy” (i). Since the clinical features of several cerebellar degenerative and nondegenerative diseases can overlap those of OPCA, and since a unifying biochemical marker is lacking, positive imaging findings have become essential in diagnosing OPCA (4-6). The diagnosis can be made when

there is a combination of the appropniate clinical picture and the appropniate brain stem and cerebellam atrophy demonstrated by means of computed tomography or MR imaging (4-8) (Figs 1, 2). A further step to-

ward

specificity

is the

demonstration

of abnormal signal intensity in structunes that are known to degenerate with this disease. Previous reports did not mention abnormal signal intensities in patients with OPCA (5,6,9,10). From pathologic studies, we know that in OPCA there is a “dying-back” process of the pontocerebellar fibers, which originate from the pontine nuclei. These fibers have a transverse course in the pons (hence their name, transverse pontine fibers) and nun to the cerebellum through the middle

Radiology

695

#{149}

cerebellar peduncles. A variable degree of degeneration of the Purkinje cells and of their fibers that project to the dentate nuclei is observed. There is also a retrograde cell loss of the infenior olives, secondary to cortical cerebellar lesions (1). The MR findings in this series of

patients

match

those

expected

on the

basis of OPCA histopathology, in terms of both atrophy and abnormal signal intensity (Figs 3, 4). Not only are the abnormalities observed where they should be, but the structunes that should be spared are spared. That is shown in the case of the pyramidal tracts (which become individually demonstrated due to degeneration

verse

of the

pontine

fibers),

tegmentum, and bellam peduncles

The

surrounding

the

trans-

of the

that emerges from this constitution of appropni-

already

reported

in the region with pathologic

weighted borders Parkinson

Marked

putamen

obtained is equal

and

nigra, on T2-

smudging

of its

toward the red nucleus disease and parkinsonian

syndromes.

the

abnormalities

of the substantia hypomntensity

images

on

hypomntensity

T2-weighted

in

images

at high field strengths to or more evident than

that that

also been observed on images obtamed at 0.5 T (15). These nonconflicting features may correspond to an increased amount

or other

paramagnetic

patients

obtained

Shy-Drager

pontine fibers. The middle tracts (solid arrows) and the

stained

with

a 1.5-T

studies

in

unit.

in the

cases

of

of

putamen.

posterior

our

data

MR

imaging

the

can

6.

7.

Of the OPCA

8.

9.

If we com-

fossa

support

5.

syndrome

pare the 0.5-T studies, available in 18 cases, we find no abnormal signal in the basal ganglia in 10 cases of OPCA without MSA, while abnormalities are seen in the putamen in four of the eight cases of OPCA with MSA. Therefore, although involvement of the basal ganglia was seen less definitely and constantly than involvestructures,

contribute

10.

11.

12.

13.

hypothesis

that

to the

di-

agnosis of MSA by demonstrating involvement of the structures of different systems. U

(cf Figs 3 and

cerebellar superior

4a). (b courtesy

pecere-

of Orso

14.

15.

Garcia de la Rocha ML, Moreno Martinez JM, Garrido Carrion A, Margalet Fern#{225}ndez P. Martin Araguz A. Present criteria for diagnosis in vivo of olivopontocerebellar atrophy. Acta Neurol Scand 1988; 77:234-238. Waespe W, Hayek J, Wichmann W, Bader JP. Die olivo-ponto-zerebell#{228}re atrophie als wichtige differential-diagnose ataktischer gangstOrungen beim #{227}lteren patienten. Schweiz Med Wochenschr 1988; 118:10321038. Savoiardo M, Bracchi M, Passerini A, Visciani A, Di Donato S. Cocchini F. Computed tomography of olivopontocerebellar degeneration. AJNR 1983; 4:509-512. Huang YP, Plaitakis A. Morphological changes of olivopontocerebellar atrophy in computed tomography and comments on its pathogenesis. In: Duvoisin RC, Plaitakis A, eds. Advances in neurology. Vol 41, The ohvopontocerebellar atrophies. New York: Rayen, 1984; 39-85.

Rutledge

1.

2.

16. DR.

Diseases

of the

basal

gan-

glia, cerebellum and motor neurons. In: Hume Adams J, Corselhis JAN, Duchen LW, eds. Greenfield’s neuropathology. 4th ed. New York: Wiley, 1984; 699-747. Finocchiaro G, Taroni F, Di Donato S. Glutamate

Shy-Drager

Tomogr

Oppenheimer

dehydrogenase

JN, Hilal

SK, Silver

AJ, Defendini

Borit

syndrome.

J Comput

Assist

1989; 13:555-560.

A, Rubinstein

tonigral degenerations: and nosology. Brain

U, Urich 1975;

H. The striaputaminal pigments 98:101-112.

in ohivopontocerebehlar

atrophies: leukocytes, fibroblasts, and muscle mitochondria. Neurology 1986; 36:550-553.

4.

Duvoisin

Chokroverty hydrogenase a particular

RC,

Nicklas

WJ,

Ritchie

V. Sage

J,

S. Low leukocyte glutamate deactivity does not correlate with type of multiple system atrophy.

Neurol Neurosurg Psychiatry 1988; 51:15081511. Duvoisin RC. The olivopontocerebellar atrophies. In: Marsden CD, Fahn S. eds. Vol 7, Movement disorders 2. London: Butter-

worths,

R.

Fahn S. Study of movement disorders and brain iron by MR. AJNR 1987; 8:397-411. Nabatame H, Fukuyama H, Akiguchi I, Kameyama M, Nishimura K, Nakano Y. Spinocerebellar degeneration: qualitative and quantitative MR analysis of atrophy. J Comput Assist Tomogr 1988; 12:298-303. Drayer BP, Olanow W, Burger P. Johnson GA, Herfkens R, Riederer S. Parkinson plus syndrome: diagnosis using high field MR imaging of brain iron. Radiology 1986; 159:493498. Braffman BH, Grossman RI, Goldberg HI, et ah. MR imaging of Parkinson disease with spin-echo and gradient-echo sequences. AJNR 1988; 9:1093-1099. Drayer BP. Imaging of the aging brain. II. Pathologic conditions. Radiology 1988; 166:797-806. Pastakia B, Pohinsky R, Di Chiro G, Simmons TJ, Brown R, Wener L. Multiple system atrophy (Shy-Drager syndrome): MR imaging. Radiology 1986; 159:499-502. Savoiardo M, Strada L, Girotti F, et al. MR imaging in progressive supranuclear palsy

and

References

ence of gliosis and cell loss-are in the putamen in stniatonigral generation (16). Stniatonigral

Radiology

are normally

with MSA, four showed increased or decreased signal, or both, in the putamen. Of the six cases of OPCA without MSA studied with the i.5-T unit, only one showed decreased signal in-

ele-

seen dedegen-

with

seven

3.

#{149}

arrows)

is perhaps the component of striatonigral degeneration. In our series, only 13 studies were

ments, as well as to cell loss and gliosis (Fig 5). Both of these featuresthe presence of neuromelanin and “hematin” pigments and the pres-

696

(open

of staining of transverse while the pyramidal

on the other hand, is an associated or constitutive finding in a high proportion of cases of ShyDrager syndrome (1). Therefore, what we see in the basal ganglia in

ment

in

in the pallidum has been reported in Shy-Dragem syndrome (1 i,i4,15). Increased signal in the putamen has

of iron

bellar peduncles Bugiani, MD.)

tensity

ate imaging support for the diagnosis of MSA. Since MSA is due to the combined degeneration of different systems, the expected MR findings should also involve different structures. Other investigators (1 i-13)

have

intensity with lack are poorly stained,

eration,

projec-

tions of the Purkinje cells to the dentate nuclei makes these nuclei become gliotic. However, the cells of the dentate nuclei are preserved and so are their projections to the red nuclei and the thalami through the supenior cemebellar peduncles (1). These peduncles are particularly well demonstrated on MR coronal sections, because their normal signal makes them visible against the degenerated background (Fig 4b).

A problem study is the

signal duncles

pontine

the superior cere(1) (Figs 3, 4, 6).

degeneration

Figure 6. (a) Intermediate-weighted MR image (SE 2,074/50) at level of pons and middle cerebellar peduncles in a patient with OPCA. (b) Similar histologic section stained for myelm (Woelcke-Heidenhain). Comparison demonstrates perfect correspondence of abnormal

1987; 249-271.

March

1990