Pekka U. Farin, MD • Heikki ]aroma, MD • Arvi Harju, MD • Seppo Soimakallio, MD
Shoulder Impingement Syndrome:
Sonographie Evaluation!
A method of shoulder sonography THE impingement (1-3) or painful
in whieh lateral and anterior eleva- arch syndrome (4) is a common
ti on of the arm is used during sean- eause of chronic shoulder pain and .
ning was demonstrated to be effec- disability. Impingement of the rota-
tive in eases of suspected Impinge- tor cuff beneath the coracoaeromial
ment syndrome. The value of the ligamentous arch without associated
method lies in its ability to dernon- rupture of the cuff is a well-estab-
strate fluid eollection in the sub- lished clinieal diagnosis. Entrapment
acromial-subdeltoid bursal system, of the soft tissues between the great-
with gradual distention of the bursa er tuberosity of the humerus and the
and lateral pooling of fluid to the coracoacromial ligamentous arch
subdeltoid portion while the arm is during abduction or elevation of the
elevated. In 102 of 381 patients stud- arm results in changes in the sub-
ied, surgieal diagnosis was available acromial bursa sueh as edema and
for eorrelation. Among this group. '. hemorrhage or fluid collection in the
there were seven faIse-negative and bursa and tendons (1,2). Impinge-
three false-positive sonographie ment syndrome is,thought to begin
findings. A eomparison of sono- in anatomically predisposed shoul-
graphie with surgical findings dem- ders and can progress to degenera-
onstrated a sensitivity of 81% and a tion and calcification witli.in the ten-
specificity of 95% in stages 1-111, dons and eventually to bone spur
and a sensitivity of 71% and a speci- formation at the undersurface of the
ficity of 96% in early stages land II acromion and the distal end of the
of the impingement syndrome. The clavicle. Subacromial bursitis, bicipi- .
results of dynamie shoulder sono- tal tendinitis, and rotator cuff disrup-
graphie examination with fluoro- tions irre common sequelae of this ab-
seopic radiography provide valuable normality.
information in patients with sus- Fundamental diagnostic imaging
pected early-stage impingement methods in this syndrome include
syndrome. fluoroscopy, arthrography, subacro-
mial bursography, tomography, and
in recent years, computed tomogra-
phy and magnetic resonance (MR)
imaging. Subacromial bursography,
in particular, has been suggested as
an accurate diagnostic modality in
evaluating this disorder (5,6); howev-
er, it has not r,ceived much attention
~)., .
The purpose of ~his paper is to de-
scribe the appropriate technique for
sonographie evaluation cf the im-
pingement syndrome and to discuss
the reliability of sonographie find-
ings. .
Index terms: Bursitis, 41.25 • Shoulder, ab-
normalities, 41.25, 41.48 • Shoulder, injuries,
41.48. Shoulder, US studies, 41.12981 .
Radiology 1990; 176:845-849
1From the Departments of Diagnostic Radi-
ology (f.U.F., S. S.) and Qrthopaedics and Sur-
gery (H.J., A.H.), Kuopio University Central
Hospital, PO Box 6, 70211 Kuopio, Finland. Re-
ceived December 5,1989; revision.requested
January 30, 1990; revision received April 17; ac-
cept~d April 24. Address reprint requests to
P.U.F.
@RSNA,1990
PATIENTS AND METHODS
Between Oetober 1985 and June 1988,
381 patients underwent bilateralshoulder
sonography followed by plain radiogra.
phy and arthrography at our radiology
department. In each ease, sonography was
performed by a radiologist (P.U.F.). Only
a few cases were videotaped. There were
221 (58%) menand 160 (42%) women,
with a mean age of 45 years (range, 18-71
years). Surgical eorrelation was available
for 102 patients (27%) (55 men and 47
women). The mean age of this latter
group was 42 years (range, 24-45 years),
The sonograms were obtained with a
real-time 7.5-MJiz linear-array scanner
(55D-280 L5; Aloka, Tokyo). Allsono-.
grams were interpreted prospeetively.
The examination teehnique for the rota-
tor cuff, intertubercular sulcus, or bieeps
tendon did not differ from that used in
earlier studies (8-11) and always included
statie and dynamie portions. Bieeps tendi-
nitis was diagnosed with sonography
when effusion surrounded the tendon in
the bicipital sheath.(9). Dislocation of the
tendon was diagnosed if the groove was
empty and the tendon was seen on the
lesser tuberosity in transverse scans of the
proximal humerus. All patients were seat-
ed during scanning, and the subacromial-
subdeltoid bursa was imaged with the
arm held in a neutral position (no inter-
nal or external rotation) and also in
maximal internal rotation, whieh was
achieved by holding the hand behind the
back. In the dynamic portion of thestudy,
the patient's arm was elevated primarily
laterally and then anteriorly, and at the
same time, attention was paid to the bur-
sal system, especially to the gradual dis-
tention ofthe bursa (Fig 1).
In 10 eases in which the bursa was dis-
tended with fluid, aspiration of fluid and
injection of corticosteroids into the bursa.
were performed (Fig 2). In five patients,
instead of injeetion of eortieosteroids, 4-6
mL of iohexol (Omnipaque; Nyeomed A5,
Oslo), with an iodine coneentration of.
240 mg/mL, was injeeted. Then burso- '
grams (Fig 3) were bbtained, at different
elevations of the arm, to verify sono-
graphie diagnosis of impingement syn-
drome. .
Two specifie criteria for the sonograph-
ie diagnosis of impingement syndrome
were developed. These were (a) fluid col-
lection in the subaeromial-subdeltoid
bursa and (b) fluid in the bursal system
. that gradually distended the bursa and
pooled laterally to the subdeltoid portion
while the arm was elevated. Nonspecifie
l',

Figure 1. Subdeltoid-subacromial bursa. (a) Longitudinal scan of the normal bursa (arrowheads) seen between the deep surface of deltoid
muscle (d) and rotator cuff (r), s = subcutaneous tissue, h = humeral head. (b) Fluid-distended bursa (0) in neutral position of the arm.
(c) Gradual distention of bursa and lateral pooling of the fluid in bursa in impingement syndrome while arm is elevated.
a.
b.
Figure 2. ' (a) Schematic diagram of the
technique used to aspirate fluid and to inject
centrast medium into the distended bursa.
(b) Note the tip of the needle in the bursa
(arrowhead) on this corresponding scan.
but suggestive findings were thick,ening
(Fig 4) or thinning (Fig 5) and hypo- or
hyperechogenicity of the rotator cuff
, without bursal fluid collection. These lat-
ter nonspecific findings were difficuit to
recognize, but comparison with the non-
affected shoulder helped in assessment of
asymmetry.
. RESULTS
The subacromial-subdeltoid bursa
is delineated by the boundary be-
846 • Radiology
tween the deltoid muscle and rotator
cuff and is poorly identified at so-
nography in its normal state. Sono-
graphically this space was depicted
as normal in 313 of 381 patients. Ar-
thrography and sonography revealed
107 and 92 rotator cuff tears, respec-
tively, in these 313 patients.
The sonographic appearance of the
bursal system was abnormal in 68
(18%) of 381 patients. Arthrography
showed normal cuffs in 38 of these
68. Thus, 30 patients with arthro-
graphically diagnosed rotator cuff
tears had fluid-containing bursal sys-
tems found only at sonography.
Inflamed and thiekened bursal sys-
tems were documented in 37 of the
102 patients who underwent surgery
(Table 1). In 30 of these 37, sonogra-
phy depieted fluid collection in the
subacromial-subdeltoid bursa, which
distended the bursa and pooled later-
ally with arm elevation. 'In two cases,
the fluid could be found onlywith
this elevation test. In addition to sev-
en false-negative findings at sonog-
raphy, there were three false-positive
findings in the 102 patients who un- ~
derwent surgery. The association be- .~
'a
b
I
d
1
t
o
t:
s.
c
a
tl
iJ '