Jean

M. Modi,

MS

Lowell

#{149}

A. Sether,

PhD

Articular Cartilage: Zones with Signal Zones

of high and low signal intenon magnetic resonance (MR) images of articular cartilage were correbated with the four histologic zones

Victor

found

in such

cartilage.

Grossly normal articular cartilage from knees and ankles of a fresh Cadaver were used in the study. The three zones identified on MR images included a low-intensity zone near the articular surface, a zone of higher signal intensity next to that, and a second zone of low intensity that was deep to the two others. The location of the superficial low-intensity zone corresponded to dense, tangentially oriented layers of collagen in the superficial histologic zone. Higher signal intensity deep to the superficial low-intensity zone correlated with cartilage in the transitional zone. The deep low-intensity zone correlated with a combination of deep radiate and calcified cartilage and cortical bone. Results of this study indicate that, with high resolulion, MR imaging may demonstrate three zones of differing signal intensity in articular cartilage. The superficial low-intensity zone may be a useful marker of the surface of normal articular cartilage. Index

terms:

Cartilage, studies, 452.121, 463.121 (MR), tissue characterization, Radiology

1991;

4521.121 #{149} Joints, MR #{149} Magnetic resonance 4521.121, 463.121

181:853-855

J.

MD

#{149}

H

correlation

ISTOLOGIC

high

and

bow

of zones

signal

Kneeland,

MD

field-of-view software were developed at our institution.) The imaging matrix was 256 x 256 for the sequences with the 500msec TR and 256 x 128 for sequences with a TR of 2,000 msec. The plane of imaging was perpendicular to the bone-cartilage interface. For preparation of histologic sections,

of

intensity

within articubar cartilage at magnetic resonance (MR) imaging has received limited study. To our knowledge, one study correlating signal intensity histologic zones

published formed

zones of different with biochemical and in cartilage has been

(1), and that study was perwith bovine tissues. In that

study,

in which

with

bong

spin-echo

repetition

sequences

times

(TRs)

We attempted intensity

to correlate

characteristics cartilage with of the cartilage.

ticulam zones

the

were initially 2-mm-thick

articubar the bone

cartilage and the cut surface of were used as landmarks on MR to correctly

acid,

AND

p.m with

with

cial

coil

METHODS

(Medical

Advances,

matrix,

and

3-mm

The

joints

were

then

the

medial

and

lateral

the distal

tibia,

and

of articular

section

Mil-

femorab

thickness.

field

the

same

of view,

the

proximal

cartilage

that

(The

1.5-T

and

solenoid

then

sec-

thick. Adjacent sections were safranin 0 (2,3) to evaluate

stained proteo-

and

with

tnichnome

zones

solution

5

(4) to

Accuracy of the histocorrelations was verioverall shape and size of the highin the

cartilage

and on

toweach

MR image were measured with the dcctronic cursors provided with the display package of the imaging system. The percentage of the total thickness occupied by each zone of differing signal intensity was calculated. In the histologic sections, the transitional,

deep

radiate,

and

calcified zones of cartilage were identified on the basis of conventional histologic criteria and were measured with a millimeter rule. Zone thicknesses on MR images were compared

with

thicknesses

of histologic

zones.

From

RESULTS

talus, were

On MR images of whole knee and ankle joints, three zones were distinguished in the amticular cartilage. One zone, characterized by bow signal in-

renor-

at visual inspection and palpation sampled for MR imaging-histologic correlation. A total of eight cylinders contaming cartilage and bone I inch in diameter were removed by using a hole saw blade on a hand drill. These cylinders were submerged in a 0.9% NaCl saline solution and imaged (500/48 and 2,000/70) at room temperature in a 2.5-cm solenoid with

and

condybes,

were

4-cm

in paraffin,

thickness.

disarticulated.

mat

coil

for

were

superficial,

waukee). Spin-echo 600/31 (TR msec/echo time msec) and 2,000/70 MR images were obtained with a 12-cm field of view, 256 x 256

cylinder

Sections

glycans

disease. MR images of the obtained at room temperature imaging unit (Signa; GE McdMilwaukee) with a commer-

extremity

with a coping Contours of the

the

a microtome.

evaluate collagen. logic-MR imaging fled by comparing of the tissue. The thicknesses

am-

Knee and ankle joints were harvested from the fresh cadaver of a 49-year-old man who died of carcinoma in the head and neck region. The man had had no history of joint joints were with a 1.5-T icat Systems,

orient

embedded

tioned

intensity

MATERIALS

cut slabs.

histologic sectioning. The 2-mm-thick slabs were fixed in 10% formatin, decatcifled with ethylene diamenetetraacetic

signal

of human the histologic

samples saw into

images

were

used, a zone of relatively high signal intensity correlated with the superficial and transitional zones in histologic sections, and a zone of bow signab intensity correlated with the deep radiate and calcified cartilage zone.

gions

From the Departments of Anatomy Q.M.M., LAS.) and Radiology (V.M.H., J.B.K.), the Mcdicat College of Wisconsin, Fnoedtert Memorial Lutheran Hospital, 9200 W Wisconsin Aye, Mitwaukee, WI 53226. Received November 19, 1990; revision requested January 14, 1991; revision received July 12; accepted July 22. Supported by a grant from Salutan, Sunnyvale, Calif. Address reprint requests to V.M.H. C RSNA, 1991

Bruce

Correlation of Histologic Intensity at MR Imaging’

sity

normally

M. Haughton,

#{149}

imaging

2-mm coil

system,

tensity with all sequences, was ent at the junction of cartilage

presand

subchondral bone. A band of camtilage superficial to this zone had moderately high signal intensity on a

section and

small

Abbreviation:

TR

=

repetition

time.

853

short tensity

TR images on long

and high TR images.

zone, characterized as a narrow of bow signal intensity on long short

TR

images,

was

three

tently short

TR

superficial sity

zones

or

identified

at the

were

consis-

on bong

and

(Fig

2). The

narrow

with

low

demonstrated

was

images

zone seen

consistently.

Results

band

surface of the cartilage (Fig 1). The thickness of the zones varied from region to region. For each region of articular cartilage sampled, there was an average two imaging-histologic correlations. For each of the eight cartilage-bone cylinders,

DISCUSSION

signal inA third

signal

of

The

transitional

inten-

A thicken

zone,

immedi-

ateby below the superficial zone, had round lacunae separated by regions of glycosaminoglycan-containing chondroid matrix that stained with safranin 0. The loose meshwork of poorly organized collagen fibrils of the transitional zone stained pabely with trichrorne solution. The third on deep radiate zone contained columns of lacunae and densely packed metachromaticalby staining collagen fibrils oriented perpendicular to the surface of the cartilage. This region stained consistently positive for glycosarninoglycan with safranin 0. The fourth or calcified cartilage zone was adjacent to the cortical bone. Colbagen, few lacunae, and minimal metachromatic staining were evident with trichnorne solution. Safranin 0 staining was variable in the calcified cantibage.

The

logic

sections occupied an average of cartilage thickness (range,

5% of the 3%-12%),

superficial

the

transitional

zone

in the

zone

histo-

occu-

pied an average of 42% of the cantilage thickness (range, 22%-68%), and the two deep zones (deep radiate and calcified cartilage) combined occupied an average of 53% (mange, 27%-72%). 854

#{149} Radiology

study

confirm

sults. The difference in the ing appearance is therefore

middle zone with medium to high signal intensity and a thick deep zone with low signal intensity were also identified (Fig 2). The superficial zone occupied an average of 16% of the cartilage thickness (range, 7%-45%), the middle zone occupied an average of 31 % of the cartilage thickness (range, 10%-75%), and the deep zone composed an average of 53% of the cartilage thickness (range, 17%-80%). The histologic sections that comebated with the MR images demonstrated normal articubar cartilage that was characterized by four zones (Fig 3). The superficial zone had compact tangentially oriented collagen fibribs that stained with trichnome solution. Staining with safranin 0 was only minimal. Elliptic lacunae were present.

of this

that

zones of differing signal intensity in anticubar cartilage can be identified at MR imaging in human beings. In the previous study of bovine articuban cartibage, only two zones were demonstrated. We identified three in human cartilage. There are no known histologic differences between bovine and human anticular cartilage that could explain the differing MR imaging me-

dent on technique. bovine cartilage, 2-3-mm 20-cm

MR irnagdepen-

In the study a 0.5-T irnager,

section thickness, field of view were

and used.

of a We

Figure

1.

12-cm

field

knee

from

Sagittal

MR image

of view,

256

a cadaver.

Three

with different signal intensities line the femur. The degenerated the dorsal aspect of the patella

tions

damage sustained from the joint.

and

chemical

shift

(5) affected

ens (1) have,

diate bates that bates fled (1), face

terized

by

superficial sponded ness,

that

a zone

with

interme-

to high signal intensity comewith the transitional zones and a bow-signal-intensity zone comewith the deep radiate and calcizones. In contrast to Lehnem et al we found a third zone at the sunof cartilage, which was chamac-

to th

low

signal

low-intensity in location, superficial

intensity.

The

zone comebut not in thickhistologic

zones

(arrowheads) cartilage on lacks the non-

mal zones.

Fluid

in the joint

sity

higher

than

that

of superficial

2.

MR

images

of the

Figure

of a

cartilage

used a 1.5-T imager, 4-cm field of view, and 2-mm section thickness in the cylinders and a 12-cm field of view and 3-mm section thickness in the intact joints. The three zones were more effectively shown with the smaller field of view used in imaging of the cylinders. The superficial bow-intensity zone on the MR images corresponded in location, although not exactly in thickness, to the superficial zone in the histologic sections. The zone of intermediate to high signal intensity in the MR images corresponded approximately to the transitional zone in histologic sections. The broad band of low signal intensity near the bonecartilage interface in the MR images corresponded approximately to the deep radiate and calcified cartilage zones and the adjacent subchondrab bone. The MR imaging evaluation of the deepest zone was affected by technique. In preliminary experiments, we found that the direction of phaseand frequency-encoding directhe appearance of the cartilage-bone interface (Fig 4). MR imaging correlates of the four histologic zones have been studied previously. The major difference between our study and that of previous investigators is the number of zones distinguished in anticular cartilage with MR imaging. We found, as oth-

(2,000/70,

x 128 matrix)

has signal

inten-

cartilage.

cartilage-bone

cylinders. Medutlary bone (b) and cartilage (arrows, arrowheads) are evident in each sample. A water standard in a tube (s) is placed within a hole in the samples (left image, 600/48; right image, 2,000/70). The high signal intensity between the cartilage and the subchondnal bone is an artifact caused by

zone.

in removal

Consistent

of the cylinder

demonstration

of the

superficial zone appears to require high spatial resolution and minimal partial volume averaging, which was obtained in our study during the ex-

perimental and calcified guished

acquisitions. cartilage

Cortical were not

bone distin-

the

cartilage

zones

from

deep

with MR imaging. What determines signal intensity in different zones of cartilage was not resolved by this study. Four histologic zones in normal articular cartilage are differentiated primarily by the orientation

nae

of collagen

(6,7).

composition

In femoral

fibers

and

cartilage,

is as follows:

the

lacu-

the superficial

December

1991

a.

c.

b.

Figure 3. (a) MR image (2,000/70), (b) tnichnome-stained section, and (c) safranin 0-stained section of a cartilage specimen. In a, two zones of low signal intensity (arrows) are separated by a zone of high signal intensity. In b, four zones of cartilage are evident. The superficial zone (s) is characterized by densely staining collagen, staining in the transitional zone (t) is paler, the deep radiate zone (d) has metachromatic staining, and staining in the calcified cartilage (c) is dense, but not metachromatic. Note tidemark (arrow) and disregard artifact (a in b and c). Below the calcified cartilage nal magnifications,

is subchondnal xl.6).

bone.

In c, safranin

0 stains

the

second,

third,

and

fourth

zone

concentration (1), but the differences in water concentration are slight (82% for the superficial and middle zones vs 76%

for

the

compared sity.

two

with

Factors

tnation

deep

those

other

must

than

in cartilage.

For the

the frequency-encoding gradient is oriented parallel to the axis of the water tube; for the right image, the gradient is perpendicular to

the tube. the deep

A marked change low-signal-intensity

in thickness of zone (arrows)

as a result chondral of cartilage

of chemical shift is evident. cortical bone and the deepest are not distinguishable.

Subzones

water

therefore

the explanation for intensity in cartilage.

of the zones

zones)

(6)

in signal

intenconcen-

be evaluated the

as

varying signal The orientation

sections

versus

5 p.m

for

1.

2.

his-

of cartilage

4.

used

in MR

imaging may be different from that on the anatomic sections. The bonders between zones in the histologic seczone, 10%; transitional zone, 40%; deep radiate zone, 40%; calcified zone, 10% (6). The distribution of waten, collagen fibrils, on proteoglycans in the cartilage corresponds inexactly to the three zones (8). The low signal

superficial explained

of signal

intensity

intensity in the bayer of cartilage may be by a bow proteoglycan con-

tions

are

arbitrarily

marrow of medublary to the subchondnal

defined. bone

The fat adjacent

cortical

bone

in

causes a chemical shift artifact that may obscure portions of the cortical bone on adjacent cartilage. The boundary

between

the

deepest

zones

lagen

study

fails

bation

of MR signal

tbe among in signal

been

zones intensity

attributed

varies

lit-

(7). The differences in cartilage have

to differences

in water

therefore

tobogic pemficial

not

be expected.

to show

zone, zone

181

#{149} Number

3

While

one-to-one

intensity

it does show that can be delineated

high-resolution techniques. Further study is indicated the superficial low-intensity

Volume

7.

of cartilage and subchondral bone, both of which have bow signal intensity, is difficult to determine on MR images. Exact correlation between MR images and histologic sections can

which

5.

6.

centration, but this would not apply to the bow signal intensity in the deep layers of cartilage (9). Signal intensity does not appear to be affected by cob-

concentration,

this come-

with

the

superficial

normal

from

zone

(onigi-

identifying in clinical needed to superficial

distinguish-

abnormal

cartilage.

References

3.

sections

not

The biochemical explanation for the low signal intensity in the superficial zone needs clarification. #{149}

tobogic sections. Also, the border between the transitional zone and adjacent zones undubates. The effect of regional variations in the four zones on

but

which may be useful for the surface of the cartilage MR imaging. A study is determine whether the zone could be a marker ing

of collagen fibers may affect signal intensity by means of its effect on magnetic susceptibility. There are some technical explanations for imperfect correlation between zones of signal intensity and the histologic zones. The section thicknesses differ: 2,000 p.m for MR imaging

of cartilage,

his-

Lehnen

KB, Rechl

Heuck

AF, Lukas

9.

GmeinwiesenJK,

Structure,

function, degeneration of bovine hyaline cartilage: assessment with MR imaging in vitro. Radiology 1989; 170:495-499. Harris R, Wesbey G. Artifacts in magnetic resonance imaging. In: Knessel HY, ed. Magnetic resonance annual. New York: Raven, 1988; 71-98. Kivinanta I, Mankku T, Jurvetin J, Saamanen AM, Helminen HJ. Fixation, decalcification, and tissue processing effects on articutan cartilage proteoglycans. Histochemistry 1984; 80:569-573. Kivinanta I, Mankku T, Jurvelin J, Saamanen AM, Helminen HJ. Microspectrophotometric quantitation of glycosaminoglycans in articulan cartilage sections stained with safranin 0. Histochemistry 1985; 82:249-255.

Gabe M; Blackith

RE, trans.

Selective

stain-

ing of collagen fibres. Histological techniques. New York: Springer-Verlag, 1976; 21 1-218. Meachim G, Stockwetl RA. The matrix. In: Freeman MAR, ed. Adult articutan cartilage. 2nd ed. 1979; 1-16. Akeson WH, Gershuni DH. Articulan cartitage physiology and metabolism. In: Resnick

D, Niwayama,

8.

HP,

HP, Kohl HP.

eds. Diagnosis

of bone and

joint disorders. 2nd ed. Philadelphia: Saundens, 1988; 758-764. Muir H. Pnoteoglycans as organizers of the intercellular matrix. Biochem Soc Trans 1983; 11:613-622. Stockwell RA, ScottJE. Distribution of acid glycosaminoglycans in human articular cantilage. Nature 1967; 215:1376-1378.

a suwith

to verify zone,

Radiology

855

#{149}