Niels

Egund,

MD

#{149} Hans

Legg-Calve-Perthes Imaging with

Wingstrand,

Disease: MR’

Magnetic resonance (MR) imaging studies were obtained to evaluate the shape of the cartilaginous femoral head and the adequacy of femoral acetabular containment in 35 children with Legg-Calv#{233}-Perthes disease (LCPD). MR results for 24 of the 35 children were compared with results at arthrography. The shape of the cartilaginous femoral head was well identified with MR imaging, but, with arthrography, less information could be obtained about the medial and lateral aspects of the cartilaginous capital epiphyses in nine and three cases, respectively. Thirteen children underwent surgery; remodeling and regained femoroacetabular congruity of the articular surfaces following derotation varus osteotomy was identified in seven of 10 who underwent repeat MR imaging. In addition, MR images and conventional radiographs of the hips were compared. The radiographically identifiable distribution of the bone necrosis of the capital epiphysis correlated well with that demonstrated on MR images. MR imaging is a valuable noninvasive procedure with which to obtain information about the status and treatment of the cartilaginous femoral head in children with LCPD. Index Bones,

terms: necrosis,

Radiology

Bones, MR 44.443

1991;

studies,

44.1214

#{149}

179:89-92

I From the Department of Radiology, Odense University Hospital, DK-5000 Odense, Denmark (N.E.); and the Department of Orthopedics, University Hospital, Lund, Sweden (H.W.). From the 1989 RSNA scientific assembly. Received December 4, 1989; revision requested January 30, 1990; revision received No-

vember

27;

accepted

print requests to N.E. .e RSNA, 1991

December

3. Address

MD

D

in obtaining early madiognaphic prognosis is a major clinical problem in the diagnosis and treatment of children with LeggCalv#{233}-Perthes disease (LCPD). Several attempts have been made to identify early prognostic signs of LCPD (14), but these signs usually appear too late to facilitate effective treatment, which is aimed at containment of the femomal head in the acetabulum (5). If surgery is considered, it is important to evaluate the shape of the cartilage, the congruence between the femoral head and the acetabulum, and the containment of the femoral head within the acetabulum. Until recently, arthrography was the only method available for such evaluation (6-9). However, arthrography is an invasive procedure and, in most hospitals, is undertaken with the patient under general anesthesia; thus, anthrographic studies are not well suited for repeat examinations. Magnetic resonance (MR) imaging has great potential for enabling the assessment of bone and joint disorders (10). Concerning the treatment of avascular necrosis of the femonal head, most reports concentrate on early detection (11-13) and assessment of the extent of the disease (1418) rather than on evaluation of epiphyseal shape and containment (19,20). The purpose of this study was to compare the potential of MR imaging with conventional arthmography for the pre- and postoperative assessment of the shape and containment of the cartilaginous femonal head in children with LCPD. Also, the extent of bone changes, which had previously been described only in the adult osteonecrotic hip (21,22), were evaluated. IFFICULTY

PATIENTS

AND

METHODS

Twenty-nine children with unilateral LCPD and six with bilateral disease (for a total of 41 abnormal and 29 normal hips) were evaluated with both conventional radiography

and

MR to assess

the

shape

and containment of the femoral head within the acetabulum. One patient who underwent two MR examinations because of movement artifacts was not included in the study. The examinations were mepeated in 26 of the children, with a time interval

child

of from

3 to

12 months.

One

examined four times; six, three times; 19, twice; and nine, once, for a total of 69 examinations. Twenty-five hip joint was

arthrographies

were

performed

in

con-

junction with the MR examinations in 24 of the children. Twelve children underwent derotation varus osteotomy and one underwent valgus osteotomy following these examinations. The children ranged in age from 4 to 1 1 years (mean, 7.3 years) at the time of the first MR examination. The five girls and 30 boys represented all stages of LCPD, but most were in the stages of condensation and fragmentation at the time of their first examination. MR imaging was performed with a 0.3T permanent-magnet imager (Beta 3000; Fonar, Melville, NY), which produced coronal views with Ti-weighted spinecho sequences, with a repetition time (TR) of 400-600 msec and echo time (TE) of 16-30 msec (TRITE 400-600/16-30; most commonly, 500/28). Section thickness was 5 mm. Images were obtained from a flexible solenoidal surface receiver coil of appropriate diameter. T2-weighted coronal images (2,000/56) were obtained in 22 children at their first examination. Use of this pulse sequence was discontinued because the images had a high noiseto-signal ratio and contributed little to the information obtained with the Tiweighted sequences. In addition, the light sedation required for all but two patients was not sufficient to keep the childmen still during the prolonged examination time required for the T2-weighted sequences in five of 22 examinations. Coronal images were obtained with the

re-

Abbreviations:

LCPD

=

Legg-Calv#{233}-Perthes

disease,

TE

echo

time,

TR

repetition

time.

89

1_

Figure 1. MR image of left hip of 8-yearold child with LCPD (Ti-weighted image; TR/TE = 500/28). Osseous manifestations of early LCPD include localized low signal intensity at the center of the capital epiphysis; the medial portion is almost unaffected, while a small fragment laterally demonstrates intermediate signal intensity. The contour of the cartilaginous femoral head is normal, as is the contour of the cartilage of the acetabulum (arrow), but the lateral edge of the acetabular labrum is not identified.

hips

in

tion bag

was that

less

than

(Marchall

by

Medico;

One

of three

was

also

(350

of the left hip

with

mg

of iodine

per

Nycomed,

the hip tion-inward

in

extension, rotations. obtained

of containment tion

head.

through

Each

at the same superimposed

hips

but

head and

were from the

the

a femonal

that

opacity

maining

hips.

All portions

of the

epiphysis

that

demonstrated

in-

contour

of ob-

of the femoral pair of tracings was compared degree of magnification while on each other.

center

According to results at radiographic examination of the 41 hips, 25 were in the stage of condensation and fragmentation, 12 were in the stage of pure fragmentation, and four were in a late stage of restitution at the time of the first MR examination. Among the 25 hips in the stage of #{149} Radiology

opacity

in the

low

signal

intensity

imaging.

Those

that

demonstrated

normal

opacity

had

had at MR

imaging

intermediate

sig-

is apparent

In three

opacity within

hips,

Among

the

12 hips

the stage of pure of the remaining epiphysis ography

had but

ate signal

a mixed mentation,

the stage

showed

that

of

mixed

were

low

or

at MR.

more

hips,

edge well

at radiintermedi-

Portions strucsig-

hips

that

of condensation, and/or healing,

mixed

low

were MR

to inter-

in fragim-

the

of the

than

medial

(Figs

of the defined

arthrogra-

remaining of the

aspect

crani-

at

of the because medium only

2, 4a).

The

(Fig

4a) but A bulging

varied

at

lateral

acetabular labrum at arthrography

MR.

hips

was in 21 of

in only

16 of 25

prominence,

in size

from

0.5 to 1 cm,

was identified at the inferior aspect of the metaphysis in eight hips (Figs 2a, 4b). A similar structure was seen at the lateral aspect of the epiphysis in five hips (Figs 2a, 4b). The signal intensity

of

sponded

to that

lage.

four

clearly

MR imaging

which

in

of the

without bone intermediate

prominent

femoral head was invisible of pooling of the contrast and, thus, was demonstrated

24

most

opacity

lateral

4a). In hips, MR

bulging

(Fig 2a, 2b). In the advanced flattening

hips

normal

intensity

Among

and

fragmentation, fragments

mixed

of the epiphysis tunes demonstrated nal intensity.

ages

pattern

was more varied the epiphysis.

and

demonstrated

nine

medially,

the

was images

al portion of the head was visualized more clearly on the MR images. In

demonstrates nearly normal In six hips there was also a area that demonstrated normal

opacity.

cranial

hypertrophic

cartilage

phy hip,

to the

cartilaginous head on MR

in 20 of 24 hips (Fig of the four remaining

lateral

normal high signal in1). Usually the central

fragment

which opacity. lateral

regard

aspects three

intensities. the

of the femoral to that seen

with

(Figs 1-3). Two hips small medial frag-

with (Fig

mediate signal At arthrography, contour identical

me-

at radiography

uniformly

small

derota-

made from the the MR images

in two

portion of the epiphysis demonstrated greater opacity. In Figure 3a, a

degree

head

at radiography

the

varying

ments tensity

RESULTS

90

opacity

nal intensity demonstrated

dy-

and abduclatter projec-

femoral

with

osteotomy. of the cartilaginous

the femoral anthrograms tamed

neutral, The

of the

Joint

with fluoroscoobtained with

to evaluate

be possible

varus Tracings

milliliter)

Oslo).

and fragmentation, demonstrated uniformly

high

creased performed in the opwith the patient under genwith use of 3-5 mL of io-

were investigated radiognaphs were

would

abduction-in-

was

namics py, and

was

condensation epiphysis

rotaplastic knees

Denmark).

examinations

(Omnipaque;

tion

neutral

a vacuum around the

Faaborg,

performed

ward rotation. Anthrognaphy erating room, eral anesthesia, hexol

5#{176} flexion;

maintained was wrapped

1-igure . ta MK image or iett nip ot lu-year-old cruiO with LL1’L) i i-weighted image; TR/TE = 500/28). Osseous lesions include low-signal-intensity lesion of the epiphysis, with a small fragment medially that demonstrates slightly decreased signal intensity (small black arrow). There is generalized cartilaginous hypertrophy (large black arrow), with a bulging prominence laterally (curved white arrow) and an adjacent defect in the cartilage (arrowhead). A structure of intermediate signal intensity on the medial aspect of the metaphysis (straight white arrow) may represent localized cartilaginous hypertrophy surrounded by a synovial recess of lower signal intensity. (b) Arthrogram of left hip of the same child. Arrows indicate the same structures seen in a. Identification of pooling of the contrast material in the cartilaginous defect laterally (arrowhead) was not possible without the MR image. The medial aspect of the cartilaginous femoral head is not visible.

The

shape

and

these

prominence during

thrography

and recess,

corre-

adjacent

carti-

maintained

position

metaphysis synovial

lesions

of the

relative

its

to

the

abduction was which

covered was

at amby also

April

a iden-

1991

.s..

.

.,

.‘.

‘

Figure

3.

Radiographs

opacity centrally, tabular roof by the acetabulum.

of left hip of 9-year-old

surrounded the denotation

by areas of normal vanus osteotomy.

stage

of healing.

Spherical

femur, and At follow-up

me-

mation while vanced capital

femoral

uated

without

with

epiphysis,

which

the performance osteotorny. Ten deformity

one

the

reeval-

follow-up

examinations; two of these strated significant deformity months later, respectively,

contraindicated

of deroof 17

were

or two

MR 11

of

DISCUSSION The aim of clinical and radiographic examinations of children suffering from LCPD is early identification of those 25% of children who need

Volume

179

Number

#{149}

the

head of the may be surgi-

1

suitable

appears

with

denotation

varus

pre-

and

for

other

serial

hand,

sedation

tial

On and

surgery

allowed

in the of our

treatment

patients

the

exclusion

because

of

ma-

it dem-

onstrated either acceptable containment of the head of the femur on advanced deformity of the head. Two of 10 patients who demonstrated normal containment of the femur at the first MR examination formity of the femoral short time; therefore,

developed dehead in a all new LCPD

patients in our hospital will undergo one MR examination at admission and, if containment is good and sur-

and

high

indicated,

follow-up

ex-

will be scheduled after 6 months, and 1 year to early progressive deforma-

3

of MR

imaging

to enable

postop-

erative evaluation of the results of derotation varus osteotomy, with the goal of containing the hypertrophic cartilage of the head of the femur unden the acetabular roof (Fig 4). Spherical remodeling, observed within a year after surgery, cannot be demonstrated with conventional radiography

only

of the

is not

surface

covered under the acethe femonal head and

tion of the head of the femur, which would indicate a need for surgery. We have demonstrated the poten-

for MR examinations, is required,

has been between

aminations months, identify

postopera-

examinations.

of the superior

epiphysis obtained

gery

or an innominate osteobe performed (23-25). MR

allows

flattening

lateral portion of the congruity has been

one patient in our series of Tiweighted sequences had to be excluded from the study because of movement artifacts on the MR images. It appears from our results that MR

jority

surgery.

treatment to contain femur. Such treatment

a femoral

examination

demon5 and which

performance

MR

only con-

epiphysis

(b) The osteotorny,

tive assessment of containment of the femoral head and enables visualization of the medial aspect of the femoral head. Since anthrography is an invasive procedure and may require the use of general anesthesia, it is not the

of the head of the femur, nine hips demonstrated addeformity of the cartilaginous

traindicated tation varus hips

hips and 17 hips defor-

(a) The

has proved valuable in the determination of the cartilaginous shape of the femoral head (20), and we have demonstrated its advantages when compared with invasive arthrography;

one was unchanged. examination of 22 pa-

tients who had 26 abnormal did not undergo surgery, demonstrated no significant

LCPD.

osteotomy tomy can

modeling was observed postoperatively in seven of 10 repeat MR examinations following the surgeries (Fig 4b-4d). Two hips showed progressive deformation of the head of

the

with

or varying opacity. (c) Two years after

cal; either

tified on the MR images in three cases. Derotation varus osteotomy was performed early in 12 hips; valgus osteotomy was performed in one hip

at the

child

except

in late

tion. The

distribution

crosis

as seen

stages

with

of restitu-

of epiphyseal

ne-

conventional

radi-

ography correlated well with areas of decreased signal intensity at MR imaging in all stages of the disease. The central portion of the epiphysis is most

affected

by

osteonecrosis,

which leaves viable periphery. A similar sient

ischemia

tral

portion

has

been

tions

Rush structure

or edema

of the value

is still

cen-

epiphysis

recently

(26).

of these

being

in the of tran-

in the

capital

reported

prognostic

fragments location

The

observa-

evaluated.

et a! (20)

noticed

at the

infenomedial

space in seven patients ed that this represented inflammatory lesion. similar abnormalities

a papillary joint

and suggesta synovial We observed at the medial

Radiology

91

#{149}

Figure

4.

(a) Left

hip anthrogram

of same

patient

as in Figure

3. With

slight

abduction-inward

rotation,

the pooling

of

contrast

material

de-

lineates the proximal and lateral aspects of the cartilaginous epiphysis but not its medial aspect. The arrowhead indicates the lateral edge of the acetabular labrum. (b-d) MR images of left hip of same patient (Ti-weighted images, TR/TE = 500/28). (1,) The central portion of the epiphysis has decreased signal intensity, which corresponds to the area of increased opacity at radiography in Figure 3a. Flattening of the superior surface of the hypertnophic cartilaginous epiphysis is present. The lateral edge of the acetabular labrum is not visible. A bulging structure with low to intermediate signal intensity appears at the medial aspect of the metaphysis (arrow). The cartilaginous shape of the femonal head was identical to that seen at anthrography except for the medial aspect of the head. (c) One year after denotation vanus osteotomy, osseous femoral head, the cartilaginous

healing is indicated which still has slight capital epiphysis

by higher flattening has been

aspect of the metaphysis and at the lateral aspect of the epiphysis. We believe that the signal intensity of the Ti-weighted image was too high to represent capsule or synovia but mesembled that of cartilage. At amthrography,

these

lesions

a synovial recess that follow the metaphysis amination. If osteotomy in such

cases,

these

were

covered

by

was found to at dynamic exis considered structures

signal intensity of its superior achieved by the

6.

7.

8.

9.

should

be subjected to histopathologic analysis to determine their nature. On the basis of our results, MR imaging has been found to be a valuable noninvasive procedure with which to obtain information about the status and treatment of the cartilaginous femomal head in children with LCPD.

10.

1 1.

12.

U

Crawford AH, Carothers TA. Hip arthrography in the skeletally immature. Clin Orthop 1982; 162:54-60. Gallagher JM, Weiner DS, Cook AJ. When is arthrography indicated in LeggCalv#{233}-Perthes disease? J Bone Joint Surg [Am] 1983; 65:900-905. Goldman AB. Evaluation of disorders of children, adolescents, and adults without prostheses. Radiol Clin North Am 1981; 19:329-348. Quain 5, Catterall A. Hinge abduction of the hip: diagnosis and treatment. J Bone Joint Surg [Br] 1986; 68:61-64. Sartoris DJ, Resnick D. MR imaging of the musculoskeletal system: current and future status. AJR 1987; 149:457-467. Bluemm RG, Falke TH, Ziedses des Plantes BC Jr. Steiner RM. Early LeggPerthes disease (ischemic necrosis of the femoral head) demonstrated by magnetic resonance imaging. Skeletal Radiol 1985; 14:95-98. Elsig JP, Exner CU, von Schulthess GK, Weitzel M. False-negative magnetic resonance

2.

3.

4.

5.

Catterall Perthes

A. The natural history of disease. J Bone Joint Surg [Br] 1971; 53:37-53. Danielsson L, Pettersson H, Sunden C. Early assessment of prognosis in Perthes disease. Acta Orthop Scand 1982; 53:605611. Dickens DR. Menelaus MB. The assessment of prognosis in Perthes disease. Bone Joint Surg (Br] 1978; 60:189-194. NevelOs AB, Colton CL, Burch PR, Woodward PM. Perthes disease: a study of radiological features. Acta Orthop Scand 1977; 48:411-421. Catterall A. Legg-Calv#{233}-Perthes disease. Edinburgh: Churchill Livingstone, 1982.

13.

14.

Pinto MR. Magnetic

15.

Radiology

Peterson resonance

of Legg-Calv#{233}-Perthes

Mitchell

1989;

MD,

HY,

necrosis

of the

scintigraphy.

Thickman Magnetic necrosis

hip:

20.

EB,

21.

22.

24.

1986;

147:67-

Magnetic

resonance

diseases. D, Axel

Clin L, Kressel

resonance of the

diol 1986; 15:133-140. Turner DA, Templeton AG,

HY,

imaging

femoral

Petasnick

26.

RE.

detection

Mag-

of carti-

avascular

necrosis:

correla-

trends

in

Perthes

of results. 67:399-401.

in orthopedics.

staging, findAG, Lon-

1972; 122-147. RK. Innominate disease:

J Bone

os-

a radiological

Joint

Surg

[Br]

Salter RB. Legg-Perthes disease: treatment by innominate osteotomy. AAOS [American Academy of Orthopaedic Surgeons]

Instructional

Moshy, NT, Singer children

Course

1973; 309. %VS, Bartal

Lectures.

E.

Hip

1986;

Pay three

et al.

abnormal findings on MR images. ogy 1989; 171:147-149.

imaging

Orthop

head

don: Butterworth, Maxted MJ, Jackson

Louis: JA.

disease:

Modern

survey 25.

Bechtold

tion of MR imaging. radiographic radionuclide imaging, and clinical ings. Radiology 1987; 162:709-715. Catterall A. Coxa plana. In: Apley

teotomy

di-

of MR.

AJR

LA,

laginous and synovial change with MR imaging. Radiology 1 988; 167:473-476. Beltran J, Burk JM, Herman U, et al. Avascular necrosis of the femoral head: early MRI detection and radiological correlation. Magn Reson Imaging 1987; 5:431-442. Mitchell DC, Rao VM, Dalinka MK, et al. Femoral

Avascular

comparison

Koman

Calv#{233}-Perthes

ed.

ME,

of

capital osteonecrosis: MR findings of difuse marrow abnormalities without focal lesions. Radiology 1989; 171:135-140. Totty WG, Murphy WA, Ganz WI, Kumar B. Daum WJ, Siegel BA. Magnetic resonance imaging of the normal and ischemic femoral head. AJR 1984; 143:12731280.

1985;

L.

cartilaginous remodeling

netic resonance imaging of pediatric hip disease. J PediatrOrthop 1985; 5:665-671. Rush BH, Bramson RT, Ogden JA. Legg-

Orthop

Steinberg

of the spherical

Toh’

disease.

A, Axel

CT, and 71. Powers

HL,

and remodeling vanus osteotomy, = 600/30).

19.

9:19-22.

Kundel

Alavi

18.

23.

of Legg-

HA, Berquist TH. imaging in earls’

Orthop

in marrow 206:79-85. 16.

stage

J Pediatr

Pediatr

Rosenberg

#{149}

in early disease.

agnosis

Kressel

17.

92

imaging

Calv#{233}-Perthes 1989; 9:231-235.

References 1.

at the periphery of the capital epiphysis surface. (d) Two years after denotation osteotomy (Ti-weighted images, TR/TE

accompanied

by

St.

pain

in

transient

Radiol-

of avascular

head.

AC, JP.

Skeletal

Seizer

Ra-

PM,

Femoral

April

1991