AJR 2000; 174:1717-1722
© American Roentgen Ray Society
Sonography of the Glenoid Labrum
A Cadaveric Study with Arthroscopic Correlation
Mihra S. Taljanovic1,2,
Kara Lubben Carlson1,3,
John E. Kuhn4,
Jon A. Jacobson1,
Lydia O. Delaney-Sathy1 and
Ronald S. Adler1,5
1
Department of Radiology, University of Michigan Medical Center, 1500 E.
Medical Center Dr., TC 2910, Ann Arbor, MI 48109-0326.
2
Department of Radiology, University of Arizona HSC, 1501 N. Campbell Ave.,
Tucson, AZ 85724-5067.
3
Radia Medical Imaging, Evergreen Hospital Medical Center, 12040 N.E. 128th
St., Kirkland, WA 98034.
4
Section of Orthopaedic Surgery, University of Michigan Health System, 24 Lloyd
Wright Dr., Box 0363, Ann Arbor, MI 48106.
5
Department of Radiology and Imaging, Hospital for Special Surgery, 535 E. 70th
St., New York, NY 10021.
Received August 9, 1999;
accepted after revision November 11, 1999.
Presented at the annual meeting of the American Roentgen Ray Society, New
Orleans, May 1999.
Address correspondence to J. A. Jacobson.
Abstract
OBJECTIVE. We assessed the usefulness of sonography in evaluating
the glenoid labrum in cadaveric specimens using arthroscopy as a standard of
reference.
MATERIALS AND METHODS. Eighty labral quadrants in 20 cadaveric
shoulders were examined by two musculoskeletal radiologists using 5- to 7-MHz
linear and curvilinear transducers. Agreement was reached by consensus. After
sonography, arthroscopy was performed by an experienced orthopedic surgeon.
Each labral quadrant was classified at the time of sonography and arthroscopy
as normal, degenerated, or torn.
RESULTS. Concordance between sonography and arthroscopy was 86%
(69/80 quadrants). In differentiating abnormal labrum (tear or degeneration)
from normal labrum using sonography, sensitivity was 63%, specificity was 98%,
positive predictive value was 94%, negative predictive value was 86%, and
accuracy was 88%. In differentiating labral tears from other labral conditions
(degeneration or normality), sensitivity was 67%, specificity was 99%,
positive predictive value was 67%, negative predictive value was 99%, and
accuracy was 98%.
CONCLUSION. Sonography has a promising role in the evaluation of the
glenoid labrum, particularly in excluding labral tears when the labra appear
normal on sonography. Further studies are required using normal and
symptomatic patients to determine the usefulness of sonography in the
diagnosis of labral abnormalities.
Introduction
In the evaluation of shoulder pain, sonography is a reliable noninvasive
imaging method and is effective in evaluating the rotator cuff
[1]. However, structures other
than the rotator cuff such as the glenoid labrum are also important in the
evaluation of shoulder pain. Current evaluation of the glenoid labrum focuses
on MR imaging with or without the use of intraarticular contrast
administration [2]. In the
diagnosis of labral tears, a sensitivity of 93% has been reported with routine
MR imaging and a sensitivity of 96% with MR arthrography using intraarticular
gadolinium solution [2]. To our
knowledge, the only studies describing the sonographic evaluation of the
glenoid labrum are published in European literature
[3,4,5,6,7,8,9,10].
Many of these studies are descriptive observations with limited pathology or
lack a standard of reference. Given the relative low cost and noninvasive
nature of sonography, we evaluated the effectiveness of sonography in
diagnosing labral abnormalities in 20 cadaveric specimens using arthroscopy as
a standard of reference.
Materials and Methods
Transcutaneous sonography of the glenoid labrum was performed in 20
cadaveric shoulder specimens (15 men and 5 women; age range, 59-94 years; mean
age, 81 years) by two experienced musculoskeletal radiologists using 5- to
7-MHz linear and curvilinear transducers (Spectra Plus; Diasonics, Santa Cruz,
CA) and ample sonographic transmission gel. The examiners were unaware of
clinical information related to the specimens. The cadaveric shoulders were
disarticulated from the thorax at the medial border of the scapula and the
sternoclavicular joint. All specimens were nonembalmed, frozen, and then
thawed before examination. Institutional review board approval was obtained
before we started our study.
Sonography consisted of both static and dynamic examinations in the
transverse plane. The glenoid labrum of the left shoulder was arbitrarily
divided into four quadrants (Fig.
1A): posterosuperior (12- to 3-o'clock position), posteroinferior
(3- to 6-o'clock position), anteroinferior (6- to 9-o'clock position), and
anterosuperior (9- to 12-o'clock position). Specimen positioning for the
sonography was as follows (Figs.
1B,1C,1D):
anterosuperior with external rotation, anteroinferior with abduction and
external rotation, posterosuperior, and posteroinferior with internal
rotation. At sonography, each labral quadrant was classified by consensus as
normal, degenerated, or torn. A labral quadrant was classified as normal if
the labrum was triangular or round in shape positioned adjacent to the
glenoid. A labral quadrant was designated as degenerative if the labrum was
irregular in contour. A labral quadrant was classified as torn if the labral
tissue was detached from the glenoid (aided by dynamic imaging to distract a
torn labrum from the glenoid) or not visible on sonography. The attachment of
the long head of the biceps tendon to the superior labrum was not assessed in
this study. Labral echogenicity was recorded as hyperechoic (equal to bone
cortex) (Fig.
2A,2B,2C),
hypoechoic (less echogenic than bone cortex and homogeneous), or heterogeneous
(less echogenic than bone cortex but not homogeneous). During sonography,
three images of each labral quadrant were obtained.
View larger version (22K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A. —Labral quadrant designation and sonographic transducer placement.
Arrows indicate sonographic imaging plane or level of transducer placement.
Drawing shows articular surface of glenoid divided into four quadrants.
|
|
View larger version (108K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B. —Labral quadrant designation and sonographic transducer placement.
Arrows indicate sonographic imaging plane or level of transducer placement.
Photograph shows patient position for sonographic evaluation of anterosuperior
(AS) quadrant in axial plane with arm abducted.
|
|
View larger version (134K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1C. —Labral quadrant designation and sonographic transducer placement.
Arrows indicate sonographic imaging plane or level of transducer placement.
Photograph shows patient position for sonographic evaluation of anteroinferior
(AI) quadrant in axial plane with arm abducted and externally rotated.
|
|
View larger version (116K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1D. —Labral quadrant designation and sonographic transducer placement.
Arrows indicate sonographic imaging plane or level of transducer placement.
Photograph shows patient position for sonographic evaluation of
posterosuperior (PS) and posteroinferior (PI) quadrants in axial plane with
arm internally rotated.
|
|
View larger version (129K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A. —The normal glenoid labrum. G = glenoid, H = humerus. Transverse
sonogram of anteroinferior shoulder in cadaveric specimen from 85-year-old
woman shows normal triangle-shaped hyperechoic labrum (arrows).
|
|
View larger version (116K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B. —The normal glenoid labrum. G = glenoid, H = humerus. Transverse
sonogram of posterosuperior shoulder in cadaveric specimen from 80-year-old
man reveals normal triangle-shaped hyperechoic labrum (arrows).
|
|
View larger version (124K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2C. —The normal glenoid labrum. G = glenoid, H = humerus. Transverse
sonogram of posteroinferior shoulder in cadaveric specimen from 85-year-old
woman shows normal triangle-shaped hyperechoic labrum (arrows).
|
|
After sonography, arthroscopy was performed from a posterior approach by an
experienced orthopedic surgeon with expertise in shoulder arthroscopy and
sports medicine. The surgeon classified the glenoid labrum as normal,
degenerated, or torn. All arthroscopic examinations were videotaped and the
examiner was blinded to clinical history and sonographic results. The
sonographic and arthroscopic results were then compared.
Results
At arthroscopy (Table 1), 56
of 80 labral quadrants were normal (anterosuperior = nine, anteroinferior =
17, posterosuperior = 12, posteroinferior = 18). Included in this group was a
single normal variant occurring in the anterosuperior labral quadrant where
there was absence of the labrum with cordlike thickening of the middle
glenohumeral ligament (a Buford complex)
[11]. Twenty-four labral
quadrants were abnormal, including three labral tears (anterosuperior = one,
anteroinferior = one, posterosuperior = one), and 21 degenerated labra
(anterosuperior = 10, anteroinferior = two, posterosuperior = seven,
posteroinferior = two).
At sonography (Table 1), 55
of 56 normal labral quadrants were correctly diagnosed (Fig.
2A,2B,2C).
One normal labral quadrant was misdiagnosed as a labral tear (anteroinferior)
(Fig. 3). Twelve degenerated
labra were correctly diagnosed at sonography
(Fig. 4) and nine were
misdiagnosed as normal (Fig.
5) (anterosuperior = six, posterosuperior = three). No
degenerative labra were misdiagnosed as labral tears. Of three labral tears,
two were correctly identified at sonography (anterosuperior = one,
anteroinferior = one) (Fig.
6A,6B,6C)
and one was misdiagnosed as labral degeneration
(Fig. 7) (posterosuperior).
The overall concordance of sonography and arthroscopy was 86% (69/80). The
percent concordance for each labral quadrant was as follows: anterosuperior
was 70% (14/20), anteroinferior was 95% (19/20), posterosuperior was 80%
(16/20), and posteroinferior was 100% (20/20).
View larger version (132K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3. —Normal labrum misdiagnosed as labral tear in cadaveric specimen from
81-year-old woman. Although labrum appeared normal on arthroscopy (not shown),
transverse sonogram of anteroinferior glenoid labrum shows no identifiable
labral tissue between glenoid (G) and humerus (H).
|
|
View larger version (123K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4. —Degenerative glenoid labrum in cadaveric specimen from 73-year-old
man. Transverse sonogram of posteroinferior glenoid labrum shows irregular
contour (arrow) of hyperechoic glenoid labrum, representing
degeneration at arthroscopy (not shown). G = glenoid, H = humeral head.
|
|
View larger version (148K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5. —Degenerative labrum misdiagnosed as normal in cadaver of 84-year-old
man. Transverse sonogram of posterosuperior glenoid labrum shows
normal-appearing labrum (arrows) proven degenerative on arthroscopy
(not shown). G = glenoid, H = humeral head.
|
|
View larger version (97K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6A. —Glenoid labrum tears in cadaveric specimen from 89-year-old woman. G
= glenoid, H = humerus. Transverse sonogram of anteroinferior shoulder shows
irregular hyperechoic anteroinferior labrum (arrows) displaced from
adjacent glenoid.
|
|
View larger version (116K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6B. —Glenoid labrum tears in cadaveric specimen from 89-year-old woman. G
= glenoid, H = humerus. Correlative arthroscopic image of A confirms
displaced anteroinferior labrum tear (arrows) or Bankart lesion.
|
|
View larger version (130K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6C. —Glenoid labrum tears in cadaveric specimen from 89-year-old woman. G
= glenoid, H = humerus. Transverse sonogram of anterosuperior shoulder shows
irregular and mixed hyperechoic and hypoechoic labral tissue (arrows)
displaced from glenoid.
|
|
View larger version (116K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 7. —Labral tear misdiagnosed as degeneration in cadaveric specimen from
74-year-old man. Although labral tear was revealed on arthroscopy (not shown),
transverse sonogram of posterosuperior glenoid labrum shows apparent irregular
hyperechoic posterosuperior labrum (arrows) adjacent to glenoid
(G).
|
|
We were able to differentiate labral abnormality (tear or degeneration)
from normal labrum at sonography with a sensitivity of 63% (15/24),
specificity of 98% (55/56), positive predictive value of 94% (15/16), negative
predictive value of 86% (55/64), and accuracy of 88% (70/80). We were able to
differentiate labral tears from other labral conditions (degeneration or
normal) at sonography with a sensitivity of 67% (2/3), specificity of 99%
(76/77), positive predictive value of 67% (2/3), negative predictive value of
99% (76/77), and accuracy of 98% (78/80).
We also recorded the echogenicity of the glenoid labrum at sonography. Of
56 normal labral quadrants confirmed at arthroscopy, 44 (79%) appeared
hyperechoic, seven (13%) hypoechoic, and four (7%) heterogeneous in echo
texture. One normal labral quadrant was not visible at sonography and its
echogenicity was not recorded. The one normal variant representing a Buford
complex appeared hyperechoic at sonography
(Fig. 8). Of 21 degenerative
labral quadrants confirmed at arthroscopy, seven (33%) appeared hyperechoic,
four (19%) hypoechoic, and 10 (48%) heterogeneous at sonography. Nine of 21
degenerative labral quadrants were misinterpreted at sonography as normal,
appearing hyperechoic in seven and hypoechoic in two. Twelve degenerative
labral quadrants were correctly interpreted at sonography, appearing
hypoechoic in two and heterogeneous in 10. Of three labral tears at
arthroscopy, two were heterogeneous in echo texture and one was hyperechoic.
The one labral tear misdiagnosed as degenerative appeared heterogeneous in
echo texture.
View larger version (152K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 8. —Buford complex in cadaveric specimen from 85-year-old man.
Transverse sonogram of anterosuperior glenoid labrum shows hyperechoic tissue
(arrows) misinterpreted as normal glenoid labrum. Hyperechoic tissue
likely represents thickened cordlike middle glenohumeral ligament with absence
of anterosuperior glenoid labrum (Buford complex). G = glenoid, H = humeral
head.
|
|
Discussion
In the late 1980s and 1990s, several authors in European literature
demonstrated that the glenoid labrum could be visualized by sonography
[3,4,5,6,7].
The normal glenoid labrum is identified as a hyperchoic triangular structure
located adjacent to the glenoid
[8]. More recently, Schydlowsky
et al. [9,
10] published two articles
describing the sonographic appearance of the glenoid labrum. One of these
studies revealed an 88% sensitivity and 67% specificity in diagnosing labral
lesions (using arthroscopy as a standard of reference); however, the
sonographic criteria used for diagnosing the labral lesions in this study were
not defined [9]. The other
study used cadaveric specimens and revealed that sonography can differentiate
normal and degenerative labra; however, labral tears were not present in this
study [10]. We attempted to
further define the role of sonography in detecting labral abnormalities.
Our results reveal an 86% concordance between sonography and arthroscopy in
the diagnosis of normal, degenerative, and torn labra. When we separated the
labrum into individual quadrants, percent concordance indicated that
sonography in the inferior quadrants (95% concordance in anteroinferior and
100% in posteroinferior) was better than sonography in the superior quadrants
(70% concordance in anterosuperior and 80% in posterosuperior).
To calculate the sensitivity and specificity of sonography in detecting
labral abnormalities, the 3 x 3 table
(Table 1) was condensed to a 2
x 2 table by combining labral tears and degeneration, and then by
combining labral degeneration and normal labra. Our results reveal that
sonography performed best when differentiating labral tears from other labral
conditions (labral degeneration or normality), with a sensitivity of 67%,
specificity of 99%, negative predictive value of 67%, positive predictive
value of 99%, and accuracy of 98%.
Echogenicity of the labrum was evaluated with sonography. Of 56 normal
labral quadrants at arthroscopy, 79% were hyperechoic, 13% hypoechoic, and 7%
heterogeneous in echo texture. Of 21 degenerative labral quadrants at
arthroscopy, 33% were hyperechoic, 19% hypoechoic, and 48% heterogeneous in
echo texture. Of three labral tears, 67% appeared heterogeneous and 33%
hyperechoic in echogenicity. Most of the normal labra were hyperechoic;
however, the appearance of degenerative and torn labra was more variable. We
cannot comment on the significance of altered labral echogenicity because we
did not perform histopathologic analysis.
Discordant results between arthroscopy and sonography include a normal
labrum in the anteroinferior quadrant that was diagnosed as a tear at
sonography (Fig. 3). Even in
retrospect, a normal labrum could not be identified on the printed image. The
examiners probably failed to identify the normal labrum and misinterpreted the
findings as a labral tear. Another discordant result included a labral tear in
the posterosuperior quadrant that was misdiagnosed as a degenerative labrum
(Fig. 7). In this case, the
area of the labral tear was probably overlooked. The remaining discordant
results included nine degenerative labra that were misdiagnosed as normal
(Fig. 5). We speculate that
the irregular contours of the degenerative labra were not visible at
sonography.
Identification of the normal labral variant was not prospectively made with
sonography. A Buford complex (absence of the anterosuperior labrum with a
cordlike middle glenohumeral ligament) was diagnosed at arthroscopy in one
labral quadrant [11]. This
quadrant was interpreted as normal at sonography
(Fig. 8). In retrospect, it
appears that the hyperechoic thickened middle glenohumeral ligament was
positioned adjacent to the glenoid, simulating labral tissue at sonography.
Moreover, absence of joint distention with fluid or contrast material can
cause difficulty in defining normal labral variants such as the Buford complex
and sublabral foramen. Based on this single case, it is unclear if sonography
can reliably differentiate normal variants from normal or abnormal labra.
Further studies may help clarify this matter. In evaluating the glenoid labra,
visualization of the anterosuperior labral quadrant was difficult because of
the close approximation of the coracoid process. This difficulty may have
affected the concordance rates in the anterosuperior region. Identification of
the long head of the biceps tendon insertion on the superior labrum and
identification of the glenohumeral ligaments was not attempted in this study.
The purpose of this preliminary study was to assess the usefulness of
sonography in visualizing the glenoid labrum and revealing labral tears.
Our study had several limitations. The small number of specimens, including
the low number of labral tears (3/80 quadrants), is a limitation.
Additionally, performing sonography on relatively inflexible cadavers was
difficult; this could have significantly contributed to the incomplete
characterization of the labrum in some locations, and we do not know whether
the freezing and thawing of the specimens affected the soft-tissues and their
imaging findings. The advanced age of the specimens with the large number of
degenerative labra may have also affected our results. In a younger
population, the fraction of echogenic labra may be much larger, enabling
better differentiation of labral abnormality. The inherent operator dependence
and learning curve associated with sonography is always present. As a
limitation, this was not assessed with multiple examiners and interobserver
variability was not calculated. Lastly, we did not assess the usefulness of
sonography in differentiating normal labral variants from normal labra or
labral tears.
In summary, sonography has modest sensitivity (67%) in differentiating
labral tears from other labral conditions (normality or degeneration). The
high specificity and negative predictive values (99%) suggest that if the
labrum appears normal at sonography, degeneration may exist but a tear can be
excluded. However, given the low number of labral tears in this study and the
use of cadaveric specimens, one should be cautious until further studies using
healthy and symptomatic subjects corroborates these data. This cadaveric study
reveals the potential of sonography in evaluating the glenoid labrum.
References
-
Wiener SN, Seitz WH. Sonography of the shoulder in patients with
tears of the rotator cuff: accuracy and value for selecting surgical options.
AJR
1993;160:103
-107[Abstract/Free Full Text]
-
Chandnani VP, Yeager TD, DeBerardino T, et al. Glenoid labral
tears: prospective evaluation with MR imaging, MR arthrography, and CT
arthrography. AJR
1993;161:1229
-1235[Abstract/Free Full Text]
-
Jerosh J, Castro WH, Jantea C, et al. Possibilities of sonography
in the diagnosis of instabilities of the shoulder joint [in German].
Ultraschall Med
1989;10:202
-205[Medline]
-
Estergalyos J, Bohar L, Kerekes V. Ultrasonic examination of soft
tissue injuries in the shoulder region [in Hungarian]. Magy
Traumatol Orthop Helyreallito Seb
1992;35:27
-30[Medline]
-
Jerosh J, Marquardt M. Sonographic diagnosis in shoulder joint
instability: possibilities and limitations [in German].
Orthopade
1993;22:294
-300[Medline]
-
Hinzmann J, Behrend R, Heise U. Sonographic assessment of typical
lesions in shoulder dislocation [in German]. Z Orthop Ihre
Grenzgeb 1988;126:570
-573[Medline]
-
Wittner B, Holz U. Ultrasound imaging of the ventrocaudal labrum in
ventral instability of the shoulder [in German].
Unfallchirurg
1996;99:38
-42[Medline]
-
Schydlowsky P, Strandberg C, Galatius A, Gam A. Ultrasonographic
examination of the glenoid labrum of healthy volunteers. Eur J
Ultrasound 1998;8:85
-89[Medline]
-
Schydlowsky P, Strandberg C, Galbo H, Krogsgaard M, Jorgensen U.
The value of ultrasonography in the diagnosis of labral lesions in patients
with anterior shoulder dislocation. Eur J Ultrasound
1998;8:107
-113[Medline]
-
Schydlowsky P, Strandberg C, Tranum-Jensen J, Bojsen-Moller F,
Jorgensen U, Galbo H. Post mortem ultrasonographic assessment of the anterior
glenoid labrum. Eur J Ultrasound
1998;8:129
-133[Medline]
-
Tuite MJ, Orwin JF. Anterosuperior labral variants of the shoulder:
appearance on gradient-recalled-echo and fast spin-echo MR images.
Radiology
1996;199:537
-540[Abstract/Free Full Text]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
|
|
 
C. M. Sofka, R. S. Adler, and M. A. Danon
Sonography of the acetabular labrum: visualization of labral injuries during intra-articular injections.
J. Ultrasound Med.,
October 1, 2006;
25(10):
1321 - 1326.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
C. Martinoli, S. Bianchi, N. Prato, F. Pugliese, M. P. Zamorani, M. Valle, and L. E. Derchi
US of the Shoulder: Non-Rotator Cuff Disorders
RadioGraphics,
March 1, 2003;
23(2):
381 - 401.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
P J O'Connor and A J Grainger
Ultrasound imaging of joint disease
Imaging,
June 1, 2002;
14(3):
188 - 201.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
