Musculoskeletal Philippe David

Kindynis,

J.

MD2 #{149} Jorg Haller, MD #{149} Debra Trudell,

Sartoris,

Osteophytosis Radiologic,

terms:

Arthritis,

degenerative, 45.77 45.76 #{149} Knee, abnormali#{149} Knee, arthritis 45.77

Chrondocalcinosis, ties, 45.77, 45.76

Radiology

#{149} Heung

Sik Tyson,

#{149} Rose

Kang, MA

of the Knee: and Pathologic

Although the radiologic manifestations of degenerative disease of the knee have been investigated, the distribution of marginal and central osteophytes has not been defined. This study included (a) 50 consecutive patients with osteoarthritis of the knee in whom routine and specialized projections were obtained prospectively, (b) 25 patients with calcium pyrophosphate dihydrate (CPPD) crystal deposition disease whose knee radiographs were retrospectively reviewed, and (c) four cadaveric knees that were dissected to assess pertinent anatomy. In this study the importance of the tunnel view in the evaluation of osteoarthritis and CPPD crystal deposition disease is demonstrated, the distribution of and the relationship between marginal and central osteophytes are discussed, and two new radiologic signs are described. As both marginal and central osteophytes may simulate intraarticular bodies, the recognition of these outgrowths is of clinical importance. Index

MD RA

#{149}

1990; 174:841-846

O that

leads

#{149} Donald

is a degenerative

of the

articular

to remodeling

cartilage processes

the Department

J.H., H.S.K., ministration

of Radiology

(P.K.,

DR., D.J.S., D.T.), Veterans Medical Center, University

Adof Cal-

ifornia Medical Center, 3350 La Jolla Village Dr, San Diego, CA 92161 and the Museum of Man (R.T.), San Diego. Received June 20, 1989; revision

requested

ceived

October

ported

in part

grant

July

26; final

31; accepted by Veterans

no. SA306.

Address

revision

November

re-

8. Sup-

ease, whose

bearing increased and other changes

bone underlying the damaged cartilage, (c) grinding down of exposed subchondral bone, (d) formation of bone overgrowths at the margins of the articular cartilage (marginal osteophytes), at the bonecartilage interface (central osteophytes), or both; and (e) variable but usually mild synovial inflammation (1,2). On radiographs, findings of joint-space narrowing, marginal osteophyte formation, bone sclerosis, and absence of bone atrophy strongly support the diagnosis of osteoarthritis

(2).

There

is a growing consensus that osteoarthnitis represents a heterogeneous set of diseases affected by the interaction of multiple risk (such as genetic susceptibility;

factors the

anatomic, biomechanical, and biochemical changes associated with aging; and previous injury) producing a common pathway of disease. The traditional classification of degenerative disease

into

primary

(idiopath-

requests

to

DR.

Current address: gie, H#{244}pital Cantonal 2

Geneva. C

RSNA,

1990

tions

ing

of degenerative

the

knee

(4-7), to our distribution

osteophytes

disease

have

affect-

been

investigated the precise of marginal and central has not been previously knowledge,

defined.

Departement Universitaire

de Radiobode Gen#{232}ve,

PATIENTS Our study adult patients knees]) with

AND included (46 men,

knee

pain

three

METHODS

were

women

25 adults

[33 knees])

ible with dihydrate

chondro-

by heavy

crystal

calcium (CPPD)

(22 men,

with

characterized

tate and linear

punc-

aggregates

compat-

pyrophosphate crystal deposition

knee

radiographs

dis-

(AP, later-

al, and tunnel views obtained) were retrospectively reviewed. Patients whose knee radiographs demonstrated jointspace narrowing, osteophytes, or both, without chondrocalcinosis or other

known

disorder

thritis,

were

group.

such

as rheumatoid

included

ar-

in the osteoarthritis

Four cadaveric

knees,

two with

osteoarthritis and two with CPPD crystal deposition disease, were dissected to assess pertinent anatomy and to illustrate the presence of marginal and central osteophytes.

In patients

with

crystal deposition tion of each femur

osteoarthnitis or disease the distal was assessed for and bone sclerosis

CPPD porosteoby

phyte formation two of the authors working together (P.K., J.H.). The presence of marginal osteophytes was recorded at each condyle (lateral, posterior, and intercondylar areas) and at the superior margin of the patellar articular surface of the femur (Fig 1). As this

to define

superior

margin

on plain

ment

of degenerative

tellar

compartment

was

radiographs,

difficult

the assess-

changes

in the

pa-

on osformation at the superior and margins of the patella. Osteowere graded as 0 = absence of os-

teophyte

inferior

phytes teophyte, 1 3 = proliferative

phyte,

was also based

small

beakbike

osteophyte,

or mushroomlike

and 2

=

osteo-

intermediate-stage those of grades

osteo-

phyte, between 1 and 3 (Tables 1, 2). The location and dimensions of central osteophytes, when present,

were

also recorded

mined joint-space eral radiographs,

(Table

3). We deterand, on lat-

narrowing the presence

of margin-

a! osteophytes in the intercondylar area seen as a line parallel to the condyles and continuous 2a, 2b).

Administration

reprint

studied

frequent loads; in the

ic) and secondary types must be interpreted cautiously, as the former designation may indicate only our inability to define important causes (3). Although the radiologic manifestaFrom

by. Also calcinosis,

in areas (b) sclerosis

MD

evidence of osteoarthritis of the knee in whom anteroposterior (AP), lateral, and tunnel views were obtained prospective-

including marginal or central osteophyte formation. Pathologic changes consist of (a) erosion of portions of the articular cartilage, which is more

joint

1

Resnick,

Anatomic, Investigation’

STEOARTHRITIS

disease

MD

Radioloi

with

the

patellar

groove

(Fig

50 consecutive four

women

and radiographic

[55

Abbreviations: =

calcium

AP

pyrophosphate

anteroposterior,

CPPD

dihydrate.

841

1

/ ..‘ .

:,

-:

:

--:.

,

.1

.5

.

i

“

,..

t :

-I .

.

‘1

b.

a. Figure

1.

Attachments

drocalcinosis and margins (arrows)

of the synovial

(b) transverse that represent

membrane

Table

In the osteoarthnitis group intercondylar osteophytes were present all but two knees. In one, only the patellan compartment was abnormal, and in the other, only an osteophyte at the lateral margin of the medial

was

noted.

In the

intercondylar

curned

as the

mality

in

842

Radiology

#{149}

single

femora

of (a) sagittab attachments

section

of the

of cadaveric

synoviab

knee

membrane

with

at the

chonarticubar

occur.

and

Central

Marginal Osteophyte Distribution

Osteophytes

Grade

in Patients

with

Osteoarthritis

1

Grade 2

Grade 3

Total

13 0 1

0 0 0

0 0 0

13 0 1

4 7 0 5

0 6 0 12

0 0 0 7

4 13 0 24

30

18

7

55

4

4

2

10

One compartment Medial

Lateral Patelbar

oc-

osseous abnon(27%) of the osteoarthnitis group and in two of the femora (6%) in the CPPD group (all of grade 1). In all groups, the number of compartments involved increased with the grade of the marginal osteophytes (Tables 1, 2). Subsequently, these marginal osteophytes formed a ridge around the articular surface. 15

showing

of Marginal

in

CPPD

osteophytes

Photographs

1

Distribution

crystal deposition disease group, intercondylar osteophytes were present in all but three knees; in these, chondrocalcinosis was present without marginal or central osteophytes. Considering all three compartments, when intercondylar osteophytes were the only osseous abnormality, they occurred in the medial compartment in nine femora (16%) of the osteoarthnitis group and in no femur in the CPPD group. Considering only the medial and lateral compartments, medial

margins.

section of cadaveric knee with osteoarthritis sites at which marginal osteophytes can

RESULTS

condyle

at the articular

Two compartments Medial/lateral Medial/patebbar Lateral/pateblar Three compartments Total femora with marginal osteophytes

No. of femora central Note-Mean

with

osteophytes age

of patients,

54.9 years.

Medial unicompartmental or bicompartmental patterns were observed more commonly in the osteoarthnitis group. Bicompartmental (medial and patellar) and tnicompartmental changes were found in both groups, more frequently in the CPPD group, probably because of the more advanced stages of disease that were studied. When present, bone sclerosis and joint-space narrowing were found mostly in the compartment

where the marginal osteophytes were the most severe. Central osteophytes were present in both groups, representing 1 1 lesions in 10 femora in the osteoarthritis

group

femora

and in

3). Although were phytes,

number

phytes

found the

nine

without

seven

(Tables

marginal

osteo-

between

with

grade

1-

osteophytes

correlation

the

in

group

no central

of femora

and

lesions

CPPD

the

the

central

of the

osteo-

marginal March

1990

Table 2 Distribution of Marginal Chondrocalcinosis

and Central

Osteophytes

in Patients

marginal remnants

with

Distribution

Grade 1

Grade 2

Grade 3

Total

0

0

0

0

0 6

1 0

0 0

1 6

0

0

0

0

2

14

7

23

8

15

7

30

1

5

1

7

One compartment Two compartments Medial/lateral Medial/patelbar Lateral/pateblar Three compartments Total

femora

No.

osteophytes

of femora

central Note-Mean

age of patients,

osteophytes

Table

3

are

not

66.2 years.

included

Three

of

entral

Location

Patients

Patients

with

with

Osteo-

Chondro-

calcinosis

arthritis

Posterior

medial

condyle

3

Posterior

3

lateral

condyle MTP LTP Other

2 2

3 1

1

1

3

1

Total

11

9

Note.-MTP = area between the tibial and patellar articular surfaces of the medial condyle, Li? - area between the tibial and patellar articular surfaces of the lateral condyle.

osteophytes was not significant in the osteoarthnitis and CPPD groups. In the osteoanthnitis group, central osteophytes were found in four of the 30 femora with grade 1 marginal osteophytes (13%), four of the 18 femora with grade 2 (22%), and two

of the

seven

femora

with

grade

3

zone

knees

with

chondrocalcinosis

but without

marginal

or

marcated

by a faint

that

rests

on

the

lateral

fully marks dial routine cavity into ments and tached oral,

extended. A similar groove the articular surface of the mecondyle but is seen less often on lateral radiographs (10). The of the knee joint is divided three communicating compart(medial, lateral, and anterior), the synovia! membrane is ataround the margin of the ferntibia!, and patellar articular sur-

faces

the

peripheral

(8,9).

when

At these

or notch border

of

the joint

margins,

the

is

an-

ticular cartilage is continuous with the synovial membrane and the penosteum. Although these marginal areas represent the sites at which margina! osteophytes develop (Fig 1), the role of the synovial membrane and peniosteum is not known. The fibrous capsule, which is not attached to these margins, is not involved in marginal osteophyte formation (2). The pathogenesis of central osteophyte formation is quite similar to that of marginal osteophytes (2); in-

The

central

deed,

central

flat

outgrowths

with

marginal

ized

small

represented

appearing

bumps

as local-

on the

niphery

articular

surface continuous with the underlying bone. The distribution of central osteophytes in the two patient groups was similar (Table 3).

On lateral arthritis and resenting intercondylar the femora 40% of the

100% Volume

radiographs

CPPD marginal

of the 174

area with femora

was

visible

grade with

femora Number

#{149}

of the

osteo-

groups, a line reposteophytes of the

in 10% of

1 osteophytes, grade 2, and

with

grade 3

3.

osteophytes

can

osteophytes

of the

merge

at the

articulation

(Fig

pe-

2). At

histologic examination, both represent outgrowths from the subchondral bone. With marginal osteophytes, subchondral hypervascularity at the periphery of the articular

cartilage

leads

adjacent

cartilage

to calcification

to new

endochondral

and

of the

bone

forma-

tion. These marginal partially covered by

osteophytes fibrocarti!age,

peniosteum,

With

or both.

growth,

border”

(2,6,1

calcification

of the

are

1). In

of the

cartilage

above In terms

deeper

is subsequently by osseous the calcified

it. of distribution

partments

involved,

tis-

of the

the

corn-

results

of

our study of the osteoarthnitis and the CPPD groups do not differ from those of other investigators (4,5,7).

The

two

groups

demonstrated

a corn-

mon pattern that included marginal ridging of osteophytes with in-

creased

grade

phytes and tercondylar

Ridging

of margina’ the presence and central

of marginal

osteoof both osteophytes.

in-

osteophytes

has

been thnitis pattern

described in cases of osteoar(4), and we found an identical in our CPPD group. Intercondylar osteophytes were present on nearly all femora and thus appear to be an early finding in degenerative disease (Fig 3). In a previous study, based on analysis of weight-bearing AP, lateral,

and

axial

patellar

views

of 130 pa-

tients with primary osteoarthnitis 107 patients with other disorders patients with rheumatoid arthritis), Altman et al (12) found that radio-

and (55

graphic evidence of osteophytes, particularly in the medial compartment, best differentiated osteoarthritis from other disorders. These investigators also found that cornbining the findings phytes altering

of medial improved specificity

of osteoarthnitis,

and lateral osteosensitivity without in the diagnosis

and

the

finding

of

joint-space narrowing provided no higher specificity than did the finding of osteophytes alone. As we demonstrated that intercondylar osteophytes may be the only type of margina! osteophytes that are evident and necessitate use of a tunnel view

for detection, specialized

subsequently

covered

cartilage is only partially resorbed and the final result is a layer of new bone between the original calcified cartilage zone and the newly formed

To understand the development of marginal and central osteophytes, anatomic considerations are necessary (8-10). The patellar and the tibia! sunfaces of the lateral condyle are de-

meniscus

surface

traoc-

Subchondral hypervasstimulates endochondral osa phenomenon termed and reduplication of the

resorbed and replaced sue; in reduplication,

one

groove

subchondral osteophytes

articular

shifting,

(28%). In the CPPD group, central osteophytes were found in one of the eight femora with grade 1 marginal osteophytes (12%), five of the 15 fernora with grade 2 (33%), and one of the eight femora with grade 3 (14%).

osteophytes

cartilage.

cartilage-bone

here.

Osteophytes

in the

by

sification, “shifting

DISCUSSION

Distribution

cur

cularity

with

osteophytes

central

with the adjacent beculae (2). Central

with

marginal

car-

tilage, and, finally, their spongy bone becomes entirely continuous

Marginal Osteophyte

osteophytes leave behind of the original calcified

we believe that projection should

routinely osteoarthnitis.

in patients with In addition,

view

provide

rnay

valuable

this be used

suspected the tunnel

inforrna-

Radiology

843

#{149}

c.

tion

concerning

for

the

may

reveal

space

the

cruciate

more

narrowing

erab

or

in

the

(13),

traarticular

Prominent #{149} Radiology

phytes diographs

4),

jointmedial

than

do

and

may

or

routine be

between

osteophytes osseous

condyles lat-

radiograph-

differentiation

intercondylar

844

the

weight-bearing

ic projections

available (Fig

prominent in

compartment

views

space

ligaments

useful large

and

bodies.

intercondybar

osteo-

in-

patellar

are

readily as

and groove

seen

a line

on

parallel

to

continuous (Fig

lateral with

2a,

2b).

ra-

eral

the

the

surface

Although

this finding is not a reliable indicator of early intercondylar osteophytes, might be a useful sign for the radiobogist who does not routinely use a tunnel view. The anterior and postenor cruciate ligaments are attached distally to a fossa in front of and bat-

to the

anterior

a depression Therefore, it

posterior

of the as

tibia, these

tibia!

spine to

the

and

respectively ligaments

in

articular

(14). are

not

attached to the tibiab spines, sharpening of these structures does not represent an enthesopathy but is secondary to formation of marginal osteophytes at the tibial condvles. These sharp tibial spines can rub against the intercondylar osteophvtes and March

1990

I Figure 2. Demonstration strates marginal osteophytes intercondylar osteophytes marginal osteophyte ginal osteophyte on

was not visible

of marginal and central osteophytes in a cadaveric knee with chondrocalcinosis. of the femur and the patella (arrows). Note a line parallel to the condyles (arrowheads). (b) Radiograph of a sagittal section through the intercondylar

(arrowheads). the lateral

on a lateral

(c) Radiograph

of a sagittal

section

of the condyle (small arrow), in a, and the barge osteophyte

margin

radiograph

of the

a small on

the

medial

condyle

demonstrating

central

osteophyte

superior

aspect

of marginal and central osteophytes in this area (barge arrow). (d) Radiograph of a sagittal groove (arrows). Note also the appearance of the osteophytes on the inferior aspt!ct of the ular surface of the femur (arrowheads), which result from merging of marginal arid central

(a) Lateral radiograph demonfemur representing prominent demonstrating prominent

of the region

an

on the articubar of the

posterior

oblique

condyle

section showing central patebla and the superior osteophytes in these

I Figure

3.

of the

mar-

(arrowheads), resulting

which

from

merging

osteophytes of the aspect of the patellar areas.

Benefit

tients with (a) AP view

section

surface

of the

degenerative

tunnel

bone

patelbar artic-

view

in pa-

changes.

demonstrating slight medial joint-space narrowing without other abnormality. (b) Tunnel view demonstrating an intercondylar osteophyte of the medial condybe (arrow).

I

.p

. ..

I

b. Figure

4.

Intercondylar

osteophyte

graph of a distal femur demonstrating bar osteophyte filling the anterior

of a patient phyte groove

patient

(arrows)

ate ligament medial

osteoarthritis

as an additional (area between

in another phytes

with

condyle

(*).

and part

its relationship

demonstrates

excrescence continuous arrows). (c) Transa.xial

with

osteoarthritis

to the cruciate

a marginal osteophyte of the intercondylar

the anterior with computed

demonstrating

the

with groove

part

and

of the

osteo-

normal intercondybar (CT) section of a knee

patellar

surrounding the posterior

(a) Photo-

intercondy(b) Tunnel view

of the intercondylar

contour tomographic

femoral

with prominent intercondylar osteophytes Note also a small central osteophyte on (arrowhead).

ligament.

a prominent (arrows).

marginal

the articular

anterior surface

osteocruciof the

) S.

1’ A

become flattened (Fig 5). This is an important consideration because a flattened tibial spine on an AP view is an indirect sign of intercondylar marginal osteophytes. Central osteophytes can occur on any articular surface covered by cartilage-including those of the tibia, femur, and patella-and may appear as

Volume

174

Number

#{149}

3

buttonlike, bumpy, or flattened osseous irregularities. Although no central osteophytes were found without marginal osteophytes, central osteophytes can form shortly after the appearance of marginal osteophytes. In the femur, they specifically occur where weight-bearing pressure is lowest during the usual gait cycle (15,16): the posterior surface of the condyles (Fig 4c) and between the tibia! and patellan articular surfaces of the femur, around the femoral notches (Fig 6). Central osteophytes of the patellar groove, as found in one of the cadavenic knees (Fig 2d), are not seen on plain radiographs. In

more advanced cases of osteoarthnitis, these excrescences can occur in the weight-bearing areas (Fig 5) because of the modified loading forces, but they are never evident in areas where the cartilage is completely eroded (2). Central osteophytes may be difficult to recognize on routine radiographs. A small central osteophyte superimposed on a prominent margina! osteophyte (particularly in the posterior condyles), central osteophytes of the patellar groove, and very flat central osteophytes may not be seen on plain radiographs (Fig 2c). Because central osteophytes can be

Radiology

845

#{149}

4.

Ahlback

S.

Osteoarthrosis

radiographic

Diagn 5.

,

1968;

Resnick

deposition

pseudogout. 6.

Radiol

C. Goergen

TG,

radiographic in calcium

dihydrate

disease

D, Niwayama

Thomas iel

RH,

C.

uation

Figure

5.

the articubar outgrowths flattening

AP

view

surface

of an

osteoarthritic

of the femur

and the underlying of the tibial spines

knee

(arrows).

condybe secondary

demonstrating

The presence is the clue to prominent

multiple

central

of continuity

to the correct intercondylar

osteophytes

between

of

the bone

diagnosis. Note osteophytes

8.

dalities.

Radiology

Warwick delphia:

9.

heads).

delphia: 10.

I Figure 6. Central osteophyte of the condybar notch area. Prominent central osteophyte at the bevel of the lateral condylar notch (arrowhead). Note also marginal osteophytes of the patella and the superior aspect of the patebbar articular surface of the femur (arrows).

tielle 11.

misinterpreted ous bodies, osteophytes

as intraanticulan the recognition has significant

importance.

The

osseof central clinical

presence

of

14.

Jaffe

HL.

flammatory

Philadelphia:

Metabolic, diseases

degenerative of bones

Lea & Febiger,

and

joints.

1972;

735-

of the

R, Asch of criteria

for 3rd

reporting

tion

of osteoarthritis

ed.

Le

I, Kitamura

N,

1986;

Resnick

D, Vint

Radio! et a!.

joint:

in the elderly 164:411-414. D, et a!. the

de-

adult. Devel-

classification

of osteoarthritis:

Rheum

surgeons. Phila-

profil du differen-

J

for

Phila-

733-859.

articulaires.

E, Bloch

and

ed.

discovertebral

opment

classifica-

of the

knee.

Arthritis

29:1039-1049. V.

The

of cartilage

Girgis

knee.

“tunnel” loss

view

in

in osteoarthri-

Radiology

FG, Marshall cruciate

anatomical, analysis. 15.

1980;

137:547-

P. Insall patellar

Maquet Femorotibial

in-

JL, Monajem

ligaments

of the

functional Clin Orthop

Aglietti A new

Geriatr and

surfaces

Anatomy

420-503.

1982;

change 1987;

plication.

2.

eds. British

1980;

Altman

The

16.

Clin

FL, 36th

548.

References Brandt KD. Osteoarthritis. Med 1988; 4:279-293.

116:585-594.

67:727-735. J, Yamamoto

tis of the

between the central osteophytes and the underlying bone and the presence of ossification rather than calcification should facilitate their correct identification (Fig 5). U ity

1.

rno-

1975; body.

assessment

continu-

a corn-

Maldague B. anatomic radiologique

generative Radiology

13.

knee:

J,

plate

Daneval-

diagnostic

WH. Anatomy back and limbs.

des

1986; Aoki End

12.

of the

Lippincott,

Malghem genou:

NP,

of available

Saunders,

Hollinshead Vol 3, The

Phila-

Compartmental

R, Williams human

of

2nd ed. 1364-1479.

of osteoarthritis study

122:1-15.

Diagnosis

D, Alazraki R.

parative

of the

also the (arrow-

Resnick

D, Greenfield

et

(CPPD):

1977;

bone and joint disorders. delphia: Saunders, 1988;

7.

a

and pathologic pyrophosphate

Radiology

Resnick

knee:

Acta

277(suppl):7-70.

D, Niwayama

al. Clinical, abnormalities

#{163}

I

of the

investigation.

perimental Surg [Am]

JN. Orthop

Van

de

PS,

weight-bearing determination. 1975; 57:766-770.

Trent

P.

and

ap-

design 1975;

Berg

joint:

experimental 106:216-231.

Walker

prothesis:

Clin PG.

and 1975;

ARS. knee

AJ,

107:175-187. Simonet areas:

J Bone

JC. cxJoint

778.

3.

846

Radiology

#{149}

Davis MA. Epidemiology tis. Clin Geriatr Med 1988;

of osteoarthri4:241-255.

March

1990